Casino Math

Abstract: Casino gaming is one of the most regulated industries in the world. Most gaming
regulatory systems share common objectives: keep the games fair and honest and assure that
players are paid if they win. Fairness and honesty are different concepts. A casino can be
honest but not fair. Casino math is very important to the casino. Casino math is used to
determine the revenue of a casino. All casino games use math to determine the drop or the
expected value of return. The casino has an advantage. Casino games are games of chance. The
probability of winning is in the house favor. The house has at least a 2% advantage over all
casino games. work. Percentages, odds, and probability are the foundation of the gaming
industry. The mathematics underlying the games assures the casinos will make money. Since
games are the products sold by the gaming industry, managers must understand how the
mathematics behind the games produces the expected revenues. Outcomes in games of chance
are, by definition, uncertain. The branch of mathematics used to describe and calculate
uncertainty is called probability.
Key words: Casino Math
Probability reflects how often we expect things to occur and can be interpreted to be the
proportion of times an event will occur in the long run. That is, the probability of an event is the
approximate proportion of times the event will occur in many repetitions of the random
experiment. Probability can be expressed as a fraction, decimal, or percentage. The probability of
heads is 1⁄2, or .50, or 50%. Sometimes probability is referred to as chance. A person may say
that there is a 50% chance of getting heads when tossing a coin.

Odds are another way to express probability. Whereas probability represents the relationship
between the desired event and all possible events, odds describe the relationship between a
desired event and all non-desired events. As a general rule, odds are stated against an event in
casinos.

Permutations refer to the number of different ways in which objects can be arranged in order. In
a permutation, each item can appear only once and each order of the items’ arrangement
constitutes a separate permutation.

Combinations refer to the number of ways that a set of items can be selected from a group,
without regard to the order in which the members of the set are arranged.

Expectation, or expected value, represents how much money a player can expect to win or lose in
the long run on a particular wager. If the player’s expectation is negative, as is typical for most
bets made in a casino, the player can expect to lose money. Expected value is a function of both
the probabilities of winning and losing the bet, as well as the amount won or lost. The following
formula can be used to calculate the expected value of a wager.
House advantage (HA) is the most commonly used indicator of the value of a particular game or
bet. From the casino’s perspective, the house advantage represents how much, in terms of
percentage of the money wagered, the casino can expect to retain.

Volatility analysis is an important tool for the casino manager. If the deviation starts to fall
outside the norm to the casino’s disadvantage, the manager should look for answers. A casino
manager needs to understand what constitutes a normal deviation from the theoretical win and
what should raise the specter of further inquiry. If the deviation starts to fall outside the norm to
the casino’s advantage, the astute regulator should ask questions. The standard deviation tells
how much variation can be expected when observing repeated occurrences of a random variable.
A small standard deviation means the outcomes tend to be similar. A standard deviation of zero,
the smallest possible value, would indicate the outcomes are always exactly the same. A large
standard deviation indicates the outcomes are very different or highly variable.

Those who subscribe to this fallacy believe that the “law of averages” will rescue them in the
midst of a losing streak by turning things around. Players experience, or attribute to themselves,
good luck when they are winning and bad luck when they are losing. As with hot and cold
streaks, however, luck is merely a natural statistical fluctuation in one direction or another –
good luck is a fluctuation of successful bets and bad luck is a fluctuation of unsuccessful bets. It
is a fallacy to think that luck exists as a tangible quantity.

A betting system is a method of placing and varying bets in a predetermined way to try and beat
a game that is favorable to the house and unfavorable to the player. It is a mathematical fact,
however, that betting systems do not work for negative expectation games.33 It is not possible
to change or manipulate the laws of probability and mathematics to turn a negative expectation
game into a positive one. Every such proposed gambling system has at least one fundamental
flaw that prevents the player
from gaining the advantage.

Another way to represent the 95% confidence limits is to report the “plus or minus” part of the
result. For example, the 95% confidence limits for the series of 100,000 wagers on red in
roulette, 4.63% and 5.89%, were derived from the calculation 5.26% ± 0.63%. This means there
is a 95% probability that the actual win percentage will be within 0.63% of the expected win
percentage. That is, 0.63% represents the maximum deviation from the theoretical win
percentage that can be expected (95% of the time) for 100,000 plays. The maximum deviation is
similar to what statisticians call the margin of error. In statistical sampling, the margin of error
represents the maximum likely error attached to a statistic. ordinarily the statistic is being used
to estimate an unknown population parameter. In our present context, the 100,000 wagers
constitute a sample from the population of all such wagers, and the observed win percentage
would serve as an estimate of the theoretical (population) win percentage. Here, however, the
objective is not to estimate an unknown parameter since the relevant parameter, theoretical win
percentage, is known. Rather, we are interested in knowing how unusual an observed win
percentage is, given the known theoretical win percentage. For the 100,000 roulette wagers on
red, an observed win percentage that is not within ±0.63% of the theoretical win percentage is
unusual since the probability of that happening is only 0.05. As long as the observed win
percentage is within ±0.63% of the theoretical win percentage, the difference can be attributed to
normal statistical fluctuations. In this sense, 0.63% serves as a 95% confidence level “margin of
error” for the win percent associated with a series of 100,000 wagers on red. Similarly, the
margin of error for 1,000 wagers is 6.32%, for 10,000 wagers it is 2.00%, and for 1,000,000
wagers it is 0.20%.

Fairness refers to the game advantage. A slot machine that holds, on average, 90% of every
dollar bet is certainly not fair, but could very well be honest (if the outcomes of each play are
not predetermined in the casino’s favor).
Two major regulatory issues relating to fairness and honesty – ensuring random
outcomes and controlling the house advantage – are inextricably tied to mathematics and most
regulatory bodies require some type of mathematical analysis to demonstrate game advantage
and/or confirm that games outcomes are random. Such evidence can range from
straightforward probability analyses to computer simulations and complex statistical studies.
Requirements vary across jurisdictions, but it is not uncommon to see technical language in
gaming regulations concerning specific statistical tests that must be performed, confidence
limits that must be met, and other mathematical specifications and standards relating to game
outcomes.
A chi-square (χ2) statistic is a test that measures how expectations compare to actual observed
data (or model results). The data used in calculating a chi-square statistic must be random, raw,
mutually exclusive, drawn from independent variables, and drawn from a large enough sample.

There are two main kinds of chi-square tests: the test of independence, which asks a question of
relationship, such as, "Is there a relationship between gender and SAT scores?"; and
the goodness-of-fit test, which asks something like "If a coin is tossed 100 times, will it come up
heads 50 times and tails 50 times?"

For these tests, degrees of freedom are utilized to determine if a certain null hypothesis can be

rejected based on the total number of variables and samples within the experiment.

For example, when considering students and course choice, a sample size of 30 or 40 students is
likely not large enough to generate significant data. Getting the same or similar results from a
study using a sample size of 400 or 500 students is more valid.
In another example, consider tossing a coin 100 times. The expected result of tossing a fair coin
100 times is that heads will come up 50 times and tails will come up 50 times. The actual result
might be that heads will come up 45 times and tails will come up 55 times. The chi-square
statistic shows any discrepancies between the expected results and the actual results.

The "mean" is the "average" you are used to, where you add up all the numbers and then divide
by the number of numbers. The "median" is the "middle" value in the list of numbers. To find the
median, your numbers must be listed in numerical order from smallest to largest, so you may
have to rewrite your list before you can find the median. The "mode" is the value that occurs
most often. If no number in the list is repeated, then there is no mode for the list. The "range" of
a list a number is just the difference between the largest and smallest values.
The “probability” includes calculating the probability for events happening. While it’s usual for
the class to include basic scenarios like playing cards and dice rolling at first, these basic tools
are used later in the class to find more complex probabilities, like the probability of contracting a
certain disease. The “statistics” part of probability and statistics includes a wide variety of
methods to find actual statistics, which are numbers you can use to generalize about a
population.

You can calculate the height of all your male classmates and find the mean height to be 5’9″ —
this is a statistic. But then you could take that statistic and say “I think the average height of an
American male is 5’9″ “. How accurate your guess is depending on many factors, including how
many men you measured and how many men are in the entire population. Statistics are useful
because we often do not have the resources to measure, survey or poll every member of a
population, so instead we take a sample (a small amount).

Theory is the branch of mathematics concerned with probability. Although there are several

different probability interpretations, probability theory treats the concept in a rigorous
mathematical manner by expressing it through a set of axioms. Typically these axioms formalize
probability in terms of a probability space, which assigns a measure taking values between 0 and
1, termed the probability measure, to a set of outcomes called the sample space. Any specified
subset of these outcomes is called an event. Central subjects in probability theory include
discrete and continuous random variables, probability distributions, and stochastic processes,
which provide mathematical abstractions of non-deterministic or uncertain processes or
measured quantities that may either be single occurrences or evolve over time in a random
fashion. Although it is not possible to perfectly predict random events, much can be said about
their behavior. Two major results in probability theory describing such behavior are the law of
large numbers and the central limit theorem.
Consider an experiment that can produce a number of outcomes. The set of all outcomes is called
the sample space of the experiment. The power set of the sample space (or equivalently, the
event space) is formed by considering all different collections of possible results. For example,
rolling an honest die produces one of six possible results. One collection of possible results
corresponds to getting an odd number. Thus, the subset {1,3,5} is an element of the power set of
the sample space of die rolls. These collections are called events. In this case, {1,3,5} is the event
that the die falls on some odd number. If the results that occur fall in each event, that event is
said to have occurred.
Probability is a way of assigning every "event" a value between zero and one, with the
requirement that the event made up of all possible results (in our example, the event
{1,2,3,4,5,6}) be assigned a value of one. To qualify as a probability distribution, the assignment
of values must satisfy the requirement that if you look at a collection of mutually exclusive
events (events that contain no common results, e.g., the events {1,6}, {3}, and {2,4} are all
mutually exclusive), the probability that any of these events occurs is given by the sum of the
probabilities of the events.
The probability that any one of the events {1,6}, {3}, or {2,4} will occur is 5/6. This is the same
as saying that the probability of event {1,2,3,4,6} is 5/6. This event encompasses the possibility
of any number except five being rolled. The mutually exclusive event {5} has a probability of
1/6, and the event {1,2,3,4,5,6} has a probability of 1, that is, absolute certainty.
When doing calculations using the outcomes of an experiment, it is necessary that all
those elementary events have a number assigned to them. This is done using a random variable.
A random variable is a function that assigns to each elementary event in the sample space a real
number. This function is usually denoted by a capital letter. In the case of a die, the assignment
of a number to a certain elementary events can be done using the identity function. This does not
always work.
Discrete probability theory deals with events that occur in countable sample spaces.
Examples: Throwing dice, experiments with decks of cards, random walk, and tossing coins
Classical definition: Initially the probability of an event to occur was defined as the number of
cases favorable for the event, over the number of total outcomes possible in an equiprobable
sample space.
For example, if the event is "occurrence of an even number when a die is rolled", the probability
is given by, since 3 faces out of the 6 have even numbers and each face has the same probability
of appearing. So, the probability of the entire sample space is 1, and the probability of the null
event is 0.
The function mapping a point in the sample space to the "probability" value is called
a probability mass function abbreviated as pmf. The modern definition does not try to answer
how probability mass functions are obtained; instead, it builds a theory that assumes their
existence.
Continuous probability distribution
The normal distribution, a continuous probability distribution.
Continuous probability theory deals with events that occur in a continuous sample space.
Classical definition: The classical definition breaks down when confronted with the continuous
case. See Bertrand's paradox.
If  is absolutely continuous, i.e., its derivative exists and integrating the derivative gives us the
cdf back again, then the random variable X is said to have a probability density
function or pdf or simply density for a set. Whereas the pdf exists only for continuous random
variables, the cdf exists for all random variables (including discrete random variables) that take
values. These concepts can be generalized for multidimensional cases on and other continuous
sample spaces.
The raison d'être of the measure-theoretic treatment of probability is that it unifies the discrete
and the continuous cases, and makes the difference a question of which measure is used.
Furthermore, it covers distributions that are neither discrete nor continuous nor mixtures of the
two.
Other distributions may not even be a mix, for example, the Cantor distribution has no positive
probability for any single point, neither does it have a density. The modern approach to
probability theory solves these problems using measure theory to define the probability space:
Given any set (also called sample space) and a σ-algebra on it, a measure  defined on is called
a probability measure.
If is the Borel σ-algebra on the set of real numbers, then there is a unique probability measure
on for any cdf, and vice versa. The measure corresponding to a cdf is said to be induced by the
cdf. This measure coincides with the pmf for discrete variables and pdf for continuous variables,
making the measure-theoretic approach free of fallacies. Along with providing better
understanding and unification of discrete and continuous probabilities, measure-theoretic
treatment also allows us to work on probabilities outside, as in the theory of stochastic processes.
For example, to study Brownian motion, probability is defined on a space of functions.
When it's convenient to work with a dominating measure, the Radon-Nikodym theorem is used
to define a density as the Radon-Nikodym derivative of the probability distribution of interest
with respect to this dominating measure. Discrete densities are usually defined as this derivative
with respect to a counting measure over the set of all possible outcomes. Densities for absolutely
continuous distributions are usually defined as this derivative with respect to the Lebesgue
measure. If a theorem can be proved in this general setting, it holds for both discrete and
continuous distributions as well as others; separate proofs are not required for discrete and
continuous distributions.
A hypothesis test is a hypothesis tests are used to determine if there is enough evidence in a
sample to prove a hypothesis true for the entire population. The test considers two hypotheses.
The null hypothesis is a statement meant to be tested, usually something like "there is no effect"
with the intention of proving this false. The alternate hypothesis is the statement meant to stand
after the test is performed. The two hypotheses must be complementary (one being the negation
of the other).

If there is a correlation between two variables, a pattern can be seen when the variables are
plotted on a scatterplot. If this pattern can be approximated by a line, the correlation is linear.
Otherwise, the correlation is non-linear. The correlations between the variable can be linear, non-
linear or skewed.
Regression analysis is a set of statistical processes for estimating the relationships between
a dependent variable (often called the 'outcome variable') and one or more independent
variables (often called 'predictors', 'covariates', or 'features'). The most common form of
regression analysis is linear regression, in which a researcher finds the line (or a more
complex linear combination) that most closely fits the data according to a specific mathematical
criterion.
Regression analysis is primarily used for two conceptually distinct purposes. First, regression
analysis is widely used for prediction to infer causal relationships between the independent and
dependent variables. Regressions by themselves only reveal relationships between a dependent
variable and a collection of independent variables in a fixed dataset.
In probability theory, the central limit theorem (CLT) establishes that, in many situations, when
independent random variables are added, their properly normalized sum tends toward a normal
distribution (informally a bell curve) even if the original variables themselves are not normally
distributed or skewed from positive to negative to normal. The theorem is a key concept in
probability theory because it implies that probabilistic and statistical methods that work for
normal distributions can be applicable to many problems involving other types of distributions.
Random variables are functions which map from the sample space of a random experiment to
numerical values. The concept of random variables is the odds of getting a certain number. In
roulette the random variable would be the odds of the ball landing on a specific number. In
blackjack the random variable is the odds of getting a certain card. In craps the random
variable is the odds of getting a certain dice combination. The concept of a random variable is
to gambling.
SUMMARY OF FORMULAS & RULES

Probability Rules
♦ Rule 1. The probability of an event will always be between 0 and 1.

♦ Rule 2. (Complement Rule). The probability of an event occurring plus the probability that
event not occurring equals 1.

♦ Rule 3. (Addition Rule). For mutually exclusive events, the probability of at least one of these
events occurring equals the sum of their individual probabilities.

♦ Rule 4A. (Multiplication Rule – independent events). For independent events, the probability
of all of them occurring equals the product of their individual probabilities.

♦ Rule 4B. (Multiplication Rule – dependent events). For events that are not independent, the
probability of all of them occurring equals the product of their conditional probabilities, where
the conditional probability of one event is affected by the event(s) that came before it.

Odds
♣ If probability in favor is X/Y, the odds against are Y–X to X

♣ If the odds against are X to Y, the probability in favor is Y/(X+Y)

Wager Expected Value

( ) EV = ∑ Net Payi × Pi

House Advantage = ×100% Wager

If payoff = x to 1 & true odds = y to 1, then

Wager Variance
[( ) ] 2 VAR = ∑ Net Payi − EV × Pi

Wager Variance (Alternative)

VAR [(Net Pay ) P ] (EV ) = ∑ i × i − Wager Standard Deviation
SD = VAR Wager Expected Value, Per Unit (House Advantage, as decimal)

wager amount EV
EV wager per unit = Wager Standard Deviation, Per Unit

wager amount SD
SD wager per unit = Series Win Percentage
EVwin% = EVper unit

Series Number of Units

units EVper unit EV = n ×units per unit SD = n × SD

Series Win Amount

win EVper unit EV = unit wager × n × win per unit SD = unit wager × n × SD

Volatility Principle

For a large series of independent wagers:

♠ 2/3 of the time actual win will be within 1 SD of expected win,
♠ Only ≈ 5% of the time will actual win be more than 2 SD’s from expected win,
♠ Virtually never (0.3%) will the actual win be more than 3 SD’s from expected win.

Casino managers must understand the very different nature of table games and slot machines
when using mathematics to assess performance. The critical difference is in the amount of
information a casino can accurately record on the play of table games versus the play of slot
machines. Someday, casinos will have the ability to track every transaction that occurs in the
casino, whether it is the amount of each bet at a blackjack table or the quality of a video
blackjack player’s play. But until technology meets these needs in a cost-efficient manner, the
casino must rely on limited or less-than-perfect information to do its analyses. This is
particularly true with table games where the casino does not generally have the advantage of the
advanced player tracking systems.

When a casino player approaches a table game to play, he usually exchanges cash for chips.
After giving the player the chips, the dealer will drop the cash into the drop box attached to the
underside of a table game. Some players are not cash players, but will gamble using credit
instruments or “markers.” A marker is an instrument used to document the extension of credit to
a player. In this case, the player will sign a credit instrument in the amount of the chips received
at the table. Like the cash received, the credit instruments also are placed in the drop box.
Besides cash and credit instruments, the drop box contains foreign chips, and credit and fill slips.
Foreign chips are chips from other casinos played or exchanged by players. A credit slip (not
related to credit extended to a player) is an instrument used to document a transfer of (usually
excess) chips from a table game to the cashier – a game credit. A fill slip is an instrument used to
document a transfer of chips from the cashier to a table game – a fill.

SLOT MACHINES
Slot machines earned their name because they have a slot that accepts coins or tokens for play
from players. Most slot machines today are equipped with a currency acceptor box, or bill
validator, that allows the slot machine to accept currency. A player will insert coins or currency
into the machine to begin play. Most slots have two types of meters: hard meters and soft meters.
The two hard meters, which resemble the odometers in cars before the 1990s, track “coin-in” and
“coin-out,” respectively. Coin-in is the money inserted by the player into the slot machine. Coin-
out is the money that is paid by the slot machine to the player. Coin-in typically goes to the slot
machine hopper, a coin storage bank inside the machine through which all payouts are made. If
the hopper is full, the machine will divert additional coins to the drop bucket stored in a locked
cabinet beneath a slot machine. If the hopper becomes empty, a casino employee will undertake
a “fill” by adding coins or tokens. Slot meters record (1) total credits by coin-in and coin-out, (2)
total credit through the bill validator, (3) total credits, total credits played, total credits won, and
total credits paid, and (4) total games played and total games won.

Handle is the total amount of money wagered or put at risk by a player. A player’s handle is
sometimes referred to as his action. In principle, handle may be calculated using the following
formula:

Handle = Pace × Average Bet × Duration

Pace refers to the speed of the game in decisions per hour in terms of hands dealt, wheel spins,
dice rolls, etc. Average bet refers to the average amount of money wagered per decision.
Handle is the total amount of money wagered or put at risk by a player. A player’s handle is
sometimes referred to as his action. In principle, handle may be calculated using the following
formula:

Handle = Pace × Average Bet × Duration

Pace refers to the speed of the game in decisions per hour in terms of hands dealt, wheel spins,
dice rolls, etc. Average bet refers to the average amount of money wagered per decision.
Duration is the length of time a player plays, in hours (a unit of time other than hours can be used
as long as the pace is also expressed in this unit). Slot machine meters and tracking systems
allow for easy determination of an individual player’s slot handle. For table games, however, it is
impractical to monitor every bet a player makes, or how long he plays, and the speed of the
particular game.

The is playing. Although technology may change this is the future, player table game handle is
currently estimated using reasonable values established through occasional observation of the
player’s average bet and duration of play. Although the speed of a table game is affected by
several factors, including the particular dealer and the number of players at the table, an average
value is often used for game pace.
Drop is the total amount of the contents in a game’s drop box (or drop bucket and currency
acceptor box in the case of slots). For table games, drop consists of the currency and foreign
chips in the drop box plus the value of credit instruments (markers) issued or redeemed at the
table. That is, drop = (cash + markers + chips) contained in the drop box.

For slots, the drop is the total amount of currency and coin in the slot machine currency acceptor
box and drop bucket. That is, drop = (coins + currency) contained in the drop bucket and
currency acceptor box. The slot drop does not include coins or tokens in the slot machine’s
hopper.

For table games, drop and handle are not equivalent. Handle, a theoretically precise notion, is
difficult to measure for table games. Drop, while empirically precise, is a conceptually elusive
term. Many factors can affect a table’s drop and it is not clear exactly what it is that drop
measures in terms of gambling volume.

Win refers to the total amount won by the casino at a table or device. For table games, win is the
amount of the drop minus the change in the table’s chip inventory, including chips issued during
fills and chips removed during credits.

Win = drop – missing chips = drop – (begin chips + fills – credits – end chips).

For slots, win is the amount in the drop bucket and bill validator less any jackpots paid and
accounting for any change in the hopper inventory, including fills.

Win = drop – jackpots – change in hopper inventory.

Theoretical win is the amount the casino expects to win and can be determined as follows:

Theoretical Win = H.A. x Handle

Since handle can be found by multiplying the average bet times the game pace times
the duration of play, theoretical win can be represented as:

Theoretical Win = H.A. x Pace x Avg. Bet x Duration

Theoretical win is also called expected win.

Win percentage is the win divided by handle:

Win Win Percentage = Win/Handle

As with the above description of win, this refers to the actual win percentage, and not theoretical
or expected win percentage. Actual win percentage represents the fraction of money wagered
that is retained by the casino over a given period of time or number of trials. The theoretical win
percentage is the house advantage, or the long-run average percentage of money wagered that is
retained by the casino. For a large number of trials, the actual win percentage should
approximate the theoretical win percentage. For slots and other gaming devices, the actual win
percentage is generally known. For table games, however, win percentage is difficult to measure
since handle is usually unknown. For further discussion on these and other related issues, see the
section later in this chapter on different ways to express win rate.

The percentage of money wagered that the casino can expect to win in the long run is the
theoretical, or expected, win percentage . This value is just the house advantage:

Theoretical Win % = Theoretic Win/Handle = House Advantage

For a large number of trials, the actual win percentage should be close to the theoretical win
percentage.

Hold, or more specifically hold percentage, is the percentage of the drop that is
won by the casino. That is,

Hold % = Win/Drop

Note that Hold percentage is a known measure of the “win rate” for slot machinestable games
and other gaming devices,is used as a rough gauge of how these games are performing. Hold
percentage for table games is represented by game revenue, i.e. the drop plus the change in chip
inventory, divided by drop.

Hold% = Drop + Change in Inventory/Drop

Change in chip inventory can be calculated by taking current inventory, adding any credits, and
subtracting fills and closing inventory. Using the expected win rather than actual win in the hold
percentage formula will yield expected, or theoretical, hold:

Expected Hold % = Expected Win/Drop

Expected Hold % = HA X Pace X Average Bet X Duration/Drop
The volatility index is the basis for calculating how much variation can be expected and for
determining whether an observed hold percentage on a given machine is normal or unusual for a
given number of plays. In general, the larger the volatility index, the greater fluctuations will be
in actual hold percentage. This section examines the derivation and use of this index. Volatility
index can be calculated by multiplying the slot wager standard deviation times the appropriate z
score, where the z score is the standard normal probability value corresponding to the desired
confidence levelIndex: V.I. = zσ

The volatility index and the derivation of the confidence limits are covered on Slot Volatility. A
90% confidence level is commonly used for the slot volatility index, with z score equal to 1.65.
Net Payi = the net payoff per coin wagered = (payout minus number of coins wagered) divided
by coins wagered, EV = house expected value (theoretical edge) for number of coins wagered,

Pi = probability of each Net Pay.

Players will need be aware of some details in order to calculate the probability of a certain event.
First, they will need a full count of the number of individual elements in the sample space. In
addition, they will need to know the number of the individual elements related to the event. Once
a player knows these values, they will be able to define the probability of the event by using the
equation:

P(event) = Size of event / Size of the sample space

What casino players should understand first is the formal theory of probability and are
recommended to start with the term that is known as the “sample space”. The sample space is
something that may occur in a large number of experiments and can normally be understood
easily when considering the nature of the experiment. Still, the sample space does not have to be
described in an explicit way.

Players will need be aware of some details in order to calculate the probability of a certain event.
First, they will need a full count of the number of individual elements in the sample space. In
addition, they will need to know the number of the individual elements related to the event. Once
a player knows these values, they will be able to define the probability of the event by using the
equation:

P(event) = Size of event / Size of the sample space

if a player owns 1 of a total of 4 tickets, their probability to win is 1 in 4. Their odds, on the other
hand, are 3 to 1. In order for a player to convert odds to probability, they need to take the
likelihood for them to win, then use it as the numerator and divide by the total number of
chances, including the ones for both winning and losingexample, let us take that the odds are 4 to
1. Then, the probability would be found by using the following equation:

1 / (1+4) = 1/5
Double-zero roulette (single number bet):
EV = (+35) (1/38) + (-1) (37/38) = -0.053
(House Advantage = 5.3%)
-zero roulette (single number bet):
EV = (+35) (1/37) + (-1) (36/37) = -0.027
(House Advantage = 2.7%)

House advantage and theoretical win percentage are the same thing, hold percentage is win over
drop, win percentage is win over handle, win percentage approaches the house advantage as the
number of plays increases, and hold percentage is equivalent to the win percentage since drop is
for slots but not table games.

· Hold % = Win/Drop
· Win % (actual) = Win/Handle
· House Advantage = Theoretical Win % = Limit (Actual Win %) = Limit (Win/Handle)
· Hold Percentage ¹ House Edge

The house edge shows how much a casino expects to win on any game.

Put simply it's the player's return divided by the same (in principle)player's loss:

House edge = 1 - (return / loss)

The formal theory of probability begins by understanding what’s known as handle. Slot hold
represents the percentage“sample space.” This is simply a description of all money wagered by
players that the machine retains.possible outcomes – everything that can possibly happen.

1. There are 2 outcomes when a coin is tossed; the sample space is {Heads, Tails}.
2. There are 6 outcomes when a single dice is rolled; the sample space is {1, 2, 3, 4, 5, 6}.
3. There are 36 outcomes when two dice are rolled (the first dice and the second dice each
produce a value from 1 to 6, so there are 6 × 6 = 36 outcomes). The sample space is {[1,1], [1,2],
[1,3], [1,4], [1,5], [1,6], [2,1], … and so on}, or more simply, using set builder notation, it is
{[x,y] | 1 ≤ x ≤ 6 , 1 ≤ y ≤ 6 }.
4. There are 38 outcomes when a roulette wheel is spun. The sample space consists of the
numbers 1 through 36, together with zero and double-zero.
5. There are 52 ways a single card can be dealt from a deck. The sample space consists of the set
of card values (rank and suit).
 Theoretic probability is based on math induced into logic thinking. for instance; if there

were three doors. You have a 3:1 chance of picking the door with a prize. When one door is

eliminated your chances change to 2:1. In theory, it is obvious that the door you least expected to

win is the winner because the probability of winning is greater than before. There is a 50%

chance of winning, but the chances of the same number reoccuring more than once is the chance

you are willing to risk when the odds are in your favor. Card counting is a method used to

determine probability theoretic win. When used with true count of the deck and counting cards

will give you a better game strategy against the house. Casino games is a game of chance. The

house has at least a 2% advantage on all the casino games.

In the casino game, Blackjack, there is a procedural process of systematic method of how

to bet from the time you sit at the table until you want to leave. In order to play, the player must

place a bet on the table. The dealer then deals the cards. The player first initial thought is to see if

he has a blackjack (21). Next, the player must look to see if the dealer has an ace showing. If the

dealer does not have an ace showing, the dealer pays the player 2 to 1. If the dealer has an ace

showing, the player should ask for even money and the bet is paid 1 to 1. If the dealer and player

both have blackjack, the bet pushes and nothing happens. If the player does not have a blackjack,

ask yourself if the dealer does? If dealer has an ace showing, take insurance. Bet ½ your original

bet-to-bet insurance. If the dealer does not have blackjack and you have a busting hand (12 or

more), surrender ½ your bet. If the dealer does not have blackjack, the player hits. If your hand

bust, the dealer will take your bet. If you have a good hand and decide to stand on your hand, the

dealer will draw to 17 or more. If the dealer draws over 21, the dealer will pay you even money.

If your hand is higher than the hand, the dealer will take your hand and your bet loses. Blackjack
can be a fun game to play. Your chances of winning lots of money is determined by how much

you want to spend. In most cases, it is best to leave the table while you are ahead. Good luck!

BLACKJACK

The game of Blackjack is played with 1, 2, 6 or 8 decks of cards. An ace is counted as 1

or 11, 10, J, Q and K is counted as 10 and everything else is counted at face value. The player is

dealt two cards. The dealer gets one face card up and one face card down. If the dealer as an Ace

showing and the player does not have blackjack, insurance is available. If the dealer has

blackjack, the player will automatically lose/push the bet or win insurance. If the player has

blackjack, he will win 3:2. The players have a chance to wager their bets before the dealer acts.

The dealer will then compare hands. If the player goes over 21, he loses. If the player stops with

a hand of up to 21, the dealer will take the cards until they reach 17 or more. The person who

comes closest to 21 wins. If the dealer has over 21, you win.

Game strategy is a known mathematical problem, given completely random cards using

basic strategy. Basic strategy is the most commonly proven game by running simulations on a

few million hands and analyzing results. The average house advantage of basic strategy uses

surrender, doubling down and splitting based on the number of decks used in the game.

Number of Decks House Advantage

Single Deck 0.17%
Double Deck 0.46%
Four Decks 0.60%
Six Decks 0.64%
 Eight Decks 0.66%

A counting system can generate an advantage. The casino’s greatest advantage is that the

player goes first (if both bust, the casino wins). The player’s greatest advantage is that the dealer

must hit 12 – 16 (making the hand bust). When there are enough high cards left in the deck, (10 –

35% of the time), the player’s advantage outweighs the casinos advantage. There is a certain

strategy decision that will increase the advantage that depends on the amount of cards remaining

in the deck.

Value of Cards in Single Deck Blackjack Basic Strategy

Removed Card Change in Player Expectation

2 +0.40%
3 +0.43%
4 +0.52%
5 +0.67%
6 +0.45%
7 +0.30%
8 +0.01%
9 -0.15%
Ten -0.51%
Ace -0.59%
The betting correlation is the advantage gained from increasing the bet at the right time.

A playing correlation is the advantage gained from playing the correct strategy at the right time.

There are levels involved in counting systems. The level is how many different values are in the
counting system. A side count determined by counting cards on the side. Some counts are

modified to take into account opportunities to place side bets such as dealer’s bust.

COUNTING SYSTEMS

Strategy 2 3 4 5 6 7 8 9 Ten Ace

Hi +1 +1 +1 +1 +1 0 0 0 -1 -1

Lo
Hi 0 +1 +1 +1 +1 0 0 0 -1 0

Opt

I
Hi +1 +1 +2 +2 +1 +1 0 0 -2 0

Opt

II
KO +1 +1 +1 +1 +1 +1 0 0 -1 -1
Omeg II +1 +1 +2 +2 +2 +1 0 0 -2 -1
Zen +1 +1 +2 +2 +2 +1 0 0 -2 -1

The Hi Lo Basic counting system is the most basic balanced system out there. It requires

running count to true count conversion. It uses a generic strategy that continues throughout the

shoe. This system has a .3% player advantage over the house depending on the number of decks,

cut penetration and level of skill.

 The KO counting system is the most simple un-balanced counting system. The initial

running count for the number of decks is 2 = -4, 6 = -20 and 8 = -24. The pivot points for the

number of decks is 2 = 0, 6 = 4 and 8 = -6.

Casinos implements several techniques to identify counters. Some of the techniques used

are for dealers to call out sudden large bet increases or for dealers to call out extremely

aggressive plays such as doubling on hard 12 or splitting 10s. Cheating in Blackjack has a

consequence of being banned from a casino.

Basic Strategy Expectation

Benchmark -0.02%
One-half deck +0.71%
Two decks -0.32%
Four decks -0.48%
Six decks -0.53%
Eight decks -0.55%
Dealer wins ties -9.34%
Natural pays 1 to -2.32%

1
Natural pays 2 to +2.32%

1
Dealer hits soft 17 -0.20%
 No re-splitting of any pairs -0.03%
Respecting of aces +0.06%
No soft doubling -0.11%
Double down -0.21%

only on 10 or 11
Double down only on 11 -0.69%
Double down on any number of cards +0.24%
Double down after pair splitting +0.13%
Late surrender +0.06%

A good rule for counting is to expect a standard deviation of +500 minimum bets unless

you spread more than 1 – 10. Most spreads are between 1 and 12. The average bet is $2 to $4

depending on the advantage play. Play a $5 minimum with a bankroll of $2500. Do not spend

more than what is intended. Double you bets when the running count is good. Bet according to

basic strategy. Use card counting to determine the amount you bet at a given time. Take

advantage of side bets to know when to double down, split or take insurance. Bet even money on

side bets. Double down on blackjack if the player has an ace showing. Do not spend more than

what is allowed.

Blackjack Risk of Ruin Formula:

The risk of ruin formula predicts the player's expectancy of the amount they win or lose within a

given rate. If the ratio is better when the ratio is above 1%. If the ratio is below 0, you are losing

your money.
K(K+1) * PE + K(K-1) * BE/BR

K = Amount of Bet, 

BE (bet efficiency - standard diviation ratio) = 9%, 

PE (player efficiency or mean ratio) = 6%, 

BR (bank roll) = Amount Planning to Spend, 

(K + 1) = amount earned, 

(K - 1) = amount lost

The word “Probability” is to use the letter “P” when referring to the probability of an event.

What is not evident by the equation for probability is how to count the size of various

collections. The following examples will help clarify the concept of probability. In Poker, there

are 52 possible first cards you can be dealt and 51 possible second cards you can be dealt.

Because the order of the cards does not matter, we divide by 2 to take account of symmetry.

(52 × 51) / 2 = 1326.

In order to have a straight, we have to get one of the hands A23, 234, 345, 456, 567, 678, 789,

89T, 9TJ, TJQ, JQK, QKA in one of the four suits. There are 12 straights and each straight can

happen in one of the four suits, so there are 12 × 4 = 48 ways to get a straight flush. The

probability of getting a straight flush is .002 percent

P(Straight Flush) = 48 / 22,100 = 0.002172

In order to count three of a kind, we have to count these directly. For deuces, we can get trips

with any of the hands [2C,2D,2H], [2C,2D,2S], [2C,2H,2S] and [2D,2H,2S]. There are four trips

for any rank and there are thirteen ranks, so there are 13 × 4 = 52 ways to get three of a kind. The

probability of getting a three of a kind is 0.002 percent.

P(Three of a Kind) = 52 / 22,100 = 0.002353.

In order get a straight we begin by counting all straights, including the straight flushes. As above,

a straight consists of one of the hand types A23, 234, 345, 456, 567, 678, 789, 89T, 9TJ, TJQ,

JQK, QKA. Look at A23. We can have any suits for the A, 2 and 3; there are four choices for the

suit of each (clubs, diamonds, hearts and spades). Therefore the A23 straight consists of picking

a card from each of {AC, AD, AH, AS}, {2C, 2D, 2H, 2S} and {3C, 3D, 3H, 3S}. There are 4 ×

4 × 4 = 64 ways of choosing these three cards. Therefore, there are 64 straights of type A23.

There are 12 types of straights and each straight can happen in 64 ways. Therefore, the total

number of straights is 12 × 64 = 768. However, this number includes the 44 straight flush hands

that we already counted, so we have to subtract those. This leaves 768 – 48 = 720 straights that

are not a straight flush. The probability of getting a straight is 0.03 percent.

P(Straight) = 720 / 22,100 = 0.032579.

In order to get a flush we need to count all the hands XYZ where X, Y and Z have the same suit

and the hand is not a straight flush. We will pick on the suit Clubs. There are 13 cards. We can

choose any of these 13 for the first card, any of the remaining 12 for the second card, and any of

the remaining 11 for the third card. Because three cards can be rearranged in six possible ways

and the order the cards were dealt does not matter, it follows that the number of ways of getting a
flush in clubs is (13 × 12 × 11) / 6 = 286. Of these flushes, 12 of them are a straight flush. This

leaves 286 – 12 = 274 flushes in clubs that are not straight flushes. Because there are four suits, it

follows that the total number of flushes is 274 × 4 = 1,096. The probability of getting a flush is

0.049 percent.

P(Flush) = 1,096 / 22,100 = 0.049593

We determine there are 6 ways of getting a pair of any rank. For example, to get a pair of deuces

the player must hold one of the pairs [2C,2D], [2C,2H], [2C,2S], [2D,2H], [2D,2S] or [2H,2S].

Once the pair is dealt, the third card has to be a card of a different rank (otherwise, the hand

would be trips). There are 48 cards that are not the same rank as the pair. Putting this together,

there are 13 possible ranks for the pair, 6 pairs of that rank, and 48 possible third cards of a

different rank. Multiplying these together, we get a total of 13 × 6 × 48 = 3,744 pairs. The

probability of getting a pair is 0.16 percent.

P(Pair) = 3,744 / 22,100 = 0.169412.

In order to get null hand, it simply means that the hand is not any of the above. We simply

subtract all of the results above from the total number of hands. This gives 22,100 – 48 – 52 –

720 – 1,096 – 3,744 = 16,440 hands that are losing hands for the Pair Plus wager. The

probability of getting a dead hand is 0.74 percent.

P(Nothing) = 16,440 / 22,100 = 0.743891.

If we add all the probabilities together, the probability that something is going to happen is 1, a

certainty.
 The probability of the ball rolling on red or black is determined by the spin of the ball,

the number of rotations in one spin, the location of the ball to the distance of where the ball will

land and the friction of the ball makes against the wheel and hitting its pockets. The probability

of the player profiting after 100 spins could be determined by the expected value to win or how

much the player wagers on each spin. The probability could be determined by the amount of

expected return from winning each spin. The probability could be determined by the number of

times the ball lands on red or black. Even if the player has a good betting strategy, the cost of

playing will not measure the player level of success against the house.

The American Roulette table has 38 slots on a roulette wheel. There are 18 red spots and

18 black spots on a Roulette wheel. If a player made a bet on red or black he has a 50:50 chance

of getting that number. The player has a bankroll of $15 and he expects to win at least $16 in 100

spins. If the player were to bet $1 each spin what would determine the outcome?

On an American Roulette wheel, there are 18 black numbers and 18 red numbers with 2

numbers that are neither. The odds are 18:36 of getting a red or black number and -18:36 odds of

that number not landing on red or black within 100 spins. The permeable probability of that

number landing on the red or black within 100 spins is deducted between 18:37 to 1:36 times of

the ball rotating 100 times.

The linear expectation for that particular bet to occur within 100 spins is determined on

return probability of it happening. There is an expected probability of 18:38 that the ball is going

to land on red or black. There is an expected probability of -20:38 that the ball will not land on

red or black. The expected probability is 5.26 percent.

 What if the player could predict how many spins he will lose? If the player were to lose

his bet four times in a row, the expected value would be 14.77 per cent. For the player to profit

we need to determine the probability of the ball landing on a red or black number. In order to

derive at that conclusion, we must first determine what the expected win is. The player will take

a positive unit 1, subtract the probability from 1, and multiply the number each by the exponent

4. The expected probability is 14.77%.

The probability of getting red or black is 18:38 or 9:19. The random variable number of

getting a red or black in 100 consecutive spins is .47 percent. Nine multiplied by 100 is equal to

900 divided by 19 is equal to 47.37 percent. You will have a 92% chance of walking away with

$16 after 100 spins. If the expected win were $2 for every $1 bet, 2 times 27.37 is equal to

$94.74. The player is expected to win $97.74 if the player bet $2 after 100 spins.

The probability of the ball not landing on either red or black determines the probability of

losing four times in a row. In order to figure this out the player will take 20:38 which fraction is

broken down to 10:19 and the percentage is 0.526%.

The probability of not getting black is .0767 percent. In order to determine this factor

multiply 10:19 four times and you have a losing strategy of 7.7%. Your expected strategy is

equal to the probability of the red or black selected multiplied by the total payout minus the cost

of playing or $0.2277.

SUMMARY OF FORMULAS & RULES

Probability Rules
♦ Rule 1. The probability of an event will always be between 0 and 1.
♦ Rule 2. (Complement Rule). The probability of an event occurring plus the probability that
event not occurring equals 1.

♦ Rule 3. (Addition Rule). For mutually exclusive events, the probability of at least one of these
events occurring equals the sum of their individual probabilities.

♦ Rule 4A. (Multiplication Rule – independent events). For independent events, the probability
of all of them occurring equals the product of their individual probabilities.

♦ Rule 4B. (Multiplication Rule – dependent events). For events that are not independent, the
probability of all of them occurring equals the product of their conditional probabilities, where
the conditional probability of one event is affected by the event(s) that came before it.

Odds
♣ If probability in favor is X/Y, the odds against are Y–X to X

♣ If the odds against are X to Y, the probability in favor is Y/(X+Y)

Wager Expected Value

( ) EV = ∑ Net Payi × Pi

House Advantage = ×100% Wager

If payoff = x to 1 & true odds = y to 1, then

Wager Variance
[( ) ] 2 VAR = ∑ Net Payi − EV × Pi

Wager Variance (Alternative)

VAR [(Net Pay ) P ] (EV ) = ∑ i × i − Wager Standard Deviation
SD = VAR Wager Expected Value, Per Unit (House Advantage, as decimal)

wager amount EV
EV wager per unit = Wager Standard Deviation, Per Unit

wager amount SD
SD wager per unit = Series Win Percentage
EVwin% = EVper unit

Series Number of Units

units EVper unit EV = n ×units per unit SD = n × SD

Series Win Amount

win EVper unit EV = unit wager × n × win per unit SD = unit wager × n × SD

Volatility Principle

For a large series of independent wagers:

♠ 2/3 of the time actual win will be within 1 SD of expected win,
♠ Only ≈ 5% of the time will actual win be more than 2 SD’s from expected win,
♠ Virtually never (0.3%) will the actual win be more than 3 SD’s from expected win.

Win refers to the total amount won by the casino at a table or device. For table games, win is the
amount of the drop minus the change in the table’s chip inventory, including chips issued during
fills and chips removed during credits.

Win = drop – missing chips = drop – (begin chips + fills – credits – end chips).
win is the amount in the drop bucket and bill validator less any jackpots paid and accounting for
any change in the hopper inventory, including fills.

Win = drop – jackpots – change in hopper inventory.

however, where win is the amount the casino expects to win and can be determined as follows:

Theoretical Win = H.A. x Handle

Since handle and drop are not generally equal, holdcan be found by multiplying the average bet
times the game pace times
the duration of play, theoretical win can be represented as:

Theoretical Win = H.A. x Pace x Avg. Bet x Duration

Theoretical win is also called expected win.

Win percentage is not the same as win divided by handle:

Win Percentage = Win/Handle

The percentage. The actual fraction of money wagered at a table game that the casino retains is
notcan expect to win in the hold percentage, but long run is the theoretical, or expected, win
percentage . This value is just the house advantage:

Theoretical Win % = Theoretic Win/Handle = House Advantage

For a large number of trials, the actual win percentage – win divided by handle. As mentioned
above, table game handle is difficult to measure and therefore the should be close to the
theoretical win percentage .
Hold, or more specifically hold percentage, is generally unknown. the percentage of the drop that
is
won by the casino.

Hold % = Win/Drop

Hold percentage is a known measure of the “win rate” for table games and is used as a rough
gauge of how these games are performing. Hold percentage for table games is represented by
game revenue, i.e. the drop plus the change in chip inventory, divided by drop.

Hold% = Drop + Change in Inventory/Drop

Change in chip inventory can be calculated by taking current inventory, adding any credits, and
subtracting fills and closing inventory. Using the expected win rather than actual win in the hold
percentage formula will yield expected, or theoretical, hold:

Expected Hold % = Expected Win/Drop

Expected Hold % = HA X Pace X Average Bet X Duration/Drop

Many factors impact hold percentage by including the theoretical win percentage, the speed of
the game, the average bet size, number of bets and the number of players at the table.

Casinos use hold percentage as a management tool for determining the performance of games.
They use it because other measurements of performance are not readily available. For example, a
comparison of theoretical and actual win percentages could assist casino management in learning
if casino personnel or cheaters at a particular table are stealing. Substantial unexplainable
deviations (either small but prolonged deviations or large and short deviations) may be warnings
of potential problems. Unfortunately, with table games, most casinos do not have the technology
to track actual win percentage because it is a function of the amount wagered on each hand
divided by the amount paid out on each hand. A casino cannot determine either measurement
without sophisticated and costly tracking systems. Therefore, casino management must rely on
more crude measurements to determine if circumstances exist that require special management
intervention.

HANDLE = total amount of money wagered

HANDLE = Pace × Average Bet × Duration
DROP = amount in drop box (or bucket plus bill validator)
DROP = amount of cash or cash equivalents exchanged by players for chips (cheques)
WIN = total amount won by the casino
THEORETICAL WIN = H. A.× Handle
THEORETICAL WIN = H. A. × Pace × Avg. Bet × Duration
WIN % = Win/Handle
Theoretical Win THEORETICAL (EXPECTED) WIN % = Theoretic Win/Handle
THEORETICAL (EXPECTED) WIN % = House Advantage
HOLD % = Win/ Drop
Expected Win EXPECTED HOLD % = Expected Win/Drop
Expected Hold% = HA X Pace X Average Bet X Duration/Drop

HANDLE = total amount of money wagered

HANDLE = Pace × Average Bet × Duration
DROP = amount in drop box (or bucket plus bill validator)
DROP = amount of cash or cash equivalents exchanged by players for chips (cheques)
WIN = total amount won by the casino
THEORETICAL WIN = H. A.× Handle
THEORETICAL WIN = H. A. × Pace × Avg. Bet × Duration
WIN % = Win/Handle
Theoretical Win THEORETICAL (EXPECTED) WIN % = Theoretic Win/Handle
THEORETICAL (EXPECTED) WIN % = House Advantage
HOLD % = Win/ Drop
Expected Win EXPECTED HOLD % = Expected Win/Drop
Expected Hold% = HA X Pace X Average Bet X Duration/Drop

Four types of slot machines are most popular.42 The traditional spinning reel machines occupy
57% of the slot market, compared to video poker with 25%, followed by multi-line, multi-play
video devices with 14%, and wide area progressives with 4%. Most video games are either poker
games or electronic versions of the traditional spinning reel machines. Other less popular video
games are electronic versions of other casino games such as keno and blackjack. Video “reel”
machines often are themed and can be based on board games (Monopoly, Battleship), television
shows (Wheel of Fortune), and rock and roll icons (Elvis). The differences between traditional
spinning reel machines and video “reel” machines are only in the way that the results are
displayed to the player. Since traditional spinning reel machines are no longer mechanical but
controlled by computer microchips like video “reel” machines, they also can be programmed for
virtually any number of “stops,” any hit frequency, and any house advantage within the limits of
applicable regulations. In other words, the symbols on each reel do not have an equal chance of
appearing on the pay line. Instead, the computer in the slot machine produces a random number,
each of which corresponds to a particular combination of symbols on the pay line of the slot
machine. Once the random number is generated, the computer instructs the slot machine to stop
the reels so that they display the combination of symbols that corresponds to the random number.

Multi-line, multi-play video devices are clearly becoming more popular, taking shares from the
raditional spinning reel. These machines come in a wide variety of types depending on how the
games pay winning players. Line games allow the player to activate additional pay lines by
playing more coins. In a typical reel slot, for example, three horizontal lines and the diagonals
might be activated to pay back with winning combinations. With today’s video slots, many more
pay lines can be activated. Multipliers pay for winning combinations on a single line only –
usually the center horizontal line. Payback for additional coins is simply a multiple of the
payback for a single coin. Multi-line, multi-play video games combine both multipliers and line
games. Buy-a-Pay games allow the player to “buy” additional winning symbols with extra coins.
For example, one coin inserted may only pay out for three cherry symbols, two coins will pay for
three cherries or three bars, and three coins will pay for three cherries, three bars, or three 7’s.
House advantage for slot machines is oftentimes referred to as the theoretical win percentage or
theoretical hold. This value represents the long-term percentage of money wagered that the
casino retains (wins). The flip side of the house advantage, the proportion of wagered money
returned to the player, is the payback percentage. For example, if a machine holds 5% (about
average for Nevada slot machines) the payback percentage is 95%.

Hold percentage is a theoretical amount that can be set to virtually any value with the computer
program that controls the machine. Whatever this theoretical hold, the actual percentage will
vary somewhat but will be close to the theoretical amount after a large number of plays.
Assessing the likelihood of a certain difference between actual and theoretical win is possible
because of the random selection of outcomes. The volatility section that follows later in this
chapter addresses how much disparity can be expected between actual and theoretical win for
slot machines.

Today’s computer controlled gaming devices depend on random number generators (RNG’s) to
select a “random” symbol combination and ensure the game’s fairness. In jurisdictions like
Nevada, regulations require machines to be tested to ensure they are sufficiently random.43 One
property of randomness that is particularly important for casino games is unpredictability.
Random does not mean totally
43 Since the computer must be programmed to select the combination, the RNG produces
pseudo-random numbers that act like random numbers. To determine whether a series of
numbers possesses random properties, mathematicians apply numerous statistical tests for
randomness Some of these tests include the runs, goodness-of-fit, and various other chi-squared
tests. Because of the law of large numbers, it is possible to make predictions about what will
happen in the long run in a completely random game – this is, in fact, an appealing feature of
certain types of random phenomenon.

The theoretical hold percentage varies across slot machines quite widely, as much as between
0.5% and 30% (some video poker machines if played with expert strategy will even pay back
over 100% – a negative house advantage), and depends on the type of machine, denomination,
and even the jurisdiction. Generally the higher denomination machines have lower house
advantages.

Hit frequency is the percentage of time the machine pays something to the player. Hit frequency
is important in assessing whether a particular slot machine will appeal to a particular audience.
There is, however, a tradeoff between hit frequency and the amount of the payout. For a given
payback percentage, hit frequency and average payout amount per hit are inversely related. One
machine can pay out small amounts quite frequently while another with the same theoretical
payback percentage may pay out large amounts very infrequently. From the casino’s perspective,
these two machines will generate the same amount of revenue in the long run, but to the player
they will behave quite differently. Casino management must decide what balance they need to
strike between hit frequency and payout amounts, and the mix of types of machines to offer to
the playing public.

A par sheet is a document prepared by a slot machine manufacturer that shows the possible
outcomes from the play of a slot machine, the probability of occurrence of each, and the
contribution of each winning outcome to the payback percentage. It also includes other statistical
information such as the hit frequency, volatility index, and confidence limits for actual payback
percentage. Par sheets are sometimes referred to as PC sheets, game sheets, specification sheets,
or theoretical hold worksheets.

The volatility index is the basis for calculating how much variation can be expected and for
determining whether an observed hold percentage on a given machine is normal or unusual for a
given number of plays. In general, the larger the volatility index, the greater fluctuations will be
in actual hold percentage. This section examines the derivation and use of this index. Volatility
index can be calculated by multiplying the slot wager standard deviation times the appropriate z
score, where the z score is the standard normal probability value corresponding to the desired
confidence level.

Volatility Index: V.I. = zσ

where z equals the z score for the desired confidence level, and σ is the slot machine’s per coin
standard deviation. The calculation of the standard deviation was discussed in detail in Chapter
1. For a slot machine, the per-coin standard deviation can be computed by: The volatility index
and the derivation of the confidence limits are covered in the section on Slot Volatility. A 90%
confidence level is commonly used for the slot volatility index, with z score equal to 1.65. Net
Payi = the net payoff per coin wagered = (payout minus number of coins wagered) divided by
coins wagered, EV = house expected value (theoretical edge) for number of coins wagered,

Pi = probability of each Net Pay.

Players will need be aware of some details in order to calculate the probability of a certain event.
First, they will need a full count of the number of individual elements in the sample space. In
addition, they will need to know the number of the individual elements related to the event. Once
a player knows these values, they will be able to define the probability of the event by using the
equation:

P(event) = Size of event / Size of the sample space

casino players should understand first is the formal theory of probability and are recommended
to start with the term that is known as the “sample space”. The sample space is something that
may occur in a large number of experiments and can normally be understood easily when
considering the nature of the experiment. Still, the sample space does not have to be described in
an explicit way.

Players will need be aware of some details in order to calculate the probability of a certain event.
First, they will need a full count of the number of individual elements in the sample space. In
addition, they will need to know the number of the individual elements related to the event. Once
a player knows these values, they will be able to define the probability of the event by using the
equation:

P(event) = Size of event / Size of the sample space

if a player owns 1 of a total of 4 tickets, their probability to win is 1 in 4. Their odds, on the other
hand, are 3 to 1. In order for a player to convert odds to probability, they need to take the
likelihood for them to win, then use it as the numerator and divide by the total number of
chances, including the ones for both winning and losing.

take that the odds are 4 to 1. Then, the probability would be found by using the following
equation:

1 / (1+4) = 1/5

-zero roulette (single number bet):

EV = (+35) (1/38) + (-1) (37/38) = -0.053

(House Advantage = 5.3%)

Single-zero roulette (single number bet):

EV = (+35) (1/37) + (-1) (36/37) = -0.027
(House Advantage = 2.7%)

advantage and theoretical win percentage are the same thing, hold percentage is win over drop,
win percentage is win over handle, win percentage approaches the house advantage as the
number of plays increases, and hold percentage is equivalent to win percentage for slots but not
table games.

· Hold % = Win/Drop

· Win % (actual) = Win/Handle

· House Advantage = Theoretical Win % = Limit (Actual Win %) = Limit (Win/Handle)

· Hold Percentage ¹

House Edge
The house edge shows how much a casino expects to win on any game.

Put simply it's the player's return divided by the player's loss:

House edge = 1 - (return / loss)

The formal theory of probability begins by understanding what’s known as the “sample space.”
This is simply a description of all possible outcomes – everything that can possibly happen.

1. There are 2 outcomes when a coin is tossed; the sample space is {Heads, Tails}.

2. There are 6 outcomes when a single dice is rolled; the sample space is {1, 2, 3, 4, 5, 6}.

3. There are 36 outcomes when two dice are rolled (the first dice and the second dice each
produce a value from 1 to 6, so there are 6 × 6 = 36 outcomes). The sample space is {[1,1], [1,2],
[1,3], [1,4], [1,5], [1,6], [2,1], … and so on}, or more simply, using set builder notation, it is
{[x,y] | 1 ≤ x ≤ 6 , 1 ≤ y ≤ 6 }.

4. There are 38 outcomes when a roulette wheel is spun. The sample space consists of the
numbers 1 through 36, together with zero and double-zero.

5. There are 52 ways a single card can be dealt from a deck. The sample space consists of the set
of card values (rank and suit).

Theoretic probability is based on math induced into logic thinking. for instance; if there

were three doors. You have a 3:1 chance of picking the door with a prize. When one door is

eliminated your chances change to 2:1. In theory, it is obvious that the door you least expected to

win is the winner because the probability of winning is greater than before. There is a 50%
chance of winning, but the chances of the same number reoccuring more than once is the chance

you are willing to risk when the odds are in your favor. Card counting is a method used to

determine probability theoretic win. When used with true count of the deck and counting cards

will give you a better game strategy against the house. Casino games is a game of chance. The

house has at least a 2% advantage on all the casino games.

In the casino game, Blackjack, there is a procedural process of systematic method of how

to bet from the time you sit at the table until you want to leave. In order to play, the player must

place a bet on the table. The dealer then deals the cards. The player first initial thought is to see if

he has a blackjack (21). Next, the player must look to see if the dealer has an ace showing. If the

dealer does not have an ace showing, the dealer pays the player 2 to 1. If the dealer has an ace

showing, the player should ask for even money and the bet is paid 1 to 1. If the dealer and player

both have blackjack, the bet pushes and nothing happens. If the player does not have a blackjack,

ask yourself if the dealer does? If dealer has an ace showing, take insurance. Bet ½ your original

bet-to-bet insurance. If the dealer does not have blackjack and you have a busting hand (12 or

more), surrender ½ your bet. If the dealer does not have blackjack, the player hits. If your hand

bust, the dealer will take your bet. If you have a good hand and decide to stand on your hand, the

dealer will draw to 17 or more. If the dealer draws over 21, the dealer will pay you even money.

If your hand is higher than the hand, the dealer will take your hand and your bet loses. Blackjack

can be a fun game to play. Your chances of winning lots of money is determined by how much

you want to spend. In most cases, it is best to leave the table while you are ahead. Good luck!

BLACKJACK
The game of Blackjack is played with 1, 2, 6 or 8 decks of cards. An ace is counted as 1 or 11,

10, J, Q and K is counted as 10 and everything else is counted at face value. The player is dealt

two cards. The dealer gets one face card up and one face card down. If the dealer as an Ace

showing and the player does not have blackjack, insurance is available. If the dealer has

blackjack, the player will automatically lose/push the bet or win insurance. If the player has

blackjack, he will win 3:2. The players have a chance to wager their bets before the dealer acts.

The dealer will then compare hands. If the player goes over 21, he loses. If the player stops with

a hand of up to 21, the dealer will take the cards until they reach 17 or more. The person who

comes closest to 21 wins. If the dealer has over 21, you win.

Game strategy is a known mathematical problem, given completely random cards using

basic strategy. Basic strategy is the most commonly proven game by running simulations on a

few million hands and analyzing results. The average house advantage of basic strategy uses

surrender, doubling down and splitting based on the number of decks used in the game.

Number of Decks House Advantage

Single Deck 0.17%
Double Deck 0.46%
Four Decks 0.60%
Six Decks 0.64%
Eight Decks 0.66%

A counting system can generate an advantage. The casino’s greatest advantage is that the player

goes first (if both bust, the casino wins). The player’s greatest advantage is that the dealer must

hit 12 – 16 (making the hand bust). When there are enough high cards left in the deck, (10 – 35%
of the time), the player’s advantage outweighs the casinos advantage. There is a certain strategy

decision that will increase the advantage that depends on the amount of cards remaining in the

deck.

Value of Cards in Single Deck Blackjack Basic Strategy

Removed Card Change in Pla

2 +0.40%
3 +0.43%
4 +0.52%
5 +0.67%
6 +0.45%
7 +0.30%
8 +0.01%
9 -0.15%
Ten -0.51%
Ace -0.59%
The betting correlation is the advantage gained from increasing the bet at the right time. A

playing correlation is the advantage gained from playing the correct strategy at the right time.

There are levels involved in counting systems. The level is how many different values are in the

counting system. A side count determined by counting cards on the side. Some counts are

modified to take into account opportunities to place side bets such as dealer’s bust.

COUNTING SYSTEMS

Strate 2 3 4 5 6 7 8 9 Ten Ace

 gy
Hi Lo +1 +1 +1 +1 +1 0 0 0 -1 -1
Hi 0 +1 +1 +1 +1 0 0 0 -1 0

Opt I
Hi +1 +1 +2 +2 +1 +1 0 0 -2 0

Opt II
KO +1 +1 +1 +1 +1 +1 0 0 -1 -1
Omeg +1 +1 +2 +2 +2 +1 0 0 -2 -1

II
Zen +1 +1 +2 +2 +2 +1 0 0 -2 -1

The Hi Lo Basic counting system is the most basic balanced system out there. It requires running

count to true count conversion. It uses a generic strategy that continues throughout the shoe. This

system has a .3% player advantage over the house depending on the number of decks, cut

penetration and level of skill.

The KO counting system is the most simple un-balanced counting system. The initial running

count for the number of decks is 2 = -4, 6 = -20 and 8 = -24. The pivot points for the number of

decks is 2 = 0, 6 = 4 and 8 = -6.

ROULETTE

Roulette is a simple game to play: players merely try to guess which number will occur as the
outcome of the spin of a ball around a numbered wheel. There are two types of wheels, the
“double-zero,” common in the United States, and the “single-zero,” favored in Europe. The
double-zero wheel contains 38 pockets, numbered 1 through 36, 0 and 00. Of those numbered 1-
36, 18 are red and 18 are black. The 0 and 00 are green. The single-zero wheel lacks the 00, and
so has only 37 numbers. The game is played essentially the same regardless of which wheel is
used, although the house advantage for the single-zero wheel is about half that of the double-zero
game (2.7% vs. 5.3%). Except where otherwise specified, the discussion and analysis that follow
are based on the double-zero game. To play the game, players wager on a number, or
combination of numbers. The following table shows the types of bets available, with their
payoffs and true odds.

Roulette (Double-Zero Wheel)

Type of Bet True Odds Payoff

Straight-up (1 number) 37 to 1 35 to 1
Split (2 numbers) 36 to 2 17 to 1
Street (3 numbers) 35 to 3 11 to 1
Corner (4 numbers) 34 to 4 8 to 1
Five-Number (5 numbers) 33 to 5 6 to 1
Double Street or Line (6 numbers) 32 to 6 5 to 1
Dozens/Columns (12 numbers) 26 to 12 2 to 1
Red/Black/Odd/Even/High/Low (18 numbers) 20 to 18 1 to 1

The dealer spins the wheel counterclockwise and releases a small ball clockwise on a track on
the upper portion of the wheel. When the ball loses its momentum, it drops, bounces, and settles
into one of the numbered pockets. Players win or lose depending on whether the winning number
was among those they selected with their wager. As can be seen from the table above, the lower
the probability of winning a particular wager, the higher the payoff. The casino makes money
because the payoff odds are less than the true odds.

A balanced wheel, spins constitute independent trials. Every bet in the double-zero game will
bring the house 5.26% in the long run, with the one exception of the five-number bet. This
anomaly House Advantage:

5.3% (double-zero)
2.7% (single-zero)

Typical Hold: 24% (Nevada)

It holds a 7.89% house advantage. Thus, the casino can expect to win about 5.3% of all money
wagered in double-zero roulette. To see how the mathematics behind the game of roulette works,
consider a $1 straight-up bet on a single number. The probability of winning this wager is 1/38
(since there are 38 numbers on a double-zero wheel), and the probability of losing is 37/38.
Notice that on the average this wager will win once and lose 37 times for every 38 spins (hence
the 37 to 1 odds in the preceding table). To break even, the player would have to be paid $37 for
the one win to compensate for the 37 times he lost. That is, the true odds are 37:1. The payoff,
however, is only 35:1, resulting in a loss to the player of $2, or 5.26% of the $38 wagered over
the 38 spins. This 5.26% house advantage can be computed using the formula for expected
value: 55 EV = (+$35)(1/38) + (−$1)(37/38) = −$0.0526.

This expected loss of $0.0526 on a $1 wager represents a house advantage of 5.26%. This, of
course, is the same percentage edge that would be obtained regardless of the size of the wager. A
$5 single number bet, for example, has a player expectation of −$0.263, or 5.26% of $5.
Expectations and advantage for other types of roulette wagers can be computed in the same way.
Except for the five-number bet, all will result in a 5.26% house edge.

$25

Corner Bet
(Payoff = 8:1 = $200, Probability = 4/38).
EV = (+$200)(4/38) + (−$25)(34/38) = −$1.316.
House advantage = $1.316/$25 = 0.0526, or 5.26%

$50 Five-Number Bet

(Payoff = 6:1 = $300, Probability = 5/38)
Expected Value = (+$300)(5/38) + (−$50)(33/38) = −$3.947.
House advantage = $3.947/$50 = 0.0789, or 7.89%.
No pattern or combination of bets will alter the casino’s 5.3% advantage.56 As an illustration,
suppose a player makes a combination bet totaling $40 by putting $20 on red, $10 on column 2,
and $5 each on 0 and 00. The possible outcomes are:

♦ Lose $40, with probability 10/38 (if a black number in columns 1 or 3 occurs
– lose all bets),

♦ Lose $10, with probability 8/38 (if a black number in column 2 occurs – win
$20 for column 2 bet, lose $30 for the red, 0, and 00 bets),

♦ Lose $0, with probability 14/38 (if a red number in columns 1 or 3 occurs –
win $20 for the red bet, lose $20 for the 0, 00, and column 2 bets),

♦ Win $30, with probability 4/38 (if a red number in column 2 occurs – win $20
for the red bet, win $20 for the column 2 bet, lose $10 for the 0 and 00 bets),

♦ Win $140, with probability 2/38 (if either 0 or 00 occurs – win $175 for the
winning single number bet, lose $35 for the red, column 2 and losing single
number bets).
The expected value is: Dividing by the total amount wagered, $40, yields a 5.26% overall house
advantage. A common “gambler’s fallacy” is to believe that black is more likely after a series of
consecutive reds (black is “due”) while another is that since red is “hot,” red is more likely on
the next spin. Both, of course, are wrong. The spins of a roulette wheel are random, independent
events, and so the outcome of one spin does not affect the outcome of any other spin. If a long
string of consecutive red outcomes has occurred, the next spin is no more or less likely to be red.
The probability of red is 18/38 on each and every spin.

A flaw in the gambler’s fallacy logic is mistaking the short run for the long run. While it is true
that the probability of, say, ten consecutive reds is small, over the course of millions and millions
of spins there will be many such sequences. If one were to examine all sequences of ten
consecutive reds over the course of millions of spins, it would turn out that for about 47.37%
(18/38) of these sequences, the next spin in the sequence (the eleventh) would be a black
number, and for about 47.37% the next spin would be red (the remaining 5.26% would be green).

The actual percentage won in a series of wagers will deviate from the expected win percentage
due to statistical fluctuations. The degree of volatility depends on the number of wagers in the
series and the type of wager. In roulette, the greatest variation can be expected for the straight-up
bet on a single
number and the least variation for even-money wagers. Other roulette wagers have standard

deviation values between these two extremes. All else equal, the larger the standard deviation,

the greater the volatility and fluctuations that can be expected in a series of wagers.

CRAPS

Craps is played by betting on the outcome of aimplements several techniques to identify

counters. Some of the techniques used are for dealers to call out sudden large bet increases or for

dealers to call out extremely aggressive plays such as doubling on hard 12 or splitting 10s.

Cheating in Blackjack has a consequence of being banned from a casino.

Basic Strategy Expectation

 Benchmark -0.02%
One-half deck +0.71%
Two decks -0.32%
Four decks -0.48%
Six decks -0.53%
Eight decks -0.55%
Dealer wins ties
Natural pays 1 to 1 -2.32%
Natural pays 2 to 1 +2.32%
Dealer hits soft 17 -0.20%
No re-splitting of any pairs -0.03%
Respecting of aces +0.06%
No soft doubling -0.11%
Double down only on 10 or 11 -0.21%
Double down only on 11 -0.69%
Double down on any number of cards +0.24%
Double down after pair splitting +0.13%
Late surrender +0.06%

A good rule for counting is to expect a standard deviation of +500 minimum bets unless you

spread more than 1 – 10. Most spreads are between 1 and 12. The average bet is $2 to $4

depending on the advantage play. Play a $5 minimum with a bankroll of $2500. Do not spend

more than what is intended. Double you bets when the running count is good. Bet according to

basic strategy. Use card counting to determine the amount you bet at a given time. Take

advantage of side bets to know when to double down, split or take insurance. Bet even money on

side bets. Double down on blackjack if the player has an ace showing. Do not spend more than

what is allowed.

Blackjack Risk of Ruin Formula:

The risk of ruin formula predicts the player's expectancy of the amount they win or lose within a

given rate. If the ratio is better when the ratio is above 1%. If the ratio is below 0, you are losing

your money.

K(K+1) * PE + K(K-1) * BE/BR

K = Amount of Bet, 

BE (bet efficiency - standard diviation ratio) = 9%, 

PE (player efficiency or mean ratio) = 6%, 

BR (bank roll of a pair of dice. There are a myriad of) = Amount Planning to Spend, 

(K + 1) = amount earned, 

(K - 1) = amount lost

The word “Probability” is to use the letter “P” when referring to the probability of an event.

What is not evident by the equation for probability is how to count the size of various

collections. The following examples will help clarify the concept of probability. In Poker, there

are 52 possible wagers that can be made,first cards you can be dealt and 51 possible second cards

you can be dealt. Because the order of the cards does not matter, we divide by 2 to take account

of symmetry.

(52 × 51) / 2 = 1326.

In order to have a straight, we have to get one of the hands A23, 234, 345, 456, 567, 678, 789,

89T, 9TJ, TJQ, JQK, QKA in one of the four suits. There are 12 straights and each straight can

happen in one of the four suits, so there are 12 × 4 = 48 ways to get a straight flush. The

probability of getting a straight flush is .002 percent

P(Straight Flush) = 48 / 22,100 = 0.002172

In order to count three of a kind, we have to count these directly. For deuces, we can get trips

with any of the hands [2C,2D,2H], [2C,2D,2S], [2C,2H,2S] and [2D,2H,2S]. There are four trips

for any rank and there are thirteen ranks, so there are 13 × 4 = 52 ways to get three of a kind. The

probability of getting a three of a kind is 0.002 percent.

P(Three of a Kind) = 52 / 22,100 = 0.002353.

In order get a straight we begin by counting all straights, including the straight flushes. As above,

a straight consists of one of the hand types A23, 234, 345, 456, 567, 678, 789, 89T, 9TJ, TJQ,

JQK, QKA. Look at A23. We can have any suits for the A, 2 and 3; there are four choices for the

suit of each (clubs, diamonds, hearts and spades). Therefore the A23 straight consists of picking

a card from each of {AC, AD, AH, AS}, {2C, 2D, 2H, 2S} and {3C, 3D, 3H, 3S}. There are 4 ×

4 × 4 = 64 ways of choosing these three cards. Therefore, there are 64 straights of type A23.

There are 12 types of straights and each straight can happen in 64 ways. Therefore, the total

number of straights is 12 × 64 = 768. However, this number includes the 44 straight flush hands

that we already counted, so we have to subtract those. This leaves 768 – 48 = 720 straights that

are not a straight flush. The probability of getting a straight is 0.03 percent.

P(Straight) = 720 / 22,100 = 0.032579.

In order to get a flush we need to count all the hands XYZ where X, Y and Z have the same suit

and the hand is not a straight flush. We will pick on the suit Clubs. There are 13 cards. We can

choose any of these 13 for the first card, any of the remaining 12 for the second card, and any of

the remaining 11 for the third card. Because three cards can be rearranged in six possible ways

and the order the cards were dealt does not matter, it follows that the number of ways of getting a

flush in clubs is (13 × 12 × 11) / 6 = 286. Of these flushes, 12 of them are a straight flush. This

leaves 286 – 12 = 274 flushes in clubs that are not straight flushes. Because there are four suits, it

follows that the total number of flushes is 274 × 4 = 1,096. The probability of getting a flush is

0.049 percent.

P(Flush) = 1,096 / 22,100 = 0.049593

We determine there are 6 ways of getting a pair of any rank. For example, to get a pair of deuces

the player must hold one of the pairs [2C,2D], [2C,2H], [2C,2S], [2D,2H], [2D,2S] or [2H,2S].

Once the pair is dealt, the third card has to be a card of a different rank (otherwise, the hand

would be trips). There are 48 cards that are not the same rank as the pair. Putting this together,

there are 13 possible ranks for the pair, 6 pairs of that rank, and 48 possible third cards of a

different rank. Multiplying these together, we get a total of 13 × 6 × 48 = 3,744 pairs. The

probability of getting a pair is 0.16 percent.

P(Pair) = 3,744 / 22,100 = 0.169412.

In order to get null hand, it simply means that the hand is not any of the above. We simply

subtract all of the results above from the total number of hands. This gives 22,100 – 48 – 52 –
720 – 1,096 – 3,744 = 16,440 hands that are losing hands for the Pair Plus wager. The

probability of getting a dead hand is 0.74 percent.

P(Nothing) = 16,440 / 22,100 = 0.743891.

If we add all the probabilities together, the probability that something is going to happen is 1, a

certainty.

The probability of the ball rolling on red or black is determined by the spin of the ball,

the number of rotations in one spin, the location of the ball to the distance of where the ball will

land and the friction of the ball makes against the wheel and hitting its pockets. The probability

of the player profiting after 100 spins could be determined by the expected value to win or how

much the player wagers on each spin. The probability could be determined by the amount of

expected return from winning each spin. The probability could be determined by the number of

times the ball lands on red or black. Even if the player has a good betting strategy, the cost of

playing will not measure the player level of success against the house.

The American Roulette table has 38 slots on a roulette wheel. There are 18 red spots and

18 black spots on a Roulette wheel. If a player made a bet on red or black he has a 50:50 chance

of getting that number. The player has a bankroll of $15 and he expects to win at least $16 in 100

spins. If the player were to bet $1 each spin what would determine the outcome?

On an American Roulette wheel, there are 18 black numbers and 18 red numbers with

different 2 numbers that are neither. The odds are 18:36 of getting a red or black number and

-18:36 odds, payoff and of that number not landing on red or black within 100 spins. The
permeable probability of that number landing on the red or black within 100 spins is deducted

between 18:37 to 1:36 times of the ball rotating 100 times.

The linear expectation for that particular bet to occur within 100 spins is determined on

return probability of it happening. There is an expected probability of 18:38 that the ball is going

to land on red or black. There is an expected probability of -20:38 that the ball will not land on

red or black. The expected probability is 5.26 percent.

What if the player could predict how many spins he will lose? If the player were to lose

his bet four times in a row, the expected value would be 14.77 per cent. For the player to profit

we need to determine the probability of the ball landing on a red or black number. In order to

derive at that conclusion, we must first determine what the expected win is. The player will take

a positive unit 1, subtract the probability from 1, and multiply the number each by the exponent

4. The expected probability is 14.77%.

The probability of getting red or black is 18:38 or 9:19. The random variable number of

getting a red or black in 100 consecutive spins is .47 percent. Nine multiplied by 100 is equal to

900 divided by 19 is equal to 47.37 percent. You will have a 92% chance of walking away with

$16 after 100 spins. If the expected win were $2 for every $1 bet, 2 times 27.37 is equal to

$94.74. The player is expected to win $97.74 if the player bet $2 after 100 spins.

The probability of the ball not landing on either red or black determines the probability of

losing four times in a row. In order to figure this out the player will take 20:38 which fraction is

broken down to 10:19 and the percentage is 0.526%.

The probability of not getting black is .0767 percent. In order to determine this factor

multiply 10:19 four times and you have a losing strategy of 7.7%. Your expected strategy is
equal to the probability of the red or black selected multiplied by the total payout minus the cost

of playing or $0.2277.

The house advantage. To understand these wagers, familiarity with the probabilities associated
with the various outcomes of tossing a pair of dice is helpful. The 36 possible outcomes and the
probabilities of each possible total are given in the following table.

Outcomes When Rolling A Pair Of Dice Total Probability

(1,1) 2 1/36
(1,2) (2,1) 3 2/36
(1,3) (2,2) (3,1) 4 3/36
(1,4) (2,3) (3,2) (4,1) 5 4/36
(1,5) (2,4) (3,3) (4,2) (5,1) 6 5/36
(1,6) (2,5) (3,4) (4,3) (5,2) (6,1) 7 6/36
(2,6) (3,5) (4,4) (5,3) (6,2) 8 5/36
(3,6) (4,5) (5,4) (6,3) 9 4/36
(4,6) (5,5) (6,4) 10 3/36
(5,6) (6,5) 11 2/36
(6,6) 12 1/36

With more combinations (six) yielding a total of seven than any other, seven is the most likely
total to be thrown, with probability 6/36, 1/6, or 16.67%. Working outwards from seven in either
direction, the totals become progressively less likely until you get to two and twelve, the least
likely outcomes, each with probability 1/36 or 2.78%. This effect is portrayed in the above table
by the pyramid shaped list of
outcomes.

The main bets in craps are the pass line, the come, the don’t pass, the don’t come and odds.
When a game or new round of play begins, the roll is called the “come-out” roll. A bet on the
pass line wins even money if the come-out roll is a 7 or 11, and loses if it is a 2, 3, or 12 (a roll of
2, 3, or 12 is called “craps”). If any other number is rolled, this number becomes the “point.”
Once a point is established, the pass wager is not resolved until the point is rolled again or a 7 is
rolled. If the point is rolled a second House Advantage:

1.41% (pass/come, no odds)

0.61% (pass/come, 2X odds)
1.52% – 16.67% (other bets)

Typical Hold: 14% (Nevada)

is always going to make a profit after every bet made. The house advantage on the pass line bet
is 1.414%. To see how this is calculated, note that there are two ways to win this bet – winning
the come-out roll or making the point. Since there are six possible points (4, 5, 6, 8, 9, and 10),
the number of ways to win the pass line bet can be further broken down into seven possibilities:

(1) win the come-out roll,

or (2) establish and make the point 4,
or (3) establish and make the point 5,
or (4) establish and make the point 6,
or (5) establish and make the point 8,
or (6) establish and make the point 9,
or (7) establish and make the point 10.

Similarly, the pass line bet could be lost in seven distinct ways. To calculate the expectation, the
probabilities for each of these outcomes are needed. Since there are eight ways to roll a 7 or 11
winning numbers for the pass line bet), the probability of winning the come out roll is 8/36, or .
222222. Since there are 4 ways to roll a 2, 3, or 12, the probability of losing the come-out roll is
4/36, or .111111. The probability of establishing and making a point depends on the particular
point number. For example, the probability of establishing 4 as the point is 3/36 (three ways to
roll a 4), and once established, the probability of making (passing) that point is 3/9 (there are six
ways to roll a 7 and three ways to roll a 4). Thus the probability of establishing then making the
point 4 is (3/36)(3/9) = 1/36. The probability of establishing then not making the point four is
(3/36)(6/9) = 2/36. Similar reasoning can be used to find the probabilities associated with the
points 5, 6, 8, 9, and 10. The following table summarizes the probabilities:

Establish then Pass Establish then don’t pass

4 3/36 3/36 x 3/9 ≈ .027778 3/36 x 6/9 ≈ .055556
5 4/36 4/36 x 4/10 ≈ .044444 4/36 x 6/10 ≈ .066667
6 5/36 5/36 x 5/11 ≈ .063131 5/36 x 6/11 ≈ .075758
8 5/36 5/36 x 5/11 ≈ .063131 5/36 x 6/11 ≈ .075758
9 4/36 4/36 x 4/10 ≈ .044444 4/36 x 6/10 ≈ .066667
10 3/36 3/36 x 3/9 ≈ .027778 3/36 x 6/9 ≈ .055556

Since winning the pass line bet pays even money, the expectation on a 1-unit pass line wager is:
EV = (+1)(.222222) + (+1)(.027778) + (+1)(.044444) + (+1)(.063131)
+ (+1)(.063131) + (+1)(.044444) + (+1)(.027778) + (−1)(.111111)
+ (−1)(.055556) + (−1)(.066667) + (−1)(.075758) + (−1)(.075758)
+ (−1)(.066667) + (−1)(.055556)
= (+1)(.492929) + (−1)(.507071) = −0.01414.
This represents a 1.414%always wins. The probability of getting a red or black number in 100
spins every time and your expected value to profit off every spin is in the house advantage on the
pass line bet. The calculation above also shows the overall probability of winning the pass line
bet is .492929 and the overall probability of losing the pass line bet is .507071.

After a point is established, a bet on the pass line cannot be removed or reduced, although it may
be increased. This latter move is unfavorable to the player since the advantage on the pass line
bet is on the come-out roll and the odds are against the pass line bet once a point is established.
Also unfavorable, a player who misses the come-out roll may make a bet on the pass line after a
point is established (called a put bet).

The odds bet is called free odds because ... well, because it’s free – the house advantage is zero.
Odds can be taken after a point has been established on the pass line wager (odds can also be
taken on come, don’t pass, and don’t come after a point is established – these bets are discussed
on the next page). Like the pass line bet, if the point is rolled before a 7, the odds bet wins. If a 7
is rolled before the point, the odds bet loses. Unlike the pass line, which pays even money, a
winning odds bet is paid according to the true odds. For example, suppose a $5 pass line bet is
made and a 4 is thrown on the come-out roll. Odds may now be taken on the point 4. The amount
of odds that can be taken varies depending on the casino, but for this example let’s assume single
odds are taken. This is an additional $5 bet that will win if a 4 is rolled before a 7 and lose if a 7
is rolled before a 4. Since there are six ways to roll a 7 and three ways to roll a 4, the true odds
against winning this bet are 2 to 1. If the point passes (a 4 is rolled before a 7), the pass line will
be paid $5 (even money) and the odds bet will be paid $10 (2 to 1). Because odds bets are paid
according to the true odds, the expectation and house advantage are zero. For example, the
expected value calculation for a $10 odds bet on the point 4 is:

EV($10) = (+$20)(1/3) + (−$10)(2/3) = $0.00.

Because the house advantage is zero on the odds bet, taking odds when a point is established will
lower the overall house advantage on the total wager, pass line and odds bet combined, from the
1.414% edge on the pass line alone. With single odds, the house advantage on the combined total
wager is 0.848%, with double odds 0.606%, with triple odds 0.471%, with 5X odds 0.326%, and
with 10X odds 0.184%.
Many casinos are now allowing a 3/4/5X odds structure, where the player may take 3X odds on
the 4 and 10, 4X odds on the 5 and 9, and 5X odds on the 6 and 8. This 3/4/5X odds structure
leads to a 0.374% house advantage on the total wager (pass line or come plus odds).

The odds bet may be removed at any time. Come odds (see come bet below) are typically “off,”
or inactive, on a come-out roll, unless the bettor specifically asks that they be “on.” Don’t come
odds are always on.
The come bet gives the player the opportunity to make what is essentially a pass line bet after a
point is established. A come bet turns the next roll of the dice into a come-out roll that wins on a
7 or 11, loses on a 2, 3, or 12, and otherwise establishes a point that will win if the point number
is rolled again before a 7 and will lose if a 7 is rolled before the point. Since the come bet works
like the pass line bet, the probabilities and resulting edge are the same as that for the pass line
(1.414% with no odds, 0.848% with single odds, 0.606% with double odds, 0.471% with triple
odds, 0.326% with 5X odds, 0.184% with 10X odds, and 0.374% for 3/4/5X odds).

The don’t pass wager works the opposite of the pass line bet, except that don’t pass bettors are
barred from winning on a come-out roll of 12. The don’t pass line bet wins on a come-out roll of
2 or 3, loses on a 7 or 11, and ties on a 12. On point rolls, the don’t pass wins if a 7 is rolled
before the point and loses if the point is rolled before a 7. An analysis similar to the one for the
pass line wager shows the house advantage on the don’t pass bet to be 1.364%.58.

Like the pass line bet, odds can be taken once a point is established. Odds on the don’t pass line
bet will win if a 7 rolls before the point and are paid at true odds – 1 to 2 for the 4 or 10, 2 to 3
for the 5 or 9, and 5 to 6 for the 6 or 8. Taking odds will reduce the house advantage on the
combined wager, don’t pass line plus odds, to 0.682% with single odds, 0.455% with double
odds, 0.341% with triple odds, 0.227% with 5X odds, 0.124% with 10X odds, and 0.273% for
3/4/5X odds. Don’t pass bets can be removed or reduced after a point is established.

The don’t come bet is essentially a don’t pass bet made after a point has been established. It
works just like the don’t pass and has the same house advantage (1.364% with no odds, 0.682%
with single odds, 0.455% with double odds, 0.341% with triple odds, 0.227% with 5X odds,
0.124% with 10X odds, and 0.273% with 3/4/5X odds).

The pass line, don’t pass line, come, and don’t come wagers, along with the odds bet, represent
the main wagers in craps. This is house advantage per roll. Since rolling 12 results in a push, the
house advantage may also be expressed as 1.403% per resolved bet. There are other games,
notably baccarat, in which the house advantage for certain wagers may be expressed either
including or excluding ties. These are house advantages per roll. The corresponding house
advantages per resolved bet are 0.691%, 0.459%, 0.343%, 0.228%, 0.124%, and 0.274% for 1X,
2X, 3X, 5X, 10X, and 3/4/5X odds.

A place bet on one of the point numbers, 4, 5, 6, 8, 9, or 10, is a bet that the number will roll
before a 7. Payoffs are less than true odds, typically 9 to 5 for the numbers 4 and 10, 7 to 5 for
the numbers 5 and 9, and 7 to 6 for the numbers 6 and 8. The house advantage is 6.67% for the 4
and 10, 4.00% for the 5 and 9, and 1.52% for the 6 and 8. A don’t place bet is the opposite of a
place bet – it’s a bet that a 7 will roll before the number. Like place bets, these wagers are paid
off at less than true odds. The don’t place on a 4 or 10 pays 5 to 11 (true odds are 1 to 2), a don’t
place on a 5 or 9 pays 5 to 8 (true odds 2 to 3), and a don’t place on a 6 or 8 pays 4 to 5 (true
odds 5 to 6). House advantages are 3.03% for the 4 and 10, 2.50% for the 5 and 9, and 1.82% for
the 6 and 8. Expectation calculations and resulting advantages are shown below.

Place 4, Place 10
EV($5) = (+9)(3/9) + (−5)(6/9) = −0.3333,
HA = 0.3333/5 = 0.0667, or 6.67%.

Place 5, Place 9
EV($5) = (+7)(4/10) + (−5)(6/10) = −0.2000,
HA = 0.2000/5 = 0.0400, or 4.00%

Place 6, Place 8
EV($6) = (+7)(5/11) + (−6)(6/11) = −0.0909,
HA = 0.0909/6 = 0.0152, or 1.52%.

Don’t Place 4, Don’t Place 10

EV($11) = (+5)(6/9) + (−11)(3/9) = −0.3333,
HA = 0.3333/11 = 0.0303, or 3.03%.

Don’t Place 5, Don’t Place 9

EV($8) = (+5)(6/10) + (−8)(4/10) = −0.2000,
HA = 0.2000/8 = 0.0250, or 2.50%.

Don’t Place 6, Don’t Place 8

Don’t place bets are not very common these days and not available in many casinos.
EV($5) = (+4)(6/11) + (−5)(5/11) = −0.0909,
HA = 0.0909/5 = 0.0182, or 1.82%.

Buy and lay bets are similar to the place and don’t place bets except that the payoffs are at true
odds and a 5% commission is charged on the amount of a buy bet and on the possible winning
amount of a lay bet. All buy bets have a 4.76% house advantage. Lay bets carry house
advantages of 2.44% for the 4 and 10, 3.23% for the 5 and 9, and 4.00% for the 6 and 8.
Expectation and resulting house advantage
calculations are shown below.

Buy 4, Buy 10
EV($21) = (+$39)(3/9) + (−$21)(6/9) = −$1.00,
HA = 1.00/21 = 0.0476, or 4.76%.
Buy 5, Buy 9
EV($21) = (+$29)(4/10) + (−$21)(6/10) = −$1.00,
HA = 1.00/21 = 0.0476, or 4.76%.

Buy 6, Buy 8
EV($21) = (+$23)(5/11) + (−$21)(6/11) = −$1.00,
HA = 1.00/21 = 0.0476, or 4.76%.
Lay 4, Lay 10
EV($41) = (+$19)(6/9) + (−$41)(3/9) = −$1.00,
HA = 1.00/41 = 0.0244, or 2.44%.

Lay 5, Lay 9
EV($31) = (+$19)(6/10) + (−$31)(4/10) = −$1.00,
HA = 1.00/31 = 0.0323, or 3.23%.

Lay 6, Lay 8
EV($25) = (+$19)(6/11) + (−$25)(5/11) = −$1.00,
HA = 1.00/25 = 0.0400, or 4.00%.

The field bet is a bet that the next roll will be a 2, 3, 4, 9, 10, 11, or 12. The payoff is typically 2
to 1 for 2 and 12, and even money for 3, 4, 9, 10, and 11. Although this wager covers seven out
of eleven numbers, the losing numbers (5, 6, 7, and 8) can be rolled more in more ways. The
house advantage is 5.56%. This can be seen with the usual expected value calculation for a $1
wager:

EV = (+$2)(2/36) + (+$1)(14/36) + (−$1)(20/36) = −$0.0556.

Some casinos offer a 3 to 1 payoff on either the 2 or 12 (but not both). A calculation similar to
the one above will show that paying 3 to 1 for either the 2 or 12 lowers the house advantage to
2.78%. (Paying 3 to 1 on both the 2 and 12 would result in a house advantage of zero.)

The big 6 is a bet that the 6 will roll before a seven; the big 8 is that the 8 will roll before a 7.
The payoff is even money and the house advantage is 9.09%, as can be seen from the following
expectation calculation:

EV = (+1)(5/11) + (−1)(6/11) = −0.0909.

The big 6 and big 8, the place 6 and place 8, and the buy 6 and buy 8 are exactly the same
wagers but with different payoffs, and hence different house advantages. The place 6 and place 8
pay 7 to 6 and have a house advantage of 1.52%; the buy 6 and buy 8 pay the true odds of 6 to 5
but charge a 5% commission, resulting in a house advantage of 4.76%; the big 6 and big 8 pay
even money and have a house advantage of 9.09%.

There are four hardway wagers: hard 4, hard 6, hard 8, and hard 10. A hardway is a wager that
the selected number will roll with doubles before the number is rolled any other way or a 7 is
rolled. For example, a hard 6 wins if 3-3 comes up before an easy 6 (1-5, 5-1, 2-4, or 4-2) or a 7.
Payoffs are 7 to 1 for hard 4 and hard 10, and 9 to 1 for hard 6 and hard 8. Expected value
calculations are straightforward. For the hard 4 or hard 10, there are eight losing combinations
(two easy ways to roll the selected number – 1-3, 3-1 or 4-6, 6-4 – and six ways to roll a 7) and
only one winning combination (either 2-2 or 5-5).

Thus the expectation is:

EV = (+7)(1/9) + (−1)(8/9) = −0.1111.

For the hard 6 or hard 8, there are ten losing combinations (four easy ways to roll the selected
number – 1-5, 5-1, 2-4, 4-2 or 2-6, 6-2, 3-5, 5-3 – and six ways to roll a 7) and only one winning
combination (either 3-3 or 4-4). The expectation is: EV = (+9)(1/11) + (−1)(10/11) = −0.0909.
House advantages are 11.11% for the hard 4 and hard 10, and 9.09% for the hard 6 and hard 8.

ANY CRAPS
Any craps is a one-roll bet that the next roll will be a 2, 3, or 12. If the next roll is not a 2, 3, or
12, the wager is lost. The payoff is 7 to 161 and the expected value is:

EV = (+7)(4/36) + (−1)(32/36) = −0.1111,

HA = 11.11%.

ANY SEVEN
Another one-roll bet, any seven wins 4 to 1 if the next roll is a 7. The expected value is:

EV = (+4)(6/36) + (−1)(30/36) = −0.1667,

HA = 16.67%.

CRAPS AND ELEVEN (C&E)

Craps and eleven (C&E) is a one-roll bet that the next roll will be a 2, 3, 11, or 12. Using payoffs
of 3 to 1 on the 2, 3, and 12, and 7 to 1 on the 11, the expected value is:

EV = (+3)(4/36) + (+7)(2/36) + (−1)(30/36) = −0.1111,

HA = 11.11%.

TWO OR TWELVE
These are two separate one-roll bets – a wager can be made that the next roll will be a two or a
wager can be made that the next roll will be a twelve. The payoff is typically 30 to 1 for each.

The expectation using this payoff is given by:

EV = (+30)(1/36) + (−1)(35/36) = −0.1389,

HA = 13.89% (for either wager).

THREE OR ELEVEN
Like the wagers on two or twelve, these are two separate one-roll bets. Each typically pays 15 to
1.63

EV = (+15)(2/36) + (−1)(34/36) = −0.1111,

HA = 11.11% (for either wager).

HORN BET
The horn is a one-roll bet that the next roll will be 2, 3, 11 or 12. It is really just four separate
bets on these four numbers, so the bet is made in multiples of four. The horn returns the usual
payoffs for the individual winning number: typically 30 to 1 for the 2 and 12, and 15 to 1 for the
3 and 11. With these payoffs and four units wagered on the horn, the bet will net 27 units for a
hit on 2 or 12, and 12 units for a hit on 3 or 11. The expected value is:

EV = (+27)(2/36) + (+12)(4/36) + (−4)(30/36) = −0.50.

Since the bet was 4 units, the house advantage is 0.50/4 = 0.125, or 12.5%.64

HORN HIGH BET

The high horn is a horn bet made in multiples of five with the extra unit wagered on the “high”
number designated by the bettor. For example, a $5 high twelve would put $1 each on the 2, 3,
and 11, and $2 on the 12. With 30 to 1 payoffs for the 2 and 12, and 15 to 1 for the 3 and 11, the
house advantage depends on which number is chosen to be “high.” If 2 or 12 are designated
high, the expected value is given by EV = (+57)(1/36) + (+26)(1/36) + (+11)(4/36) + (−5)(30/36)
= −0.6389.

If 3 or 11 are designated high, the expectation is:

EV = (+26)(2/36) + (+27)(2/36) + (+11)(2/36) + (−5)(30/36) = −0.6111.

These expectations yield house advantages of 12.78% (0.6389/5 = 0.1278) for the high two and
high twelve, and 12.22% (0.6111/5 = 0.1222) for the high three and high eleven.

HOP BET
The hop is a one-roll bet that the next roll will be a certain combination. There are two types, an
“easy” or two-way hop, such as 6-1, and a “hard” or one-way hop, such as 2-2.66 Typical
payoffs are 15 to 1 for an easy hop and 30 to 1 for a hard hop.

Thefavor. The player is expected value calculation for the easy hop is:
EV = (+15)(2/36) + (−1)(34/36) = −0.1111.

For the hard hop:

EV = (+30)(1/36) + (−1)(35/36) = −0.1389.

House advantages are 11.11% for the easy hop and 13.89% for the hard hop.

Baccarat’s overall house advantage is about 1.2% – one of the lowest in the casino. Play
proceeds as follows. After bets are made, two cards are dealt to each side (player and banker).
All cards are community cards – there are no individual hands. Face cards and Tens are worth
zero, Aces count as one, and all other cards are worth face value. The total for a hand is the sum
of the values of the individual cards in the hand, modulo 10 (meaning that the first digit is
dropped if the total is greater than nine).

A total of 8 or 9 on the first two cards is called a natural. If either side has a natural, the hand is
over and wagers are resolved. If neither side has a natural, a fixed set of rules (following) is
applied to determine if either side takes a third card. Once the hands are complete, the side with
the higher total wins. If the two sides have equal totals, player and banker bets are a push. The
payoffs on winning player and banker bets are even money, but a 5% commission is charged to
banker wins.68 A winning tie bet pays 8 to 1.

Baccarat’s hitting and standing rules are presented in the following table. To read the table, start
with the player’s first two cards total in the left column and proceed to the right, column by
column to determine the drawing/standing rules. The rules favor the banker and the casino’s
advantage on the banker bet would be negative without the 5% commission. With this
commission, the house advantage on the banker bet is 1.06% (including ties). House Advantage:
Player Bet: 1.24% (1.37% ignoring ties)
Banker Bet: 1.06% (1.17% ignoring ties) Tie Bet: 14.36% Typical Hold: 18% (Nevada)

*If either player or banker totals 8 or 9 on first two cards, no cards are drawn and hand is over.
As mentioned above, since the drawing rules are fixed, baccarat offers no opportunity for
strategic decisions like blackjack does. Furthermore, although the game involves dependent trials
(similar to blackjack), it has been shown that card-counting techniques are not effective in
baccarat.
Since baccarat is typically played with eight decks of cards, a proper mathematical analysis of
baccarat can be performed by examining all possible combinations of player and banker hands
for all possible six-card sequences in an eight-deck shoe. The analysis can be limited to six-card
subsets since no hand can consist of more than six total cards for player and banker combined,
and since hands comprised of less than six cards can be accounted for by summing the
appropriate probabilities.

However, since suits and the distinction between Tens, Jacks, Queens, and Kings are irrelevant,
only 5,005 distinguishable six-card subsets are possible, significantly easing the computational
burden. A complete analysis, then, consists of (1) identifying the 5,005 distinguishable six-card
subsets, (2) playing through the 180 possible hands for each of these subsets, (3) recording the
player and banker scores and winning.

The outcome (player, banker, or tie) for each hand, and (4) weighting each of the (5005 x 180)
outcomes by the appropriate number of repetitions (corresponding to non-distinguishable six-
card subsets).

The analysis described above shows that the player wins 44.625% of the hands, the banker wins
45.860% of the hands, and 9.516% end up tied. A complete breakdown by banker total is given
in the following table.

As is the case with the don’t pass and don’t come wagers in craps, the player and banker
advantages can be expressed either including or excluding ties. The 1.24% and 1.06%
advantages reported above include ties. If ties are ignored, the player will win 49.3175% of the
hands and the banker will win 50.6825%. The expected values per resolved bet for these two bets
are:

EVPlayer = (+1)(.493175) + (−1)(.506825) = −0.01364965

EVBanker = (+0.95)(.506825) + (−1)(.493175) = −0.0116916.

House advantages ignoring ties, then, are 1.36% and 1.17% for player and banker,
respectively.71 These advantages per resolved bet also could be obtained by dividing the house
advantages per hand by the proportion of untied hands:

.1 235% .0/ 9048 = 36.1 % and % .1 058% .0/ 9048 = 17.1 .

Regardless of whether house advantage is expressed per hand or per resolved bet, expected
casino earnings are the same. For example, if a player bets $25 a hand on the banker for 10,000
hands, the 1.06% house advantage per hand dealt results in a casino expected win of about
$2,645 (.010579 x 10,000 hands dealt x $25 per hand). Applying the 1.17% house advantage
per resolved bet to the untied hands yields – aside from rounding error – the same expected win
(.011692 x 9,048 resolved bets x $25 per bet = $2,645). Obviously a similar equivalency exists
for the player bet.

The vitality for baccarat since it is a game favored by high rollers and represents one of the few
situations where the casino can be put in a high-risk situation. This section elaborates further on
fluctuations in casino win in baccarat.

For the sake of simplicity, the analysis that follows assumes an overall game standard deviation
of 0.94 per unit and a casino advantage of 1.15%, approximately the values for a large series of
bets equally divided between player and banker. Actual total casino win over the course of many
hands will deviate from the expected win by an amount that depends on the per unit standard
deviation, the amount wagered per hand, and the number of hands. Mathematically speaking, in
repeated series of n equal wagers, the distribution of outcomes of total dollars won will be
normally distributed (bell-shaped) around the expected win with standard deviation equal to win
per unit SD = unit wager × n × SD . Confidence limits for the actual win can be obtained from
the following: ( ) win Expected Win ± z × SD , where z is the appropriate standard normal value
depending on the confidence level. For the usual 95% confidence, this is z = 1.96. Values of z
corresponding to other confidence levels can found in a table of standard normal probabilities. In
a series of 1,000 hands at $500 per hand, for example, the expected casino win is $5,750
(assuming a 1.15% house advantage), with an associated standard deviation of 94.0 ×$500× ,1
000 ≈ $14,863. The 95% confidence limits for the casino’s win are given by $5,750 ±
1.96(14,863), or −$23,381 to +$34,881.to lose

Casinos can increase the theoretical paybacks on slots by either increasing the number of
winning combinations or increasing the amounts of paybacks on winning combinations. A casino
can do the former on many slot machines by modifying the computer chip that operates the game
so as to produce more or higher paying winning combinations. A casino can do the latter by
increasing the paybacks on combinations. For example, the payout on lining up four bars may be
increased from 100 coins to
200 coins. In either case, over time, the casino decreases the theoretical win, pays out more in
winnings and retains less revenue as a percentage of handle.

Slot machines with progressive meters do not have a set theoretical win in the same sense as
other slot machines. On non-progressive slot machines, the theoretical payout to players is the
same on every play because the opportunity to win, the winning combinations, and the amounts
of the payouts on winning combinations remain constant. On progressive machines, the
progressive meter increases with each pull of the handle. For example, on a dollar progressive
machine, a nickel of every dollar played may be added to the progressive jackpot. Therefore, on
any pull of the handle, the theoretical pay back will increase as the progressive meter increases.
If no player wins the progressive jackpot, at some point, the increasing size of the jackpot creates
a theoretical payback that will result in a statistical advantage to the players. From this point to
the time a player wins the jackpot, the total amount of the paybacks to the players including the
progressive jackpot should exceed the total amount of money played by the players.

This slight statistical advantage is what attracts the slot teams. They find progressive jackpots
that have arisen to a level that favors the players. Only certain progressive slot machines will
meet the slot team’s criteria. First, the number of slot machines that are linked to the progressive
jackpot must be manageable. The team needs to monopolize all the slot machines to avoid the
risk that a non-member will win the jackpot. Second, the statistical frequency of hitting the
jackpot must be consistent with the slot team’s bankroll. No slot team has an unlimited bankroll.
Based on probabilities, the team needs to have enough cash on hand to play all the machines on
the carousel until one hits the progressive jackpot.77 For example, the previous chart shows a
progressive slot machine that has a cycle of 884,736 (and a probability of a little better than one
in one million of hitting the progressive jackpot). If the 77 The Poisson probability distribution
can be used to determine the probability of hitting the jackpot over the course of a specified
number of plays. See the section on Slot Machines earlier in this chapter.

The probability of hitting the progressive jackpot is one in ten million, the slot team would need
an astronomical bankroll to have reasonable assurances that it would hit the jackpot before
running out of money. Therefore, slot teams avoid the “mega” jackpot progressive that have very
infrequent hits or winners. To minimize their potential exposure, they concentrate on progressive
carousels that feature lower jackpots with a higher frequency of payouts.

It is an advantage to the player must act first in blackjack, because the player will lose anytime
there is a “double bust” in which both player and dealer totals exceed twenty-one. To get a rough
idea of how much this is worth, consider a player who plays his hand exactly like the dealer must
play. This player and the dealer will each bust about 28% of the time and so there will be a
double bust about 8% of the time (.28 x .28 ≈ .08). All else being equal, this gives the casino
roughly an 8% advantage. All else, however, is not equal. The 3 to 2 payoff for a player natural
is worth about 2.3%. Thus, a player who mimics the dealer is giving the casino about a 5.7%
edge. The player can further reduce the house edge by taking advantage of the strategy options
not available to the dealer. The approximate potential benefits of each of these strategy options
are:

♠ Proper hitting and standing, worth ≈ 3.5%.

♠ Doubling down when advantageous, worth ≈ 1.7%.
♠ Splitting when advantageous, worth ≈ 0.5%.

Additionally, surrender, if allowed, is worth a fraction of a percent (late surrender is worth ≈

0.06%, early surrender is worth ≈ 0.62% but is rarely offered) and insurance can be a positive
expectation move for a player who is tracking cards. Proper use of these options can, for most
games offered, whittle the casino advantage down to anywhere between 0.0% (single deck, Las
Vegas Strip rules) and 1.0% (six decks, dealer hits soft 17, double down on ten and eleven only),
depending on the number of decks in play and the particular rules in use. A casino advantage of
0.5% is a reasonable representative figure that can be applied to expert basic strategy for most
games. For the average player, the casino advantage is about 2%.

Basic strategy is the player’s strategy that maximizes his expectation, or average gain, playing
one hand against a freshly shuffled pack, without keeping track of the cards. It is a complete set
of decision rules that tells the player whether to hit, stand, double down, split, or surrender (if
offered) depending only on the player’s present cards and the dealer’s upcard. It does not depend
on other players’ cards or previously played cards. Basic strategy was developed by analyzing all
possible combinations of decisions and outcomes for a given set of player’s cards and dealer’s
upcard. The decision resulting in the largest expected win (or smallest average loss) is the basic
reports 5.5%. In the long run, a player who follows basic strategy perfectly will maximize his
average gain. Basic strategy charts86 for single and multiple deck games under typical rules
(surrender in these charts refers to late surrender) are given below.

Other things being equal, a blackjack game with multiple decks will result in a larger house
advantage than a single deck game. The reasons for this are:

♣ Naturals are less prevalent in multiple deck games (dropping from 0.0483, or
once every 20.7 hands, for a single deck, to 0.0475, or once every 21.1 hands
for six decks), reducing the player advantage on 3 to 2 payoffs for naturals.87
♣ Doubling down is less effective.
♣ Judicious standing with stiff totals is less successful.

These outcomes result because the impact of removing individual cards is less for a multiple
deck than for a single deck. To see this, supposethan he wins. If the player doubles down with a
(6,4)wanted to play in order to win $1 profit, the chances of having a good betting strategy will
not win against the dealer’s upcard of 5. The player is hoping for a Ten or Ace. With the three
cards already in play, the probability of drawing a Ten or Ace in a single deck game is 20/49 = .
408, since there are still 16 Tens and 4 Aces in the remaining 49 cards. For an eight-deck game,
however, this same probability is only 160/413 = 0.387. In terms of the chances of receiving the
desired Ten or Ace hit card on the double down, the effect of removing the three cards is more
dramatic for a single deck than for the eight decks. For similar reasons, the player is less
successful when standing with stiff totals. In this situation the player is hoping the dealer will
bust, but with eight decks, there are more small cards available (percentage-wise) for the dealer
to avoid busting than with a single deckhouse.
The previous section explained why the house advantage increases with more decks. This section
addresses the effect of different rule variations. In a single deck game with old Las Vegas Strip
rules (dealer stands on soft 17, double down on any first two cards but not after splits, split non-
Ace pairs up to four times, split Aces receive only one card, no surrender), basic strategy results
in essentially a zero house advantage.

For each rule variation in effect in a game, add the corresponding percentage from the table to
the 0.00% benchmark value to arrive at the approximate overall house advantage for the game.
For example, a 6-deck game where dealer stands on soft 17 and double down is allowed on 10
and 11 only, the house advantage is 0.54% + 0.26% = 0.80%. This same game with dealer hitting
soft 17 would have a 1.00% house advantage.

The insurance bet in blackjack is offered when the dealer’s upcard is an Ace. This side bet is
made for up to one-half the value of the original wager and pays 2 to 1 if the dealer has a natural.
This wager amounts to a bet on whether the dealer’s hole card is a Ten (value) and is completely
independent of the original wager. The casino The advantage on the insurance bet depends on the
number of decks in play. The reasoning is illustrated for the case of a single-deck game. If the
dealer shows an Ace and the player is holding two 10s, then there are 14 Tens and 35 non-Tens
among the 49 remaining unseen cards. In this case the probability the dealer has a natural is
14/49, the probability the dealer does not have a natural is 35/49, and the expected value of the
insurance wager is:

EV = (+2)(14/49) + (−1)(35/49) = −0.14286.

Similarly, the expectations for the cases when the player is holding a Ten and non-Ten, and two
non-Tens are:

EV = (+2)(15/49) + (−1)(34/49) = −0.08163,

and EV = (+2)(16/49) + (−1)(33/49) = −0.02041.

Thus, the house advantage on the insurance wager is 14.286% when the player is holding two
Tens, 8.163% when the player is holding a Ten and a non-Ten, and 2.041% when the player is
holding two non-Tens. To find the overall house advantage, these three values must be weighted
by the probabilities the player will be in the three situations. These probabilities are (for a single
deck game):

Player holds two Tens: 094118 16( 51/ ) × 15( 50/ ) = .0 .

Player holds a Ten and a non-Ten: 439216 16( 51/ ) × 35( 50/ ) × 2 = .0 .
Player holds two non-Tens: 466667 35( 51/ ) × 34( 50/ ) = .0 .
The overall house advantage for the insurance wager, then, is:
HA = .094118(14.286%) + .439216(8.163%) + .4666667(2.041%) = 5.88%.
Insuring a natural (for the full half-unit insurance wager) is sometimes referred to as taking
“even money.” The reasoning for this is as follows. If the dealer does not have a natural, the
player loses the half-unit insurance bet but wins 11⁄2 to 1 on the original wager for a net profit of
1 unit. If the dealer does have the natural, the player wins 1 unit (2 to 1 on the half-unit wagered)
for the insurance wager and pushes on the original bet, for a 1-unit profit. Thus, insuring a
natural results in the player being paid even money regardless of whether or not the dealer also
has a natural. Taking even money may result in a 100% return to the player, but insurance is still
a negative expectation play even when holding a natural. The figures above show the
disadvantage on the insurance bet (alone) to be about 8% in this situation. When the original bet
and the insurance bet are considered together, the overall expected value of not taking insurance
when holding a natural is:

EV = (0)(15/49) + (+1.5)(34/49) = +1.0408.

That is, not taking insurance will return 104% to the player in the long run compared to 100% by
taking even money. These computations on insurance bets assume the player has no knowledge
of cards other than his own first two cards and the dealer’s upcard. For card counters, however,
the insurance bet could be favorable since they’ll know when there are more Tens than usual
remaining to be played.

The average squared outcome yields the variance (since the expected value is effectively zero)
for a single hand of blackjack: Variance ≈ (±2)2

(.10) + (1.5)2
(.05) + (±1)2
(.75) + (0)2
(.10) ≈ 1.26.

The standard deviation is the square root of 1.26, or about 1.1 (i.e., 1.1 units or 110% of the
original wager). Using this per unit standard deviation, confidence limits can be constructed for
the actual win in a large series of independent blackjack wagers.

For 1,600 hands, for example, the standard deviation is 0.0275, or 2.75%. This is equal to 44
units (2.75% of 1,600). Assume now a player betting $10 per hand and a 1.0% house advantage.
Then the expected casino win is 16 units (1% of 1,600) or $160, with an associated standard
deviation of $440. Adding and subtracting one standard deviation from the expected win, about
68% of the time the house win from this player (after 1,600 hands) should be between −$280 and
+$600. Adding and subtracting two standard deviations from the expected win indicates that
about 95% of time the casino win will be between −$720 and $1,040. This type of analysis was
used to construct the following table showing approximate 95% confidence limits for casino
units won for various numbers of wagers assuming expert basic strategy against a rather typical
0.5% house edge game.

The limits in the preceding table give a rough guide as to the variation that can be expected. For
example, the casino may lose as much as $17,000 or win as much as $27,000 for a $100 per hand
basic strategy bettor over the course of 10,000 hands. A loss or win outside this range may be
due to something other than normal statistical fluctuations. Similarly, over the course of 1,000
hands, 95% of the time a $100 per hand player will find his winnings between −$7,500 (casino
win) and $6,500. For a given number of hands played, win values occurring between the lower
and upper confidence limits are to be expected. Win amounts outside these 95% confidence
should alert the casino to the possibility that something may be amiss. An unusually large win by
a player over the course of this many hands, for example, may be due to cheating or card-
counting.

Odds in blackjack vary significantly depending on the skill level of the player, which can range
from a slight advantage for a skilled card counter (see below) to a substantial disadvantage for a
new player with no knowledge of basic strategy. Most casinos should expect that over time
actual win percentage will exceed the theoretical house advantage because many, if not most,
blackjack players will not always play with the highest player expectation (i.e., according to
basic strategy ignoring card-counting). How much this benefits the casino will depend on many
factors, such as whether the casino’s clientele are predominately tourists versus locals, the
maturity of the market, and the size of the table limits. This is because local blackjack players
tend to have a better grasp on average of basic strategy than tourists. The same generalities can
be said of players in mature markets as opposed to new markets, and
those who play at higher limit games than lower limit games.

Because successive hands in blackjack are dependent trials, cards that have been played on
previous hands affect the probabilities of what will happen on later hands. This is not true for
games such as roulette and craps, where the spins of the wheel and the throws of the dice are
independent trials. A series of nine consecutive red numbers in roulette does not make red any
more or less likely on the tenth spin. Similarly, the probability of rolling a seven is the same, 1/6,
regardless of what was thrown on the previous roll (or series of rolls) of the dice. For cards being
dealt from the same deck or shoe as in blackjack, however, the statistical environment is quite
different. Once a card is removed, the composition of the remaining deck changes. If four Aces
have been dealt (without replacement) from a single deck, the probability that the next card will
be an Ace is zero. A knowledgeable player can beat blackjack using a memory-based card-
counting system because the probabilities of different outcomes change depending on what cards
have already been played.
Changes in player expectation in the above table reveal that, generally, removing low cards from
the deck will increase the player’s expectation while removing high cards will reduce the
player’s expectation. Thus, when a preponderance of low cards has been played and the
remaining deck is rich in high cards, subsequent hands will tend to favor the player. Conversely,
if a preponderance of high cards has been played, subsequent hands will tend to favor the dealer.
These observations form the basis for card-counting strategies. A player keeping track of the
cards that have been played can recognize when the remaining deck is rich in high cards, and can
then take advantage of the increase in player expectation at these times by increasing the bet size
and/or altering playing decisions. When the remaining deck contains proportionally more small
cards, the counter should lower the bet. With proper use of such a strategy, it is possible to play
blackjack with a positive overall expectation (i.e., the casino advantage will be negative).

Why does a deck rich in high cards favor the player? The reasons are:

♦ Naturals are more prevalent.

♦ The dealer is more likely to bust when hitting stiffs.
♦ Player double downs are more effective.
♦ Player splits are more effective.
♦ Insurance can be effective.

To elaborate on these reasons, although both dealer and player will get more naturals when the
deck is rich in Tens and Aces, the player is paid a 3 to 2 premium (the dealer is not). The dealer
must hit stiff totals (hard 12-16), and is more likely to bust if there is a preponderance of Tens
remaining, whereas the player can choose not to hit these stiff hands. Player double downs and
splits are more effective since the player is often looking for a high hit card in these situations
(and also because in many of these situations the dealer’s upcard is weak, leading to more dealer
busts). More Tens than usual also means that insurance can be a positive expectation wager for
the player. Standard expected value calculations show that insurance is advantageous if more
than one-third of the remaining cards are Tens.

The edge gained by a card-counter depends on several factors, including the number of decks in
play, the penetration (how far into the pack cards are dealt before reshuffling), the rules of the
game, the bet spread and the system used. A skilled counter will typically have a 0.5% to 1.5%
advantage over the house. More decks and shallow penetration (not dealing very far into the pack
before reshuffling) tend to decrease the counter’s advantage. Any advantage created through
card-counting, of course, is negated when the deck is shuffled. Possible casino countermeasures
against card-counting include preferential or at-will shuffling.

Poker does not pit the player against the casino. Instead, players compete against each other and
money won or lost merely is transferred from one player to another. The casino provides a
dealer, who does not play, and makes its money by (a) taking a percentage of each pot, (b)
charging an hourly fee, or (c) collecting a flat amount for every hand. The first of these is most
common, and the percentage extracted is called the rake. A rake of 5% to 10% is typical.

The object of poker is not necessarily to have the best hand, but to win the most money. Having
the best hand usually helps, but does not guarantee victory because of the importance of skill and
psychology. Indeed, it is fair to say that skill is more of a factor in poker than in any other casino
game. For skilled players, poker can be a positive expectation game and it attracts many
professional and semi-professional players. There are not many people who earn a living or
consistently make money gambling, but the majority of those who do are poker players.

With most casino games – roulette, craps, baccarat and keno are examples – it is not
mathematically possible to win in the long run. Blackjack and video poker offer the opportunity
for a slight player edge (one to two percent) and a few players can make money at these games
with perfect play (and counting cards in the case of blackjack). The going is tough, however, and
the earnings are small. Skill does play a part in these games, but luck is a large factor. In
cardroom poker, skill is a much greater factor and the better players will win out over the weaker
players in the long run.

Casino poker games are played table stakes, which means a player may bet only with the chips
(or money) he has on the table during a hand. If a player runs out of chips when calling or
betting, he cannot add more until the hand is over, and must go all-in to stay in the hand. When a
player goes all-in, all subsequent wagers by other players go into a separate side pot in which the
all-in player has no interest – he may win the main pot, to which he contributed, but may not win
the side pot even if his hand is the best.

The limits, or absence of limits, on how much a player may bet and raise will dramatically affect
the game dynamics, including players’ decisions and strategies, and the relative balance of luck
versus skill in the game.

In a fixed-limit game, no bet or raise may exceed a specified amount. This amount usually varies
with the betting round, with later rounds allowing higher bets and raises than early rounds. In a
$5−$10 fixed-limit game, for example, players may bet or raise exactly $5 in early rounds and
exactly $10 in later rounds.

Spread limit games are similar to fixed-limit, but allow any bet between the two amounts at any
time. Thus in a $10−$20 spread-limit game, bettors may make wagers of any amount between
$10 and $20 at any time, with the provision that raises must be at least equal to the preceding bet.
In pot-limit games, bets or raises are limited only by the amount of money in the pot at the time
the wager is made. In no-limit games, a player may bet or raise any amount he has in front of
him (table stakes limit betting in a hand to the chips and money on the table). Pot limit and no
limit formats are generally used only for more serious games (no-limit is used in the World
Series of Poker, the premier high-stakes tournament). In most limit games, a bet and either three
or four raises per betting round are permitted.

The probabilities of various poker hands (given in the table above) are determined
using the usual probability rules and methods of counting, including combinations and
permutations. The calculations for the probabilities of the different types of five-card
poker hands are outlined below.

Royal Flush = 1 in 640,000

Straight Flush= 1 in 72,000
Four of a Kind = 1 in 4,000
Full House = 1 in 700
Flush = 1 in 500
Straight = 1 in 250
Three of a Kind = 1 in 47
Two Pair = 1 in 20
One Pair = 1 in 4
Bad Hand = 1 in 2

In Hold ‘Em, each player is dealt two cards face down, and then a total of five community cards
are dealt face up in the center of the table. Each player uses the five community cards in
combination with his two hole cards to form the best five-card hand. After the first two cards are
dealt to each player there
is a round of betting. After the first round of betting, the first three community cards (the flop)
are exposed, followed by a second round of betting. After the second betting round, the fourth
community card (the turn) is exposed, followed by another round of betting, and then the fifth
and final community card (the river) is exposed, followed by a final round of betting. Each
betting round begins with the first active player to the left of the dealer (or in a game dealt by a
house dealer, the first active player to the left of the button used to indicate dealer position).
Suppose a player has four cards to flush after the flop. Consider the following three questions:

(a) What is the probability he will make the flush on the turn?
(b) If he doesn’t make the flush with the turn card, what is the probability he will make it on the
river?
(c) What is the probability he will make the flush on either the turn or river?

Suppose the player holds two spades and two of the three flop cards are spades. Since the player
has seen five cards – his two hole cards and the three flop cards – there are 47 remaining unseen
cards, of which nine are spades (i.e., there are nine outs, or cards that will complete the flush).
Thus the probability he will make his flush on the turn card is 9/47 = .191, for odds against of 38
to 9, or about 4.2 to 1, answering question (a). To answer (b), note that if he doesn’t make the
flush on the turn, there are still 9 spades left in the 46 remaining cards, so the probability he
makes it on the river is 9/46 = .196, for odds against of 37 to 9, or 4.1 to 1. To answer (c), first
compute the probability he doesn’t make the flush on the turn or river, in Hold ‘Em game, two
blind bets are posted before the cards are dealt – a small blind by the player to the dealer’s
immediate left and a large blind by the next player to the left of the small blind. A blind is a
forced bet made before the player sees his cards used to start the pot and stimulate action. The
small blind is usually equal to one-half the amount of the big blind. Since the deal rotates around
the table (even in a casino where the dealer is not a player, a button used to signify the nominal
dealer rotates after each hand), all players participate equally in the posting of any forced blind
bets. We will refrain from discussing further details regarding betting amounts and structure as it
is not necessary for this example. Because the first two players to the left of the dealer (or
button) have already acted by putting in blind bets, the player one to the left of the big blind is
the first with any choices (to call, raise, or fold in the first round of betting) on the pre-flop
betting round.

The probability of making the flush on either the turn or river is almost 35%, for odds against of
1.86 to 1. This last calculation illustrates how some probabilities can be easier to determine by
first computing the probability of the opposite (complement), then subtracting the result from
one. Note that the probability that the flush is made with one card to come depends on whether
we look at making the flush on the turn card or the river card (having not made the flush on the
turn). In the example above, the former probability is .191; the latter is .196.126 The following
tables show probabilities and odds for making hands with a given number of outs (cards that
will make the desired hand), with the “one
card to come” probabilities and odds computed assuming that one card is the final river card.

Examples of games in which rule variations can affect the odds are blackjack and craps. In
blackjack, the dealer hitting a soft seventeen increases the house advantage 0.2% compared to a
comparable game where the dealer must stand on soft seventeen. The number of decks used, no
soft doubling, and no re-splitting of pairs are other examples of rule variations in blackjack that
affect the overall price of the game to the player.

The “free odds” bet in craps can vary the price of the game product by the amount of odds that
can be taken. A player who bets the pass line and takes single odds is at a 0.85% disadvantage,
but only a 0.61% with double odds, and 0.47% with triple odds. This means for every $100
wagered on the pass line with single odds ($50 pass line and $50 odds), on the average, the
player will pay a price of about $0.85 while $100 bet on the pass line with triple odds ($25 pass
line and $75 odds) will cost about $0.47. A casino allowing triple odds offers a better priced
craps game than one that permits only single odds. Some casinos have offered as high as 100X
odds – a player taking full 100X odds will face only a 0.02% house advantage on the combined
pass line (or come) and odds wagers.
Early slot machines were mechanical and the odds of winning depended on the number of reels,
number of stops on each reel, and payouts for the winning combinations. For example, if a slot
machine had 3 reels with 20 stops each, and each stop had a different symbol, there would be (20
x 20 x 20) = 8,000 possible combinations. If 30% of these combinations were winners with a
combined total payout of 7,500 coins, the hold percentage (house advantage) would be (8,000 –
7,500)/8,000 or 6.25%. The average price of playing this machine would be $6.25 for every $100
bet. Casino operators could adjust this price by changing the number of winning combinations,
changing the symbol configurations on the reels, or changing the payout for winning
combinations. With today’s microchip-controlled slots, the odds of winning can be adjusted
merely by altering the computer program that runs the gaming device.

A factor that is not in the casino’s control that affects game price is player skill in those games
involving both chance and skill. In a typical six-deck game, for example, the average blackjack
player gives about a 2% edge to the house but a basic strategy player is at a 0.5% disadvantage.
Game speed, or number of decisions per hour, affects the cost per hour to the player (and the
expected casino win per hour), although it does not alter the basic price per unit wagered (i.e.,
the house advantage). In roulette, for example, a $5-per-spin player betting at a double-zero table
making 40 spins per hour can expect to pay ($5 x 40 x .0526) = $10.52 per hour. At a table
completing 60 spins each hour, the same player would spend an average of $15.78 per hour. The
basic price in both cases is 5.26% of the amount wagered. Similarly, at a base price of 1.15%, a
$100-per-hand baccarat player will pay, on average, about $92 per hour if playing 80 hands per
hour, but it will cost this same player $161 per hour if dealt 140 hands per hour.

Because pricing decisions in the casino require mathematical calculation, either a failure to make
such calculations or errors in doing so can have negative consequences. One casino wanted to
increase their baccarat play and so lowered the commission on winning banker bets to 2%. It was
not long before the players made them pay for this mistake. As the following expected value
calculation shows, this 2% commission gives the player a 0.32% advantage:

EV = (+.98)(0.4462) + (−1)(0.4586) = 0.0032.

A player betting $100 per hand at 60 hands per hour can expect to win $750 per hour. A $500
bettor at this rate would take in $3,750 per hour. An Illinois riverboat reportedly lost $200,000 in
one day with a “2 to 1 Tuesdays” promotion in which blackjack naturals paid 2 to 1 (instead of
the usual 3 to 2). A similar promotion with similar results occurred at a Las Vegas Strip casino.
Without other compensating rule changes, paying naturals 2 to 1 can increase the player
expectation enough to give the player about a 2% advantage over the house.
In most businesses, if the operator lowers the price of his goods or services, the volume of sales
will increase, provided that his competitors also do not lower their prices. Whether the increased
sales volume will justify the reduced price, however, is often unknown. A casino manager may
consider lowering the house advantage and therefore the price to the player to increase volume of
play. In some cases, the motivation for this is to keep the tables full. Here the casino manager
may see idle tables and come to the conclusion that he is better off filling the empty tables at
worse odds because the casino already is paying for the pit bosses, dealers and others. But,
ultimately, the casino manager needs to consider the impacts of lowering the price, via a lower
house advantage, on operating profits. Unfortunately, this is not a simple equation – a twenty
percent decrease in price is not necessarily made up by a twenty percent increase in volume.

On casino craps odds, with one casino ultimately trumping the others that offered 5X, 10X and
then 100X odds by offering unlimited odds. These more extravagant promotions were eventually
terminated for the reason that actual earning potential per hour based on the same player with the
same average combined (pass line plus odds) bet per hand substantially decreases as the back
line odds increase and the front line bets decrease. This phenomenon is illustrated below.

2X ODDS 5X ODDS 10X ODDS: Bet Win Bet Win Bet Win

Pass Line (1.41%) $110 $1.56 $55 $0.78 $30 $0.42

Odds (0%) $220 $0.00 $275 $0.00 $300 $0.00

TOTAL $330 $1.56 $330 $0.78 $330 $0.42
Avg. Bet Size* $257 $238 $230

*Average bet size is 2/3 of pass line amount plus 1/3 of total amount. As can be seen from the
table, the expected casino win on the same $330 total wager decreases from $1.56 for double
odds, to $0.78 for 5X odds, to $0.42 for 10X odds. This is because with higher odds, the player
can put more of the $330 on the back line odds and less on the pass line. Since the back line odds
are “free” and only 146 Assuming the player takes full advantage of the odds offered.

Volatility And Casino Operations the pass line portion of the combined amount is subject to a
casino edge (1.41%), the overall casino advantage and expected win decrease.147 The figures in
this example are for one $330 wager – earning potential per hour differences are much more
substantial.

Another major promotional game is single deck blackjack. On the proposition that offering a few
single deck blackjack games may help attract potential blackjack players. This may be true, he
concluded, but the reduced house advantage coupled with the propensity to attract more
knowledgeable players would require 30% more traffic on these games to have a positive
economic effect. He also concluded that to justify the same gross gaming win, the casino would
need to have about five times the volume just to break even because the house advantage is
significantly reduced in single deck blackjack compared to six-deck blackjack. With average or
poor players, the differences are not as extreme.

Still, the example shows the need for the casino manager to understand certain principles. The
first is how simply altering rules impacts earning potential. The second is that increased volume
from lowering the house advantage requires more personnel to deal, supervise and control the
games. Likewise, he needs to understand the cost of complimentaries, particularly beverages,
may increase with the additional volume. If a casino is not operating to full capacity, it may be
better to reduce the work force than to lower house odds.

The value of a particular player to the casino is simply how much the casino can expect to win
from that player. This is often called earning potential or theoretical win. This should not be
confused with actual win. Consider two players that visit a casino for the first time. Casino
management, not knowing either player, observed the first gambling for a few hours at roulette,
betting $200 per spin and losing $2,000. The second also was observed playing roulette for a few
hours, betting $400 per spin and winning $4,000. The two players approach the casino manager
and ask for the last available ticket to a casino-sponsored concert that night. Who should get the
ticket? The correct answer is the player that won the $4,000, but not necessarily because he
happens to have $4,000 of the casino’s money. All other things being equal, although he
happened to win, the casino has a better chance of winning more money
from him over time due to his $400 per spin wagers compared to the other player’s age of $200
per spin. While the house advantage over each player is the same over time, the total value of the
$400 average bettor is greater to the casino. As this example shows, what a casino actually wins
(or loses) will often differ greatly from its theoretical win projection due to statistical value.

Casino management needs to understand the factors that go into calculating player value beyond
the mere formula for determining theoretical win. Recognizing which factors go into earning
potential that are under the control of the casino and which are under the control of the player is
particularly important. With this knowledge, casino management can both make adjustments in
procedures that can increase the prospect of maximizing player value and make adjustments to
formulas that estimate player value to assure they more accurately reflect actual player value.

Casinos can control player value by changing the house advantage. Simply, reducing the house
advantage will lower the player value of those players taking advantage of the lower house
advantage. Likewise, the opposite is true. While this may seem obvious, casinos may make
marketing decisions they hope will attract additional players without adequately anticipating the
resulting decrease in player value. For example, a casino could decide to implement double-deck
blackjack games for which the basic strategy house advantage is 0.32% compared to 0.54% for
six-deck. When it does, it needs to consider that (a) the volume of play will need to increase,
often substantially, to meet the theoretical win generated before the change and (b) the individual
worth of the blackjack players who take advantage of the double-deck games will be less than
before the rule change. For example, all other things being equal, a blackjack player betting a
total of $12,000 and playing a six-deck game is worth about the same as blackjack player betting
a total of $20,000 and playing a double-deck game.

Another factor in player value that is partially casino controlled is rate of delivery, or hands per
hour. This can apply to both table game play and slot machines. With table games, it refers to
how the casino can increase the speed of play by dealing more hands per hour. This could
involve the use of automated shuffle machines, dealer training or game procedures. With slot
machines, it is the speed which the machine displays the play outcome and the time needed for
the player to activate a new play.

Most factors that impact player value are within the control of the player. Understanding how the
player controls these factors is important to the casino to assess the earning potential from the
player. Of course, all these factors are important, but an emphasis is often placed on time played.
This is because volatility is reduced the longer the player is betting. Again, the casino’s best
friend is the law of large numbers. The number of hands a player plays is of fundamental
importance to casino operations. Playing many hands takes time. Time is the casino’s best ally.
The more hands a person plays, the greater the casino’s earning potential. This makes sense
because if the house advantage on a given bet is 1.36%, the expected win on a single hundred-
dollar bet is $1.36 and the expected win on 100 hundred-dollar bets is $1.36 x 100, or $136.
Moreover, as the number of hands played increases, the more likely it is that the casino’s actual
win from that player will approach the theoretical win and the more likely the casino will realize
the full earning potential from the player. Returning to the example of the two roulette players,
because the second player is betting more than the first player, over time the casino will win
more from the $400 average bet player than the $200 player. Therefore, he is more important to
retain as a player.

For slot machines, tracking the exact earning potential of players is easy with the player tracking
systems in use at most casinos because all factors needed to make the determination are
available. Slot machine manufacturers state the house advantage of the machines in the “par”
sheets. The meters on the machine record the denominations, number and total of all coins
played.

Determination of player value is more difficult for table games. Unlike slots, where each coin or
credit played can be recorded, a player’s table game betting handle, betting pattern (the particular
bets placed) and, in cases where relevant, the skill of the player, are not precisely known and
must be estimated through observation.
For many table games, actual earning potential is the amount of each bet made by the player
multiplied by the number of hands played multiplied by the house advantage for the game played
as adjusted for the skill level of the player. For example, take a very good blackjack player who
bets $10 per hand and plays 70 hands at blackjack. If he is playing six-deck blackjack, the house
advantage is about 0.54%.
The earning potential of this player is computed as follows:

$10 x 70 x 0.54% or $3.78.158

In other words, earning potential = $3.78.

A number of uncertainties, however, come into play in determining all of the factors in the
formula. For example, can the casino accurately track the amount of every wager made by the
player? Some players will not bet the same amount for every hand.159 This can be accomplished
by either sophisticated tracking systems or simple manpower that is needed to track every bet of
every player. The ability to obtain perfect information on all the player-controlled factors can be
expensive, either the cost of manpower or of the tracking system. In most cases, casino
management has historically decided that these costs are greater than the costs of errors resulting
from more rudimentary systems. Therefore, most casinos must rely on assumptions and averages
to estimate theoretical win or actual earning potential of table game players.

Casinos often will develop a system to measure theoretical win in a manner that is simple to
administer. The system a casino uses to estimate theoretical win from players is generally
referred to as a rating system. This is because under most of these systems, each player whose
play is observed for purposes of estimating theoretical win is rated according to some scale.
Casinos may use different methods and terminology for rating players according to the casino’s
theoretical win. A “rated” player is one whose play will be reviewed by the casino, usually for
the purposes of determining comps and to qualify players for marketing purposes.

Conceptually, estimating the average amount of each bet by a player may appear simple enough.
Some players, however, will periodically vary their bets throughout the course of their game
play. This requires the pit personnel to observe their play and come up with a reasonable figure
for the player’s average bet for the time period observed. An error may occur if the pit personnel
observes only a limited sample of the hands, is careless in his observations, or makes a mistake
in calculation, among other things. For example, a player hoping to obtain a higher player rating
(to maximize comps) may manipulate his average bet. This can include wagering larger amounts
when the supervisor is watching, such as when he begins play, or when the supervisor is
obviously recording ratings. The player can also hide chips so that it appears he is losing rapidly
to deceive the rating supervisor into believing he is betting at the rate of his highest bet.
Casinos rely on averages to determine the number of hands played by the player. Many factors
can influence how many hands are played, including the skill level of the dealers, the number of
players at the table, the speed of players in making decisions, and whether a shuffler is used. For
example, a person playing alone at the blackjack table may average about 200 hands per hour,
while the same player playing at a full table may average only 50 hands per hour. The difference
in the earning power between these two players is about 400%. Despite that the player playing
alone is worth four times as much to the casino as the same player playing at a full table, many
casino evaluations do not account for these differences. Players hoping to obtain higher player
ratings may use methods to play as few hands as possible in a given time period. These methods
can include (a) playing on crowded tables, (b) choosing dealers who are slow or are distracted
by talkative players, (c) playing slow, (d) taking many breaks, or (d) other similar tactics. The
casino needs to be sensitive that sometimes the player has no choice but to play at a crowded
table when the casino is busy. In this case, the casino may want to comp the player consistently
between trips despite that his play is slower during busy times in the casino.

Casinos face another difficulty in determining how much they can expect to win from their
players. In certain games, like blackjack and craps, the house advantage varies. With blackjack,
the edge varies with the skill of the player, while at craps the house advantage depends on which
bets a player selects. Assessing the overall advantage the house enjoys over an individual player
in these games is a formidable task.

Incredibly, some casinos will use the same house advantage for determining theoretical win for
every table game, regardless of whether it is blackjack (average HA ≈ 2%), craps (average HA ≈
1.5%), baccarat (average HA ≈ 1.2%) or roulette (average HA ≈ 5.3%). While undoubtedly
easier to administer, it allows for certain problems and abuses. One problem is that it
overcompensates certain players at the expense of other players. For example, a good $100 per
hand blackjack player with, say, a 0.6% disadvantage, has a lower value to the casino than a
$100 per hand baccarat player because baccarat has a typical house advantage of 1.2%. All other
things being equal, the expected return on the blackjack player will be about half that of the
baccarat player. Yet, the comp privileges for each would be the same under a rating system that
uses the same house advantage for each game. This has led some commentators to suggest
methods whereby players can exploit deficiencies in casino rating systems to obtain more in
comps than their actual expected value to the casino.

Probably the most popular form of rebates is based on theoretical loss. In principle, this is a
straightforward method to comp valued players. It is a program that can be easily adapted to both
table and slot machine play. Slot play can be tracked through the use of player slot club cards.
Once the player inserts his or her card into the slot machine, it tracks every coin the player plays.
From this, the casino can pay back a small percentage of each coin played.
As explained previously, accurately determining a player’s theoretical win (earning potential) for
table games is not possible without tracking every bet made by the player. In the case of high
rollers, however, the only efficient use of manpower may be to have a casino employee
physically observe and record every single bet placed by the premium player so that theoretical
win can be accurately assessed. This is because the amount of the rebate to high rollers can be
significant. For example, if a casino is willing to pay back 50% of theoretical loss, then the
player would receive 50% of the house advantage for each $1 wagered over the course of play.
Thus, if the edge were 1.2%, then 0.6% of each dollar wagered would be paid back to the player,
regardless of whether the player won or lost. The casino would retain the advantage at 50% of
the usual value, or 0.6%. When the player is a million dollar player, this amount becomes worth
tracking by hand.

Dead chips are non-negotiable gaming chips that can be used for gambling but cannot be
exchanged for regular chips or cash. In a typical program the player purchases a certain amount
of non-negotiable chips at a discounted rate, usually through a voucher called a chip warrant, and
uses these chips for wagers just like normal chips. Once the player has used all the non-
negotiables, the remaining regular chips from winning wagers may be redeemed for cash. Some
casinos, particularly in Asia and Australia, use dead chip programs as a marketing strategy to
attract higher-end players. The general idea is to give players a bonus or refund on non-
negotiable gaming chips in return for a commitment to a certain level of play. These programs
can be effective, but because playing with dead chips alters the game advantage, a casino must
be careful in setting appropriate bonus amounts. Awarding a bonus or refund can sometimes
result in an effective house advantage that is too low to offset operating costs, and may even
result in the player having the edge.

Dead chip programs are usually targeted toward ‘junkets.’ A junket is a group of players who
travel to a casino specifically for the purpose of gaming. The travel is prearranged through a
junket representative and the players’ costs associated with traveling to and staying at the casino
are absorbed by the casino. The general process of a typical dead chip program is as follows. A
junket group arrives at a casino and players deposit front monies with the casino in exchange for
chip purchase vouchers, also known as cheque credits, or chip warrants. The vouchers are then
used to purchase non-negotiable chips which the players use for wagers during the course of
play. Losing wagers are taken by the dealer while winning wagers are paid in regular (‘live’)
chips and the original wager remains as non-negotiable chips. Once a player has exhausted his
(most junket players are men) non-negotiable chips, he takes his live chip winnings and
exchanges these for more non-negotiable chip vouchers at the casino cage. The value of this
exchange is recorded and added to the player’s/group’s non-negotiable chip purchase voucher
schedule. When the junket group departs, the non-negotiable chip purchase voucher is totaled
and all remaining non-negotiable chip purchase vouchers are deducted as well as any non-
negotiable chips held. This provides a value for the turnover – the cumulative amount of money
wagered – upon which a commission is paid to the junket group operators. All chip purchase
vouchers and chips are converted to cash or deducted from personal cheques held with a new
cheque value written out, which may or may not include commission. The commission payment,
then, is calculated essentially on the loss of non-negotiable chips over the period.

A few of the larger casinos in jurisdictions outside the Asia-Pacific region have recently
implemented or are considering implementing dead chip programs for premium players. A
version that surfaced in Las Vegas gives players a ‘bonus’ of additional free non-negotiable
chips at the time of purchase. A player wagering at baccarat, for example, may receive $103,000
in non-negotiable chips for his $100,000 in cash, a 3% dead chip bonus. Subsequent purchases
may carry a reduced bonus. Thus, once the $103,000 in non-negotiable chips is gone, the player
may receive $102,000 in non-negotiable chips for the second purchase of $100,000 in cash (2%
bonus). For the third and all subsequent purchases of $100,000 in cash, the player may be given
$101,000 in non-negotiables (1% bonus). Note that the player receives the entire buy-in amount
in dead chips, and not, for example, $100,000 in negotiable chips and $3,000 in dead chips.
Although it is obvious the effect of such a bonus program is to reduce the advantage the casino
holds over the player, the precise amount of the effect is not immediately evident.

The reduction in house advantage associated with the dead chip bonus model is a function of the
average length of time a chip will be wagered before lost, which in turn depends on the
probability of losing the wager. The logic can be seen clearly with a simple coin-tossing
experiment. Consider a wager on the outcome heads in a series of tosses of an honest coin. With
heads and tails equally likely, a bet on heads will lose every other toss on average; thus the
average number of tosses before the wager is lost is two. If the coin is biased so that the
probability of heads is equal to, then the average number of tosses before a wager on heads is
lost is ten. Generally, if PL is the probability of losing for each of a series of independent wagers,
then 1/PL is the average number of trials (wagers) until a bet is lost. This result can be viewed in
terms of the well-known geometric probability model. If p represents the probability of success
on each of a series of independent Bernoulli trials and X = the number of trials until the first
success occurs, then the expected value of X is equal to 1/p. In double-zero roulette, for example,
a straight-up single number bet will last on average 1/(37/38) = 1.027 trials and an even-money
wager will last on average 1/(20/38) = 1.90 trials. Thus, the ratio of actual turnover to non-
negotiable losses is a function of the probabilities associated with the individual game and
particular wager placed. For roulette this ratio is close to one on single number straight-up
wagers, close to one-and-a-half for dozens and columns bets, and close to two for even-money
wagers.

This helps explain why in a game like baccarat, casino operators are willing to pay commissions
– usually 1.5% to 2.0% – that are greater than the house advantage (which is around 1.2% to
1.8%) on non-negotiable chips. This is because for baccarat, the effective house advantage on
non-negotiable chips can be viewed as being twice that of the normal value due to the ratio of
actual versus non-negotiable turnover being about 2:1 (slightly more for banker bets; slightly less
for player bets).

The following example may help in understanding the formal mathematical results. Suppose a
player pays $100,000 on buy-in and receives $101,500 worth of non-negotiable chips, reflecting
a 1.5% bonus. Assume further this player wagers these non-negotiables, one chip per wager, only
on the baccarat banker, with probabilities of winning and losing equal to PW = .458597 and PL =
.446247 respectively (probability of tie is .095156). On average, it will take 227,452.7 wagers
(101,500/.446246) before all the non-negotiable chips are lost, after which time the player will
have received 99,093.7 live chips for wagers won (.458597 x 227,452.7 x .95). Thus the
casino profit is $100,000 – $99,093.7 = $906.30, equating to an effective house advantage of
0.40% (906.3/227,452.7 = .0040). This reasoning generalizes to the following result.

Fundamental Theorem of Dead Chips: Suppose a player receives X(1+B) non-negotiable chips in
exchange for X dollars, so that B represents the dead chip bonus.

Suppose further the player wagers these non-negotiable chips until lost, one per wager on a series
of independent wagers each with probability of winning w units equal to PW and probability of
losing one unit equal to PL. It is assumed that all wagers are made with non-negotiable chips;
losing wagers are taken by the dealer, and winning wagers are paid in regular chips with the
original wager remaining as a non-negotiable chip.

Proof: To prove (1), note that on average a non-negotiable chip unit will be wagered 1/PL times
before it is lost, and therefore the average number of units wagered with X(1+B) non-negotiable
chips before all are lost is X(1+B)/PL. Thus the player’s expected win is [X 1( + B /) PL]wPW
units and the house retains X − [X 1( + B /) PL]wPW units.

The fundamental theorem can be extended to results for multiple rounds of dead chip purchases,
wagers involving multiple winning outcomes (for example, slots, keno, video poker, and certain
bets in craps), and mixtures of wagers. Proofs are analogous to that of the preceding fundamental
theorem and are omitted. In the results that follow, it is assumed that all wagers are made with
non-negotiable chips, losing wagers are taken by the dealer, and winning wagers are paid in
regular chips with the original wager remaining as a non-negotiable chip.

Multiple Rounds. Suppose a player purchases k rounds of non-negotiable chips, exchanging Xi

dollars for Xi(1+Bi) non-negotiable chips on round i, for i = 1, 2, ...k. Thus B1, B2, ...Bk
represent the dead chip bonuses on each of these purchases. Suppose further the player wagers
these non-negotiable chips until lost, one per wager on a series of independent wagers each with
probability of winning w units equal to PW and probability of losing one unit equal to PL.
In some programs, the dead chip bonus is awarded as a commission in cash after the termination
of play rather than at the time of the buy-in. In this situation, the effective house advantage
computation is straightforward, and can be shown to be HA = HA − BPL* .

Note that giving the dead chip bonus as a commission at the end of play will always result in a
lower effective house advantage than giving the bonus at the time of buy-in, all else equal.

Although the laws of probability warrant that the actual win percentage will converge on the
theoretical win percentage, management needs to be concerned about potentially large deviations
from the expectation that may occur in the short run. The margin of error figures in the preceding
tables provide upper bounds, with 95% confidence, on the size of this deviation for several
numbers of trials (hands) and can be used to obtain 95% confidence limits for the actual win rate.
About baccarat volatility, however, the level of turnover required for the actual win percentage to
fall within a small level of tolerance from the mean can be astronomical, so analysts and senior
management need to have very strong stomachs if involved in baccarat high end play.180 Even if
the house advantage holds through the duration of play, the casino will still have to foot the bill
for room, food and beverage, airfare allowance and gaming taxes. Thanks to the higher house
advantage in roulette, a casino is on more solid ground when it offers dead chip bonus to roulette
players. In fact the maximum bonus in roulette for a player betting ‘evens’ (any of the even-
money bets, such as red or black) could be twice as high as a baccarat player betting on the
banker. Having thus determined the house advantage and the resulting bonus a casino could
safely live with for the various games, bonuses should be based on specific games as well as a
punter’s total buy-in for dead chips. Of course, the house advantage can be even higher in other
games – keno and some slots, for example – but these are not usually the games of choice for
high-rollers.

The situation gets most uncertain from a casino’s standpoint when dead chip players are allowed
to play a variety of games. Since house advantage is a statistical function of the particular game
played and particular wagers made, establishing a realistic bonus for a player betting on both
blackjack and craps becomes an extremely unwieldy proposition. Unless the precise mixture of
games to be played can be controlled in some way, a casino may need to be more conservative if
allowing a mixture, perhaps using as the basis for the dead chip bonus the game that would result
in the lowest bonus percentage.

Otherwise, it makes sense to design dead chip bonus programs under the condition that the chips
could be used for play in only one specified game. Note too, that for games that involve an
element of skill, the house advantage will vary depending on the player. In such situations,
calculations assuming optimal strategy will provide a baseline reference for skilled players, but
the actual house advantage will be higher for those players who do not use optimal strategy. Of
course, for many games, including baccarat, the game most associated with dead chip program
play, skill is not an issue.
Another implication of the mathematics of dead chips relates to the wisdom of providing the
dead chip bonus at the time of buy-in as opposed to at the termination of play. Giving a dead
chip bonus at the end of play will always result in a lower effective house advantage than giving
the same bonus at the time of buy-in. Given slim operating margins, casinos are better off giving
commission at the time of buy-in. Although many players may be unaware of the mathematical
intricacies surrounding this choice, more savvy players and junket representatives will seek out
dead chip programs offering commission at the end of play rather than at buy-in.

Besides considering the win rate volatility in deciding commissions, other factors such as
volume, bet-limits, incentives and credit risk also need to be factored in. In setting betting limits
associated with high-rollers, one should pay attention to capital reserves of the company, the
company’s propensity for risk, fixed and variable expenses associated with the operation, the
total handle, and so on. Not too many companies (or their shareholders) would be thrilled in the
event of being forced to utilize their capital reserves to pay for table losses or fixed expenses.
Incentives accrued to high-rollers should be accurately estimated in order to arrive at the bonus
figure to award on dead chips. Incentives typically total somewhere between 50% and 70% of
theoretical win. This allows little room for bonuses or commissions. Bear in mind that dealing
with high net-worth individuals means that you are often
negotiating with very influential businesspeople who demand (and usually get) the best of
everything. Costs associated with airfare, hotel suites, golf course fees, and food and beverages
do add up, and offering these as comps cuts significantly into the casino’s theoretical win.
Furthermore, the credit risk associated with high-rollers needs to be factored in when
determining the bonus amount. As the industry saying goes, ‘You have to win the money twice.’
Since most high-rollers play on credit, casinos have to win the money over the tables and they
also have to win by getting paid, which may be neither very prompt nor certain.

The mathematics underlying dead chip programs sheds further light on the profitability and risk
associated with high-end play. In offering incentives to high-rollers, casinos may sacrifice a
substantial portion of their house advantage, thus creating a significant dent in profitability.
Some authors181 have suggested that instead of continuing to offer high incentives primarily to
premium players, casinos may wish to consider marketing strategies and programs that target the
$25 to $200 bettors.

These mid-level punters constitute the backbone of the casino gaming business; offering rebate
and discount programs will lower the price of the game for these players and, at the same time,
have a positive impact on casino revenues and customer retention. Although it is natural to view
programs such as the dead chip offerings in terms of a player’s single visit to the casino, it would
not be difficult to extend the analysis to apply to multiple visits. With the current emphasis on
customer relationship management (CRM), it is advisable to consider player profitability over
the entire length of a player’s relationship with the casino. Given the volatility associated with
short-term play of the high-end segment, it would make sense to make every attempt toward
enhancing the lifetime value of high-rollers. At a time when casinos all over the world fiercely
compete for a share of the high-rollers’ wallets, engendering loyalty among high-rollers is a
formidable task.

Dead chip programs have been used by casinos for some time – particularly in the Asia-Pacific
region – but the mathematics underlying these programs has not in the past been well-
understood. Comprehending the effect these programs have on game advantage is crucial to the
casino’s ability to offer an attractive dead chip package while still conducting sound business. As
the popularity of dead chip programs grows, so does the need for a clear understanding of the
precise impact of this marketing initiative on the mathematical edge. Recent history of several
casinos ending up on the losing side of the table vis-à-vis high-rollers underscores the
importance of thoroughly understanding the mathematical realities behind incentive programs.
Many casino executives assume that the high-roller segment is inherently profitable. Combine
that assumption with a misunderstanding of gaming mathematics
and you have the makings of a downward profit spiral.

Bear in mind that the mathematics only guarantees the actual win rate will be close to the house
advantage after a very large number of trials. In the short term, deviations from the theoretical
win will inevitably occur, making dealing to high-rollers an inherently risky proposition for most
casinos. Decisions on commission should therefore be made in conjunction with decisions on bet
limits, other incentives and credit risk. All these decisions should be formulated against the
backdrop of a casino’s financial reserves, its tolerance for risk, shareholder expectations and
expected turnover volumes. As one author notes, ‘This (high-roller segment) can be a high risk.

By tracking both the minimum equivalent bet and the observed loss, the casino can give comps
when the player loses a substantial amount of money in a short period. In this case, because play
is shortened by a run of bad luck, the player’s theoretical loss would be far less than their actual
loss. The observed loss could still be used to reward the player for his play and meet his
expectations.

Most casinos will track both theoretical win and observed losses. The reason for this is simply
that some players will lose most if not all of their bankroll quickly either because they are
extremely bad players or they experience an incredible streak of bad fortune. Although the house
advantage represents the average outcome in any game, there is a bell-shaped (normal)
distribution of outcomes centered on the expected house win. Some players will end up at the
lower end of the house win distribution with a large player win. Others, perhaps merely
“unlucky,” will be at the upper end, with a resulting large player loss and house win. Some
casinos will grant comps or rebates if the player loses all or more (often 80% or more) of their
credit line or front money deposit even if they do not meet the criteria that the casino sets for
minimum theoretical loss.
As one of the most heavily regulated industries and one built on the mathematics of games,
mathematical issues often arise in the regulatory process. While regulation often attempts to
achieve many goals, most regulatory systems share common objectives: keep the games fair and
honest and assure that players are paid if they win. Fairness and honesty are two different
concepts. A casino can be honest, but not fair. Honesty refers to whether the casino offers games
whose chance elements are random. For example, a slot machine is honest if the outcome of each
play is not predetermined in the casino’s favor. Fairness is very much a mathematically based
consideration.

Regulations related to honesty are generally twofold. The first is to assure that the games
produce random results and players have an equal opportunity to win each prize offered on any
given play. Moreover, these results should conform to the approved or represented odds of the
game being offered. These regulations can range from standards for random number generators
in gaming devices to methodologies to test whether the games are performing as mathematically
expected. This chapter
covers these regulatory issues.

Fairness refers to the games being designed so that they do not take unreasonable advantage of
the player. For example, some slot machines in traditional casinos are designed to pay back on
average 95% of all wagers accepted. The five-percent retained is the casino’s profit. This is a
reasonable amount for the casino to retain to pay for its capital costs, operating expenses and a
fair profit. Setting the machines to retain 40%, however, may not be reasonable.

In most industries, regulatory price controls mean the government sets prices for services. For
example, a state public service commission may set basic telephone rates at $9.00 per month. In
the casino industry, government sets rates by dictating odds of the games. For example, a
jurisdiction may require casinos to offer no more than double odds on craps, and their roulette
table must have 0 and 00. These regulations set prices in the casino industry; so do requirements
that a casino use multiple decks in blackjack. With slot machines, these standards are usually
defined by setting a minimum payout, i.e., prohibiting the casino from theoretically retaining
more than a set amount of each dollar wagered. For example, many jurisdictions provide that
gaming devices must have a minimum payout, such as 70%. This means that the maximum
average price that the casino operator can charge is $.30 for each dollar
played.

While good policy reasons may exist in some jurisdictions to set maximum price, the
government may have difficulty in fairly setting standards. Where the government decides to set
the minimum rate, it should consider the maximum permissible bet, the maximum average price
(or minimum payout), and hours of operation. Most jurisdictions attempt to set maximum prices
by mandating the maximum theoretical win the casino can derive from a game. The word “fair”
is subject to consideration of many factors, many of which are difficult, if not impossible, to
define in regulations.

Regulations that set maximum price restrictions tend to be overly general. For example,
regulations can require casinos to pay back 80% on slot machines. This approach has practical
problems. These types of required paybacks may or may not be fair to the player depending on
the minimum wager. If an 80% payback applies to a nickel gaming device, the average cost per
hour for the player to play this machine with one coin-in and 325 pulls per hour is $3.25. This
may be fair to the player, but unacceptable to the casino operator if the cost of maintaining a
gaming device exceeds $3.25. As a result, the casino may not offer 5-cent slot machines if it is
limited to a 20% advantage. In contrast, if the player is playing a dollar machine at a 20%
advantage with three coins in, the average price is $195 per hour.188 This is not a fair advantage
over the player, If a jurisdiction wishes to set prices, it must consider the gaming device
denomination or table limits in deciding what is fair to both the player and the casino.

The result of placing broad maximum price restrictions that apply regardless of denomination of
game or whether the game is a table or gaming device, risks being unfair to both the player and
casino. For example, such pricing requirements may actually increase the price of playing if the
casino must eliminate low denomination machines or table games that become unprofitable to
offer to players because of maximum price restrictions. This can be partially addressed by having
different maximum pricing for different denominations. Other factors influence price. These
include the pace of the game and, in many cases, the player’s skill level. Therefore, on table
games or other games, the government may want to consider the average number of plays per
hour and allow the casino to have higher theoretical win in the slower games.

Most frequently, regulators set prices in the gaming industry by requiring the casinos to
theoretically pay out a mathematically demonstrable percentage of all amounts wagered, e.g., not
less than eighty percent (80%). This approach presents some problems. Using theoretical wins
will provide an average price per play, but it does not provide meaningful comparisons between
different types of games or an easy standard for calculating costs. For example, a casino may
charge more “per hand” for blackjack than a gaming device but make more per hour on a gaming
device because the player can play two to six times more plays per hour.

Some regulators may realize all the practical problems with trying to set maximum price by
dictating theoretical win and look for some other standard. The most obvious is to review the
average “hold” from the game as experienced over a trial period or other time period. Attempting
to set minimum prices by dictating the maximum “hold” for casino games raises other problems.
As noted earlier, hold is important to the casino operator. It gives the operator a rough estimate
of how the table games are performing. It is not, however, an accurate tool to test the fairness of
the games. Casino operators use “hold” for casino games only because no better method of
evaluating performance is available. Unlike gaming devices, no method exists to calculate handle
on casino games. Thus, casinos rely on hold percentages to give them basic information on how
the games are performing.

Regulators, however, should not rely on hold as a substitution for theoretical win percentage in
deciding whether the game meets minimum standards. Many factors other than theoretical win
can influence hold percentage, such as game interest, dealer conduct, and theft. A better standard
is average price per hour. Price per hour is a function of average price per play multiplied by the
average number of plays per hour. A second set of problems concern the inherent difficulty of
setting maximum prices on table games. Governments can easily set maximum theoretical wins
based on the denomination of the game when dealing with gaming devices. It is far more
difficult, however, when dealing with table games. Except in jurisdictions with maximum bet
restrictions, table games often allow a range of bets, e.g., from $5 to $100. Thus, one player at
the table may be betting $5 per play while another player is betting $500 per play (five different
crap bets at $100 per bet).Another practical consideration is how to treat a game, such as craps,
that allows for different bets. Some craps bets have low house advantages, while others are high.

Cheating in casino gambling has three broad categories. The first is in simply predetermining the
outcome of the game. This would occur, for example, where the deck in a card game has the
cards in a predetermined order or a skilled dice player can “slide” the dice so that a desired
combination will appear.

The second category is “past posting,” or in other words, placing or removing a bet after the
outcome has already been determined. A skilled cheat can increase or decrease his bet after the
game ends. For example, a cheat with a losing hand in blackjack can "pinch the bet" by palming
one chip in the stack wagered. If the cheat adds a chip after learning that he has a winning hand,
he pressed or past-posted the bet. Both pressing and past posting are illegal.

The third category is to obtain knowledge unavailable to other players that gives that player a
theoretical advantage over other players or the house. A good example of this is shining in
blackjack. Shining most often occurs in blackjack. Shining is the method of learning the dealer's
hole card using any reflective device. This could include a metallic cigarette lighter, a facet on a
ring, a polished fingernail, or a small mirror. If a person knows the dealer's hole card, he has an
advantage over the casino through having additional information to figure out whether to "double
down," i.e., relinquishing his hand and only losing half his bet, or determining whether to hit or
stand. The statistical advantage gained by the player that can learn the dealer’s hole card is
substantial. Knowledge of the dealer’s hole card on every hand can give a skilled player who
would otherwise play about even or at a slight disadvantage more
than a 9% advantage.

To cover these possibilities, most casino states have laws that prohibit a person from placing,
increasing, or decreasing a bet or to decide course of play after acquiring knowledge, not
available to all players, of the outcome of the game or any event that affects the outcome of the
game.

The gimmick, however, was that the alteration in other rules more than compensated for the
player’s advantage in seeing the dealer’s hole card. Rule modifications used to offset the player
advantage have varied, but include the dealer winning tied hands (except possibly a player’s
blackjack), paying even money instead of the normal 3 to 2 for blackjacks, and restrictions on
doubling down and resplitting. Depending on the combination of rules in effect and the number
of decks used, Double Exposure Blackjack typically gives the casino a 0.3% - 2.0% advantage.

Advantage play is often associated with card counting in the game of blackjack. The edge gained
by a card-counter depends on several factors, including the number of decks in play, the
penetration (how far into the pack cards are dealt before reshuffling), the rules of the game, the
bet spread, and the system used. A skilled counter will typically have a 0.5% to 1.5% advantage
over the house.229 More decks and shallow penetration (not dealing very far into the pack before
reshuffling) tend to decrease the counter’s advantage.

Video poker as played in casinos is a game of mixed skill and chance. In a typical video poker
game, the player randomly draws five cards from a deck of 52 cards and has the option of
keeping or discarding up to all of those cards. For each card discarded, the player will randomly
draw an additional card from the cards remaining in the original 52 card deck. The combination
of cards remaining in the player’s hand is compared to a pay chart. If the player has any of the
required combinations, he is paid according to the pay chart. The skill element involved in video
poker is the determination of which cards to hold or discard. Some casinos offer video poker
machines that when played at optimum skill pay back more on average than they would collect.
The casinos rely on players not being able to play at optimum skill to insure a profit.230 Despite
this, some players can play at optimum skill and have an advantage albeit slight (about .5%) over
the casino.

Progressive slots are slot machines in which one or more of the payouts increase by a set amount
for each coin or credit played that does not result in the player winning that payout.231 For
example, suppose a progressive slot machine has a top payout of $100 and 5 cents of every dollar
bet increases the progressive jackpot. If the player plays the machine for $1 and does not win, the
progressive jackpot will increase to $100.05. The progressive jackpot will increase in this way
until it is won and it will then be reset to its original starting point. If the progressive jackpot
increases to a certain point without being won, then the theoretical payout results in a positive
player advantage.

This slight statistical advantage is what attracts the professional slot teams. These are organized
and financed teams of professional slot players that attempt to exploit those progressive slots that
have a positive player expectation. Only certain progressive slot machines will meet the slot
team’s criteria. First, the number of slot machines that are linked to the progressive jackpot must
be manageable. The team needs to monopolize all the slot machines to avoid the risk that a non-
member will win the jackpot. Second, the statistical frequency of hitting the jackpot must be
consistent with the slot team’s bankroll. No slot team has an unlimited bankroll. Based on
probabilities, the team needs to have enough cash on hand to play all the linked machines with
the progressive jackpot until one team member hits the progressive jackpot. The slot team can
use the Poisson probability distribution to determine the probability of hitting the jackpot over
the course of a specified number of plays. The Poisson distribution (named after the 19th century
French mathematician Simeon Poisson) is commonly used to calculate the probability of rare
events.

The probabilities in the table above refer to the behavior of an individual slot machine for which
the probability of hitting the jackpot on a single play is a little better than one in a million
(1/884,736 = .00000113). Over the course of one-million plays on this machine, the probability
of hitting the jackpot is 67.7%, with this probability increasing as the number of plays increases.
If several machines are considered, the probability that at least one hits the jackpot during the
course of a fixed number of plays will be greater than the probability that one specific machine
will hit during this same number of plays.

If the probability of hitting the progressive jackpot is one in ten million (as might be the case on
a different machine), the slot team would need an astronomical bankroll to have reasonable
assurances that it would hit the jackpot before running out of money. Therefore, slot teams avoid
the “mega” jackpot progressive that have very infrequent hits or winners. To minimize their
potential exposure, they concentrate on progressive carousels that feature lower jackpots with a
higher frequency of payouts.

Many commercial gambling games rely on proper randomization to ensure fairness and maintain
the desired house advantage or players’ win/loss rate. Gaming regulators, for example,
acknowledge the importance of randomization in gaming devices such as video poker, slot
machines, and shuffle machines by requiring that they meet minimum confidence levels on
standard statistical tests. Indeed, the greatest regulatory concern is that the device selects the
cards or symbols within acceptable levels of randomness.

In card games, the random aspect of the game is the shuffle of the cards by the casino. This is
done either by a mechanical shuffler or by hand. Mechanical shufflers are usually reviewed and
approved by regulatory authorities to assure that the outcome of every shuffle will appear
random. Non-machine card shuffles are not typically subject to such regulatory scrutiny. A
nonrandom shuffling process may alter the odds of a game, if players can use a predictable
pattern inherent in the shuffle process as an advantage play.
Shuffle tracking may allow a player to follow segments of cards through a shuffle and by so
doing know roughly when these segments will appear during ensuing rounds of play. If such
tracking can be employed in a situation when slugs of favorable or unfavorable cards have been
identified in the pre-shuffle pack (for example, in blackjack where multi-deck shoe games are
quite common), a player could then utilize knowledge of the slug locations during the play of the
next shoe by adjusting strategy and bet sizes accordingly. Such shuffle-tracking techniques are
not possible, of course, if the shuffle is random.

Using standard casino playing cards, two professional dealers shuffled a variety of six-deck, two-
deck, and single-decks. Cards were ordered from 1 to n prior to each shuffle iteration and the
post-shuffle position recorded for each card. Thus for each iteration of each shuffle procedure,
data consists of the permutation array π(1), π(2), ..., π(n)), where π(i) is the pre-shuffle position
of the card whose post-shuffle position is i. Six-deck shuffles were chosen to reflect procedures
in use at selected major Las Vegas Strip casinos in early 1997. Thirty shuffle iterations were
obtained representing six different shuffle procedures (see below for further details). While the
data here reflects shuffles executed according to these procedures, inconsistencies exist in
dealers’ adherence to the house procedure on live games.

Given casinos have fixed shuffle procedures, when one such procedure results in a non-random
shuffle (particularly as non-random as those in the empirical study cited above) an advantage
player could observe, then simulate and study this procedure to determine where specific clumps
of cards from the pre-shuffle deck will likely be located in the post-shuffle deck. Once this
“mapping” from the pre- to post-shuffle decks has been accomplished, a player can utilize this
knowledge as a method of advantage play by predicting when favorable and unfavorable cards
will be dealt after the (non-random) shuffle and then varying bet size and strategy accordingly.
This “shuffle tracking” method of exploiting non-random shuffles to identify clumps of
favorable or unfavorable cards in the post-shuffle deck has been refined to advantage play
techniques such as ace prediction, ace tracking, sequence tracking, and key
carding.

Hole-carding is a technique used primarily in blackjack to learn the value of the dealer’s hole
card before the player needs to make a decision on how to play his or her hand. The advantage
derived from hole-carding can be more substantial than card counting.

Most hole-carding is done intentionally. One surveillance expert described a typical team that
took advantage of a weak dealer. In that case, the team consisted of two people. One person was
in the “third base side,” meaning the seat nearest to the dealer’s right hand, and the other in the
first spot, nearest the dealer’s left hand. “Not only was the third base player slouching in his seat,
his signals to the other player were simple. The hole-carder on the third-base side of the game
had a stock of green ($25 chips) and red ($5 chips). To signal stand, he touched the red
(stand/stop), to signal hit, he touched the green (hit/go) and for insurance he tapped the green
stack with a chip pinched between his fingers to represent an ‘I’ (insurance). The dealer later was
proven to be a weak dealer and not in the group....

In the case of the player sitting on the third base side, slouching, he was right “74 percent of the
time at reading the hole card.” If a person knows the dealer’s hole card, the player has an
advantage over the casino through having additional information to figure out whether to double
his bet, surrender a hand (thus, only losing half a bet), or determining whether to hit or stand.

The impact that “hole-carding” has on the house advantage will vary depending on the advantage
player’s proclivity to correctly identify the hole card and to what advantage he uses the
information. If the player knows the dealer’s hole card every time and played every hand to
maximum advantage, it would result in a 10% advantage over the house. The typical advantage
player, however, would not play to maximum advantage because hitting a 19 against the dealer’s
20 would be too suspicious and draw inquiry.

The “cherry squeeze” or “handle popping” involved the handle manipulation of mechanical or
electro-mechanical slot machines to control the reel alignment. The facts of one case revealed
that a player pulled the slot machine handle down two-thirds with his right hand, and then gently
hit the handle with his left hand. This activity did not damage the machine. By using this
technique, the manipulator stuck two reels on the payout line while the other one was spinning,
thus setting the machine up for a jackpot.

According to a 1986 Nevada Gaming Control Board memorandum, one suspect admitted
removing more than $60,000 per year using this method. The memorandum revealed that there
were about 40 other such cheats in Northern Nevada alone. The cheaters were careful not to
draw attention to themselves, and rarely emptied the contents of the machine. In a series of cases
involving slot handle manipulation, the Nevada Supreme Court began down a very curious path.
The court held that “handle popping” was not covered by the criminal statute because “[t]he
physical characteristics and potential pay offs of slot machines are not altered by handle
manipulators.”

The Nevada Supreme Court did not follow the same reasoning as slot manipulation in a case
involving dice sliding. In craps, players are given the opportunity to "toss" two dice that will
result in numbers from 2 to 12. This is the random event upon which the gaming contracts
between the casino and the players at that table will be decided. In dice sliding, the skilled player
is able to slide one or both dice across the table rather than tossing them and allowing them to
roll.262 Thus, a predetermined "roll" may be chosen and the outcome of the game manipulated
by the slider.

In Skipper v. State, a dice slider challenged his cheating conviction on the same ground as that
raised by the handle popping slot players in the earlier handle popping cases. Claiming that the
criminal statutes264 were unconstitutionally vague, Skipper argued that the statutes failed to
alert the average dice player that dice sliding constituted criminal conduct.

The Nevada Supreme Court rejected the argument finding that the rules of craps clearly require a
roll of the dice (the random event that is the basis of the contract). As the dice do not roll when
they are slid, dice sliding was found to be a clear violation of the rules of the game providing
adequate notice to the average player that sliding violates the anti-cheating statutes.

Again, the basis of the crime of dice sliding is that the person intentionally slides the dice to
impact the random element of the game. If a player improperly throws the dice so that they slide
instead of being rolled, the condition of a random event would not be met and the dice would be
thrown again. The thrower, however, has committed no crime.

Dice sliding should be contrasted with dice setting. The latter is a process by the shooter or
thrower in a dice game to exert some control over the numbers that the dice will show after they
have been tossed.266 The concept is that the casino issues rules that the thrower arranges the
dice in their hand before throwing them. By controlling (and standardizing) the motion of the
throw, dice setters claim to be able alter the random selection of results to favor certain dice
combinations that will give them an advantage over the casino.

Dice setting is different than dice sliding because the dice setter is working within the prescribed
rules of play to attempt to influence the roll. The casino prescribed the rule with the belief that
regardless of what proper method that the advantage player employs, it will not impact the
random-aspect of the throw. In other words, the casino considers dice setting to be more based
on superstition than science.

While some forms of advantage play such as card counting are “not considered cheating, nor is it
illegal,”267 Nevada casinos and casinos in most other states routinely exclude suspected
advantage players from gaming. These casinos may exclude advantage players at their discretion.
For example, no Nevada statute requires casinos to admit suspected card counters. Allowing a
casino to bar advantage players and others has a common law origin. “At common law,
proprietors of privately-owned places of entertainment and amusement were not obligated to
serve the general public.” In many states, the major gambling establishments are racetracks.
Cases involving exclusion of patrons from racetracks have provided substantial support for this
common law principle.

Casinos may counter some forms of advantage play simply by changing the rules of the play of
the game. These countermeasures are often statistically designed to reduce or eliminate the
player advantage.
Any advantage created through card counting, of course, is negated when the deck is shuffled.
Possible casino countermeasures against card counting include (1) preferential or at-will
shuffling,272 (2) using multiple decks,273 (3) changing maximum bet or restricting the bet size
to the table minimum for new players to the game,274 (4) having shills occupy all other seats at
table, (5) limiting players to a single wager,275 (6) prohibiting players from joining game in
midshoe, (7) short cut – placing cut card further from back,277 and (8) using an automatic
shuffler or a continuous shuffling shoe.

“The shuffle-at-will occurs when the dealer is instructed to reshuffle prior to the cut-card manner
because the casino card-counting team has determined that the shoe is player-favorable and card-
counters are suspected to be playing at a given table.” Technology also has evolved that could
allow casinos to achieve advantages over the advantage player. One example is the use of
“smart” gaming tables. These tables can use a variety of technologies to track wagers made,
cards dealt, and payouts. For example, one smart table can track every card that is dealt out of a
shoe in the game of blackjack. This table potentially could be used by a casino to call for a
reshuffle anytime the deck favors the advantage player.

Two strategies are available for slot teams that play progressive slot machines with a positive
player expectation. The first strategy is for the casino to retain the discretion to limit players to
playing one machine. Thus, when a slot team tries to monopolize a carousel, they will need more
team members and incur greater expense in doing so.