ABSTRACT

THE EFFICACY OF COUNTER-PRESSING AS AN OFFENSIVE-DEFENSIVE

PHILOSOPHY

By James Warwick

Analytics in soccer have been revolutionized over the last two decades. In
particular, models that can measure success using thousands of data points have emerged.
Counter-pressing as a philosophy has been linked to success. Despite this, there is
disagreement in the analytics community over the best method of measuring the
effectiveness of counter-pressing in soccer. This study will examine whether counter-
pressing is effective as an offensive-defensive philosophy. We will also explore how
video analysis can be used to measure the efficacy of counter-pressing. Video analysis
was used to interpret the performance of Miami University’s women’s soccer team
during the 2018 Fall MAC season. It was found that teams who pressed efficiently and
higher-centrally were more likely to have a successful game outcome. This paper
presents how counter-pressing is an effective offensive-defensive philosophy through the
successful use of video analysis.

THE EFFICACY OF COUNTER-PRESSING AS AN OFFENSIVE-DEFENSIVE
PHILOSOPHY

Thesis

Submitted to the

Faculty of Miami University

in partial fulfillment of

the requirements for the degree of

Master of Science

by

James Warwick

Miami University

Oxford, Ohio

2019

Advisor: Adam S. Beissel

Reader: Brody J. Ruihley

Reader: Christopher Henderson

©2019 James Alan Warwick

 This Thesis titled

THE EFFICACY OF COUNTER-PRESSING AS AN OFFENSIVE-DEFENSIVE

PHILOSOPHY

by

James Warwick

has been approved for publication by

The School of Education, Health and Society

and

Department of Kinesiology and Health

____________________________________________________
Dr. Adam S. Beissel

______________________________________________________
Dr. Brody J. Ruihley

_______________________________________________________
Dr. Christopher Henderson
 Table of Contents

Table of Contents Column1

List of Tables IV
List of Figures V
Introduction 1
Literature Review 2
Soccer Analytics 2
The Phenomenon of counter-pressing 4
Counter-pressing and success 5
Methodology 6
Results 8
Counter-pressing is an Indicator of Success 8
Discipline is Crucial 9
Counter-pressing across 90 minutes decreases 10
Discussion 12
References 15
Figures 19
Tables 29
Glossary of Terms 36

iii
 List of Tables

Table 1: Total counter-pressing actions of Miami University women’s soccer team and
opponents
Table 2: Counter-pressing efficiency of Miami University women’s soccer team and opponents
Table 3: Presses leading to shots for Miami University women’s soccer team and opponents
Table 4: Counter-pressing zones and percentage of presses for Miami University women’s
soccer team
Table 5: Counter-pressing averages per 15 minutes for Miami University women’s soccer team
Table 6: Counter-pressing action averages across final six games for Miami University women’s
soccer team
Table 7: High-central counter-pressing across entire MAC season for Miami University
women’s soccer team

iv
 List of Figures

Figure 1: Counter-pressing action symbols

Figure 2: Individual/team counter-pressing template
Figure 3: Example individual counter-pressing map
Figure 4: Example team counter-pressing map
Figure 5: High, mid, and low counter-pressing zones
Figure 6: Right, central, and left counter-pressing zones
Figure 7: 4-2-3-1 formation example
Figure 8: Miami University counter-pressing heatmap vs Kent State
Figure 9: Miami University counter-pressing heatmap vs Northern Illinois
Figure 10: Example of the FWD to CDM gap

v

Introduction

In contemporary society, big data is all around us. This year 2.5 quintillion bytes of data
were created every day (Marr, 2018). All aspects of life are quantified to make decisions:
predicting the stock markets, healthcare industries, supply chain management, etc. This big data
transformation has seemingly reduced human beings, their actions, and wider society to a
quantifiable cadre of statistics, metrics, and algorithms. Computers use machine learning to
calculate millions of data points. These machines can quantify decisions and simulate all results
to systematically determine the correct outcome (Mills, 2019).
This analytical revolution is taking over sport as well. Data and analytics are changing
the way the sports are contested and administered in a range of ways by: identifying labor market
inefficiencies, determining win optimizing in-game strategies, improving the physiological
development and health of athletes, and maximizing profitability through ticket sales,
sponsorship, and media reach. The sports industry will only increase its use of analytics as
technology for data collection improves. An NFL game can often yield over a million data points
using modern analytical technology (Ricky, 2019). In Baseball, the popularized term Moneyball
describes the analytics revolution and subsequent success of the Oakland A’s who used data to
scout and analyze players (Kleinschmidt, 2019). The Houston Rockets implemented
‘Moreyball’, a tactic based on analytics to produce points more efficiently (Lozada, 2016).
Individual teams such as Houston Rockets and Oakland A’s are prime examples of how analytics
are successful in sport. Analytics are often used to inform practice, determine player selection,
regulate player workloads, and create tactics. Indeed, you would be hard pressed to find a single
North American sport franchise not using data analytics in at least some aspect of their
operations.
Despite such transformations, the sports analytics revolution has been slowly adopted in
the sport of soccer. Although the sport analytics revolution has come later than in other sports
(Kidd, 2018) - some have argued this was due to the low scoring in each game (Brooks et al,
2016) - analytics today in soccer are commonplace, with numerous big data companies such as
OPTA providing data feedback to teams live. Analytics have provided success to those who
embrace it at all levels of soccer. For example, analysts and their data feedback at Lincoln City
FC was accredited for their English FA Cup success in 2016/17 (Brown, 2017). On the other end
of the soccer scale, elite level clubs such as Liverpool FC are also heavily investing in analytics.
Liverpool FC recently hired astrophysicist Ian Graham as the team’s Director of Research.
Through use of analytics, Graham built a model to track the performance of over 100,000 players
worldwide to help Liverpool’s scouting process (Schoenfeld, 2019). For Liverpool, this model
helped determine their tactics in the famous 18/19 Champions League semi-final against
Barcelona. Analytics in soccer have advanced from basic statistics such as passing percentages
and shooting efficiency to advanced predictive models such as xG, which tracks the probability
of scoring from a certain position on the field based upon a database of previous shots taken
(Goff, 2018). Analytic companies and soccer clubs are building large datasets to create models
that interpret the performance attributes of players. These models allow professional soccer clubs
to quantify performance attributes such as counter-pressing effectiveness.
Among the most important aspects of soccer analytics is the emergence counter-
pressing. Counter-pressing is an offensive-defensive philosophy which involves players
attempting to win possession from the opposing team as high up the field as possible
(Whitehouse, 2014). The philosophy emerged through Rinus Michels and the Dutch team Ajax

1

during the 1960’s. In recent years, the counter-pressing philosophy has been used successfully by
world-class soccer teams such as Liverpool, Manchester City, and Bayern Munich. Even in the
2018 World Cup, teams that do not counter-press have been decimated by those who can
(Morales, 2018). Perhaps the coach most famous for counter-pressing in today’s game is Jurgen
Klopp at Liverpool and Borussia Dortmund. In a 2016 interview, Klopp stated that ‘no
playmaker in the world can be as good as a good counter-pressing situation’ (Nalton, 2016). The
counter-pressing philosophy emerged in response to the extremely successful philosophy of Tiki-
Taka, a tactic popularized by Barcelona (Whitehouse, 2014). Tiki-taka is a possession-based
philosophy, where the team dominate possession of the ball for the majority of the game. In
comparison, counter-pressing is used to combat a possession-based team. The best example of
counter-pressing against Tiki-Taka was the 2013 Champions League semi-final where Barcelona
faced a strong counter-pressing Bayern Munich team. In the semi-final, Bayern set-up to play
reactively and counter-press against Barcelona’s high levels of possession with great success,
winning 7-0 across two legs (Wilson, 2014). With the recent Champions League success for
Liverpool FC and domestic success for Manchester City FC, it is clear that counter-pressing is
revolutionizing the soccer world.
Analytics in soccer are often based upon subjective knowledge of an attribute that is then
quantified into large datasets. This means that measuring attributes such as counter-pressing
efficiency can be difficult. Despite this, counter-pressing efficiency and success has been
measured by large data companies and numerous studies. One of the most notable measures of
counter-pressing is the PPDA (passes allowed per defensive action) model. Using this model,
analysts can measure counter-pressing efficacy by calculating the total amount of opponent
passes successfully completed per defensive action, across 3/5 of the field (Trainor, 2014).
However, a major flaw of PPDA is that for teams who dominate possession in the opponents
final third of the field, they are simply more likely to have a higher PPDA because they occupy
the area more frequently (Harkins, 2016). Another way of measuring counter-pressing is to used
player positional data using a variety of technologies available. In a study by Low, Vilas Boas,
Meyer, Lizaso, Hoitz, Leite, and Gonacalves (2018) on deep-counter-pressing versus high-
counter-pressing, GPS technology was used to plot the position of a press on the field, and the
player’s velocity and position compared to an opponent to verify a press occurred.
AnfieldIndex.com, soccer analysts who research Liverpool FC, also use positional data in their
analysis. In their case counter-pressing is also measured using player positional data, with a press
being verified when the counter-pressing player is within five yards of an opposing player
(Gurpinar-Morgan, 2018). Using positional data of players to calculate counter-pressing
efficiency seems to be the most effective measurement of the counter-pressing philosophy in
soccer today. With limited research on the efficacy of counter-pressing using positional player
data, this project will help fill the gap in research in this field.
The primary purpose of this study was to examine the efficacy of counter-pressing as an
offensive-defensive philosophy. This study aimed to demonstrate how the analytical data of
player positioning can be used to measure the efficacy of counter-pressing in soccer. We have
two primary research questions for this study. Firstly, how effective is counter-pressing in
soccer? Secondly, how can video analysis be used to measure the effectiveness of counter-
pressing in soccer. This study will measure the counter-pressing effectiveness of a women’s
collegiate soccer team across one season. Using video analysis, we will be measuring the
counter-pressing effectiveness of the team during the Fall 2018 season.

2

 The results of this study show that counter-pressing is an effective philosophy when
implemented correctly. Counter-pressing decreased during the 90-minutes of regular season
games. Counter-pressing efficiency, total presses divided by successful presses, was an indicator
of successful game outcome. Presses that led to shots was also an indicator of successful game
outcome. Finally, our video analysis determined that high-central counter-pressing is crucial for
a successful game outcome.

Literature Review

Soccer Analytics
Although soccer is by far the world’s most popular sport, published work in soccer
analytics has not reached the levels of other professional sports (Brooks, Kerr, & Guttag, 2016).
Most published research on sports analytics has focused on sports that have easily categorized
variables such as baseball, American football, and tennis (Brooks et al, 2016). Sports such as
these have match variables that are strongly correlated with scoring outcomes (Brooks et al,
2016). A key issue of building predictive analysis models is the sparseness of positive outcomes
in soccer. Soccer is a low-scoring sport with very few positive outcomes in comparison to
basketball, for example (Brooks et al, 2016). A team may dominate possession or shots and fail
to score. The weak correlation between match variables and goals, means that statistical models
are more difficult to create than in other sports (Brooks et al, 2016).
Despite this, many software systems have been developed for soccer to enable coaches
and trainers to analyze training and matches to improve performance (Stensland et al, 2014). For
example, in 2013 ProZone automated a previously manual task of quantifying player actions
such as speed and position (Stensland et al, 2014). Using motion-detection cameras, the
measurement of these actions was successfully trialed in stadiums such as Old Trafford and the
Reebok Stadium, for Manchester United and Bolton respectively (Salvo et al, 2006). Video
technology can also capture the movement of all players during training and matches. For
example, Carey et al. (2001) analyzed the footedness of 236 male players across 16 teams at the
1998 World Cup Finals, through reviewing video footage. The researchers determined that
decision making did not differ when comparing right and left footed players. Professional soccer
clubs are now hiring video analysts to review previous matches with the purpose of improving
team performance and detecting team deficiencies. Traditional methods of video analysis, such
as the studies mentioned previously, have been extremely labor intensive, often with analysts
reviewing footage by hand to document actions of interest (Carling et al., 2012).
A major challenge for analytics in soccer is the creation of an automated system of data
collection. Recent technological advancements have meant that movements of interest can now
be captured and analyzed with feedback provided to the coach, through video analysis (Carling
et al., 2012). In 2015, Stein et al. (2015) called for an automated system that identifies phases
and patterns of play in soccer. This was in response to the difficulties of using analysis to
correlate statistics with scoring, as all matches and players differ highly. The researchers
proposed an interactive visual analysis system to identify patterns of play the analyst is interested
in. Through user feedback and a learning stage, the analysis process was found to be enhanced
when analyzing real-world soccer matches (Stein et al., 2015). Many studies have focused on
phases of play. For example, a recent study by Lucey at al. (2015) developed a new technique to
measure the factors that increase the probability of a chance in soccer. The researchers analyzed
a whole season of player and ball tracking, specifically focusing on the seconds before a chance,

3

meaning a situation where is an increased likelihood of a goal being scored. They concluded that
classification errors in soccer analysis can be significantly decreased by analyzing variables that
are associated with creating chances.
Alongside video footage, GPS technology has advanced analytics in soccer. GPS devices
are often worn by players during training and matches to track movement (De Silva et al., 2018).
GPS systems collect this data, then transmit it live to coaches and trainers alike. A major reason
for the advancement of analytics through GPS technology is the accessibility of the information
live during practice or games. Following the 2010 World Cup in South Africa, several
professional teams in the country adopted GPS technology analytics to improve performance
(Ferrari, 2017). Numerous teams in the region such as Ajax Cape Town FC have even allowed
ProZone to analyze performance through GPS technology. This form of analysis is
predominantly used to inform coaches on movement patterns (Ferrari, 2018). De Silva et al.
(2018) found that there are significant positional differences in the physical activity levels of
players during competitive matches using GPS technology to track movement. Using wearable
GPS technology on 150 players, movement was tracked across four seasons from 2014-2018
from Chelsea Football Club’s academy. Other studies have focused on how player movement
can indicate match outcome. GPS technology was also used to track 27 players from different
national teams during matches in the 2016 European Championships to find that lower total
distance covered (TD) led to a significantly higher chance of defeat or draw (Smpokos,
Mourikis, & Linardakis, 2018.
In recent history, analytics in soccer have used data points from analytic technology to
create probability models to predict the outcome of matches (Brooks et al., 2016). One of the
first examples of probability modelling, Reep and Benjamin (1968) developed a model that
would measure the success rates of passing sequences of different lengths. However, this study
was hampered due to the lack of video analysis technology. Hughes and Franks (2004) built
upon Reep and Benjamin’s work in their analysis of passing sequences at the 1990 and 1994
FIFA World Cup Finals. The researchers found that more goals were scored from longer passing
sequences than from shorter passing sequences. Similar research from Redwood-Brown (2008)
examined the passing frequencies five minutes prior and post scoring a goal. It was found that on
average, in the five minutes preceding a goal, the scoring team had a significantly high level of
possession than their overall total possession. Whereas, the conceding team had a significantly
lower number of passes prior to conceding. More recent work from Bialkowski et al (2014) used
performance analytics by tracking players’ movements in formations, to identify teams within a
given game, with 70% accuracy. Lamas et al. (2018) also focused on probability modelling in
soccer analytics. The researcher evaluated the decision-making performance of two professional
goalkeepers, Gianluigi Buffon and Julio Cesar, during national team matches. The football field
was divided into six sections, with goalkeeper positioning evaluated based upon the final action’s
(shot/goal/save) location within the six sectors. The two goalkeepers’ decisions were compared
against a dataset from a similar study, with Buffon and Cesar placing in the lowest probability of
conceding a goal in all six sectors (Lamas et al., 2018). The other major area of probability
modelling is injury prediction and prevention. In Hawkins & Fuller’s (1999) study, they defined
the most likely causes of injury to players based upon data from previous Premier League
seasons. Reviewing 1994-1997 seasons, the researchers discovered 67% of injuries occurred
during competition, with 8.5 injuries per 1000 hours of gameplay. This data can be used to
manage injuries and allow coaches to predict injury occurrence across the season.

4

 Creating models to examine performance of soccer players has been the new frontier in
soccer analytics. Statistical models are now used by professional soccer teams in all areas of
coaching, from player selection to areas of focus in training (Brown, 2017). For example, xG
(expected goals) and xGA (expected goals against) have emerged over the last five years as an
important model to measure the amount of goals a team were expected to score and concede in
any given game. OPTA, a large soccer analytics measure xG and xGA on numerous variables
such as assist type, whether they define the chance as a ‘big chance’, and what type of shot it was
(OPTA, 2019). However, some statistics are harder to measure and quantify than others. One of
these measures is counter-pressing effectiveness.

The Phenomenon of Counter-pressing

Counter-pressing in soccer is a philosophy of attack-minded defense. Following the loss
of possession, the defending team will attempt to recover the ball from opponents as high up the
field as possible.
The philosophy of counter-pressing has developed throughout the history of soccer.
Perhaps the first elements of counter-pressing arose during the 1950’s. The Hungarian ‘golden
team’ dominated international soccer throughout the 50’s with their attacking form of defense
(Borbely, Ganczner, Singer, & Hrebik, 2018). Through the use of two formations for attack and
defense, the Hungarians were able to increase movement intensity and press following the loss of
possession. The use of two formations required significant tactical discipline and physical
fitness. This tactic influenced and inspired the historical use of counter-pressing through present
day counter-pressing.
During the 1960’s, Ajax and the Dutch Tulips applied ‘total soccer’ under the guidance
of Rinus Michels. With the notion that attack is the best form of defense, the Dutch teams would
rotate positions to continuously press opponents. While implementing this philosophy, coaching
staff found that a 4-3-3 formation was the most efficient tactical shape to press. The formation
allowed players to ‘opponent hunt’. This term refers to counter-pressing opponents as high up
the field as possible to regain possession. Trigger players were also introduced by these Dutch
teams, where the press would be activated by an attacking player counter-pressing an opponent
with possession (Winner, 2012).
Inspired by these examples of counter-pressing, more recently the counter-pressing
philosophy has become a global phenomenon. Counter-pressing has now become synonymous
with coaches such as Jurgen Klopp (Borussia Dortmund & Liverpool), Pep Guardiola
(Barcelona, Bayern Munich, & Manchester City), and Mauricio Pochettino (Southampton &
Tottenham) (Borbely et al., 2018).
This philosophy cannot be more relevant to soccer today. Coaches such as Guardiola and
Klopp are amongst the most successful coaches in world soccer currently. As counter-pressing
philosophy is so successful currently, this research focuses on how analysis of counter-pressing
in women’s collegiate soccer can be used to evaluate performance.

Counter-pressing and success

Counter-pressing as a philosophy has been linked to increased goals, and therefore
increased success throughout history. In recent history, analytics have been able to prove the link
between counter-pressing and success. In a study by Caley (2014), video analysis was used to
identify the likelihood of a shot based upon pressing locations. Through video analysis, he found
that pressing in a high-central location on the field led to a 20% increase in shots. According to

5

xG models, only 2% of attacking moves lead to shots on goal (OPTA, 2019). Therefore, a 20%
increase in shots through this offensive-defensive strategy can prove highly successful for teams.
An increase of this magnitude in shots, leads to a higher xG, and therefore goals and game
success. This research backs up analytic models that show counter-pressing is linked to success.
In research by Gurpinar-Morgan (2018) using PPDA, the most successful teams at counter-
pressing in the 2017/18 English Premier League season were Manchester City, Tottenham
Hotspur, and Liverpool by a significant margin. All three teams had particularly successful
seasons, with Manchester City breaking the premier league points record and Liverpool reaching
the final of the Champions League. Skeptics of this link between counter-pressing and success
would argue that all three of the previously mentioned teams have very talented squads and
reached these successes simply because they were better than the other teams. It would be argued
that teams with weaker players such as Burnley could not possibly achieve success by counter-
pressing. However, Burnley are also near the top of the PPDA rankings, so did they achieve
success using counter-pressing? In the 2017/18 Premier League season, Burnley finished 7th and
qualified for the Europa League in a highly successful season for the club. In fact, according to
Gurpinar-Morgan (2018), Burnley were second in the total counter-press possessions per game
on 18, only behind Manchester City.
This model illustrates the link between counter-pressing and success. However, PPDA
does have its limitations. For example, teams who possess the ball in higher areas of the field
more frequently, have much higher PPDA than other teams simply because that is where they
lose possession most often. Therefore, a statistical model that measures pressing efficacy is
required. Anfield Index, a team of data analysts who often focus on counter-pressing, have
designed an effective model of measuring pressure in male professional soccer. Through video
analysis, the team plot counter-pressing actions and locations on a template to collect data. A
press is verified by tracking that the pressing player is within a five-yard radius of the opponent
in possession. This method of data collection, although time consuming, is extremely effective in
measuring counter-pressing effectiveness. The limitation of this model is that the team
predominantly focus on the counter-pressing of Liverpool.
This study will use similar video analysis to examine the efficacy of counter-pressing in
soccer. By sampling a collegiate soccer team in the US, this study will explore whether counter-
pressing will lead to success at a lower level of soccer than male professional soccer. We will
also determine whether video analysis can be used as an effective measure of counter-pressing.

Methodology

This research study used quantitative analysis as a framework for data collection. The
purpose of this study is to examine and quantify the efficacy of counter-pressing as a philosophy
in soccer. To complete this purpose, we specifically used quantitative content analysis
categorize, code, and interpret counter-pressing efficacy, using existing footage of the Miami
University’s women’s soccer team during the 2018 Fall Mid-American Conference (MAC)
season.
A convenience sampling was used for this study. This study was written at Miami
University therefore it was convenient and accessible for the research to focus on the Miami
University women’s soccer team. Contact with the coaching staff at Miami University was
initiated via email, with a meeting to discuss the study with the coaching staff shortly afterwards.
Following the meeting, footage of all games involving Miami University women’s soccer team

6

across the MAC season were shared by the coaching staff using the secure server VidSwap.com.
All players for the women’s team that played any number of minutes during the Fall 2018 MAC
season were analyzed as part of the study’s sample.
To protect player identities, all players were kept anonymous during this process, with
players being assigned a number code that only the researcher and Head Coach were aware of.
This number coding will be encrypted and stored securely on the researcher’s computer. The
data created from this study is protected on a secured Google spreadsheet, with only the
researcher and the coaching staff having access. No players were contacted during this study.
From the footage, counter-pressing actions will be counted and categorized as either
successful or unsuccessful. Successful counter-pressing will include the following;

1) Press that led to shot - The defending team wins the ball through counter-pressing and
has a shot within 10 seconds of the turnover.

2) Press that led to goal - The defending team wins the ball through counter-pressing and
scores within 10 seconds of the turnover.

3) Press won possession – The defending team win possession of the ball through
pressuring the opponent. Possession is gained through directly turning over the ball on
the field, or pressuring an opponent to concede a set-piece (throw-in, free-kick, corner).

4) Press pushed opposition back - The defending teams’ pressure forces the opponent to
pass or dribble the ball sideways or backwards.

An unsuccessful press will include the following;

1) Failed press - The offensive team player either dribbles or passes the ball in a forward
direction beyond the defensive counter-pressing player.

2) Failed press led to opposition free kick - The counter-pressing player commits a foul
during a press.

These individual counter-pressing actions will then be plotted onto a virtual soccer map
manually using Microsoft PowerPoint and video footage. Counter-pressing actions will be
plotted according to the successful and unsuccessful counter-pressing symbols (Fig. 1). All
counter-pressing actions of individual athletes will be collated onto one virtual soccer map using
a template that has been created on Microsoft PowerPoint (Fig. 2). The counter-pressing actions
of all individual members of the soccer team will also be displayed on one virtual soccer map, to
display the team successful and unsuccessful counter-pressing actions during one game. Fig. 3
illustrates an example individual counter-pressing map, using Liverpool FC central-midfielder
Georginio Wijnaldum’s counter-pressing actions against Tottenham Hotspurs (A) in the 18/19
Premier League season. Once individual counter-pressing actions have been mapped, they can be
collated onto one team counter-pressing map. Fig. 4 demonstrates a team counter-pressing map
for Liverpool FC against Tottenham Hotspurs (A) from the 18/19 Premier League season.
Video analysis for each season game took approximately three hours, totaling 33 hours
across 11 MAC games. Two computer screens were used during the video analysis. One screen

7

displayed the video footage provided by the coaches. The other screen was used to plot the
location of presses on the Microsoft PowerPoint template, as well as to input the category
(successful or unsuccessful) and time the press occurred into a Google spreadsheet. Once all
individual player press locations had been input on the PowerPoint template, presses for that
particular game were collated onto one template to show the team’s counter-pressing locations
overall. To determine the zones presses had occurred in, the field was divided into high, mid, and
low (Fig. 5) counter-pressing, and right, central, and left counter-pressing (Fig. 6). Opponents
counter-pressing data was also collected as a tool for comparison against Miami University’s
counter-pressing efficacy. This data was not mapped for counter-pressing location due to time
constraints, with presses only categorized into the Google spreadsheet to calculate effectiveness.
Following the completion of video analysis, data collated was shared back to the coaching staff
of Miami University women’s soccer team.
Due to time constraints our convenience sample only consisted of the Miami University
women’s soccer team. Therefore, the generalizability of our findings is limited. By collating
counter-pressing data from opponents, the efficacy of the opponent’s counter-pressing against
Miami University can be discovered. However, the sample of data is inadequate to generalize to
all women’s collegiate soccer within the United States. There are numerous women’s collegiate
soccer conferences within the US and of differing competitiveness compared to the MAC. This
means that findings should not be used to infer the efficacy of counter-pressing as a philosophy
within women’s collegiate soccer as a whole. Despite this, the study provides a template to
measure the efficacy of counter-pressing for a wider study of all women’s collegiate soccer in
the US.

Results

Introduction
The aim of this research was to examine the efficacy of counter-pressing as an offensive-
defensive philosophy. The research also wanted to explore how video analysis of positional data
can be use d to determine counter-efficacy. The following findings will explore these two
research questions.

Finding 1 – Counter-pressing is an indicator of success

The analytics performed inferred that counter-pressing that led to shots was an indicator
of success. The counter-pressing totals for Miami University and each opponent are displayed in
Table 1. Total counter-pressing actions did not consistently reflect the game outcome, win, loss,
or draw. For example, Table 1 shows higher total counter-pressing actions than their opponents
across their first three MAC games with a loss outcome. However, the following three games
Miami University had fewer total counter-pressing actions than their opponents with the
outcomes of a win, draw (overtime loss), and loss.
Counter-pressing efficiency is a more accurate indicator of how counter-pressing
outcomes influence game success. The counter-pressing efficiency of Miami University and their
opponents is indicated in Table 2. Throughout the MAC season, the team who had a higher
counter-pressing efficiency had the successful game outcome. This in itself is unsurprising, as
common sense would indicate that the stronger team presses more efficiently. What confirms
that counter-pressing more efficiently leads to success, is the large differences in counter-
pressing efficiency the greater the score differential in the game. In Miami University’s greatest

8

score differential in their favor, their efficiency was 71.4% to the opponents 60.19%. On the
other side, in Miami University’s largest score differential against, their efficiency was 54.57%
to their opponent’s 80.18%. Therefore, perhaps the larger the difference in counter-pressing
efficiency between Miami University and its opponents, the greater the level of game success
positively or negatively.
The most interesting indicator that counter-pressing leads to success is the category
‘Press led to shot’. As the literature reflects, counter-pressing as a philosophy is a team tactic
where players pressure opponents who have possession as high as possible up the field to win
possession. Through winning possession high up the field, the team is closer to the opponent’s
goal for their attacks. Therefore, it would be hypothesized that counter-pressing that led to shots
would indicate the game success of that team. The analysis of the MAC season for Miami vs
opponents confirms this hypothesis in Table 3. Similarly, to counter-pressing efficiency, the
team who had the larger total of presses leading to shots had the positive game outcome. In the
greatest scoreline differentials, presses that led to shots peaked. For example, the highest presses
to shots for opponents was eight, in a 1-4 win for Kent State against Miami University. For
Miami University, they peaked at six presses to shots in a 4-0 win against Northern Illinois. For
the 2018 MAC season, Miami University’s counter-pressing zones were analyzed to locate
where counter-pressing occurred most during their season games. As mentioned previously in
the methodology section, the game field was divided into three zones; High counter-pressing,
mid counter-pressing, and low counter-pressing. The counter-pressing location season averages
are displayed in Table 4. With 38.8% of presses in the high zone, and 54.8% of presses in the
mid zone, it can be hypothesized that Miami University’s lower totals of presses leading to shots
occurred predominantly due to the location of their counter-pressing during games. The
philosophy of counter-pressing would infer that the majority of counter-pressing should occur in
the high press zone, but Miami University’s majority occurred in the mid press zone.
Following the feedback of the analysis, a tactical change to increase counter-pressing in
the high zone could mean more presses leading to shots, and thus a greater likelihood of game
success according to our findings. However, there are perceived limitations to deploying high
counter-pressing. A coach of a technically weaker side than their opponents may argue that
counter-pressing higher up the field may create more space behind the pressing players for
opponents to take advantage of. This may seem the case from a coaching perspective, but a
counter-pressing philosophy requires all players to hold a higher line up the field including the
defense and goalkeeper. With defenders higher up the field behind the pressing midfielders and
forwards, this perceived space for opponents to exploit does not exist. Therefore, a higher
counter-press is recommended for game success.

Finding 2 – Discipline is crucial

According to the literature, teams who use the counter-pressing philosophy must be
physically fit and disciplined. Therefore, all players in that team must press as one working unit
for the press to work. Counter-pressing should occur following a ‘trigger player’, typically the
forward or central attacking midfielder, with other players counter-pressing based upon this.
During the analysis it was found that counter-pressing was not equally distributed.
During analysis, counter-pressing zones were created to map where counter-pressing occurred.
During the MAC season, 39.7% of Miami University’s presses occurred within the central zone
of the field (Table 4). Counter-pressing on the left side of the field occurred 27.7% of the total
presses, with 32.6% occurring on the right side. This may seem like a fairly even distribution of

9

counter-pressing, however this is dependent on the formation. With Miami University most
frequently deploying a 4-2-3-1 formation (Fig 5), only two players press the left zone (left
midfielder [LM] and left back [LB]) and two in the right zone (right midfielder [RM] and right
back [RB]). With this formation, four players press the central zone (forward [FWD], central
attacking-midfielder [CAM], and two central defensive-midfielders [CDM]). Therefore, it would
be expected that counter-pressing would be far more frequent in the central zone than the wide
zones, but this was not the case. In fact, Miami University’s worst game outcome occurred in a
1-4 loss to Kent State, where just 4% of presses occurred in the high-central zone of the field. In
comparison, Miami University’s best game outcome was a 4-0 win against Northern Illinois,
where 17% of presses occurred in the high-central zone. The counter-pressing heatmaps for the
Kent State (Fig 6) and Northern Illinois (Fig 7) games clearly demonstrate this difference. As
mentioned previously, the high-central zone is the location most presses would most commonly
occur based upon the high-press mentality and overload of players centrally.
Perhaps the best explanation for this lack of high central counter-pressing is a noticeable
gap between the FWD and CDM’s in their counter-pressing positioning. As mentioned
previously, once the trigger player as activated the press, all members of the defending team
must follow the press until they regain possession, push the opposition back, or press the ball out
of play. If the press is not activated following a trigger, the press has a very high likelihood of
becoming unsuccessful due to the overload of players the team with possession has. However, if
all defending players from the FWD to CDM’s press, the defending team will have a counter-
pressing overload and are far likelier to regain possession through a successful press. Throughout
the analysis, it was clear that following a trigger, not all players between the FWD and CDM’s
activated a press. An example of this gap is shown in Fig 8, with Miami University in red. In this
example, the RM has triggered the press and team counter-pressing should now be activated.
However, the CAM, two CDM’s, and LM have not activated. Between the trigger player and two
CDM’s, there are three opponent players and two defending pressers. With an overload for the
team in possession, they are likely to break the press and attack the defending team. In this
scenario, the two CDM’s would push higher towards the halfway line, the LM would be counter-
pressing the opponents RM, and the RM would be counter-pressing the opponents LM. This
press would have a much greater likelihood of success due to the counter-pressing overload of
seven players to the opponents four.
Following a meeting with the coaching staff of Miami University at the start of analysis,
it was hypothesized that after the first two games of the Fall MAC 2018 season, Miami
University had implemented a ‘line of counter-pressing’ that would occur at the halfway line.
This would mean that no counter-pressing trigger should occur until the opponents attempt to
pass the line of press with possession of the ball. However, through video analysis it was clear to
see that this counter-pressing tactic was not working for Miami University. As Fig 8 shows, the
attacking players (FWD, CAM, LM, & RM), often triggered the press beyond the halfway line,
thus creating this large gap between Miami University’s CDM’s who held the line of press, and
Miami University’s attacking players who pressed beyond it.
We recommend that coaches should decrease this FWD to CDM gap, through the CDM’s
counter-pressing higher after the trigger. This would mean that the ratio of presses in the high-
central zone will increase and led to a higher likelihood of a successful outcome following the
press - a goal or shot. However, following a meeting with the coaching staff of Miami
University, it was revealed that Miami University’s ability to implement an effective counter-
pressing philosophy is hampered due to various constraints. Due to NCAA rules on women’s

10

collegiate soccer, coaches are only given one to two weeks of preseason with their players before
their first season game. Counter-pressing requires high discipline as an entire team to
implement. One to two weeks of pre-season is not enough time to integrate this philosophy into
the team with existing players graduating and new arrivals potentially not having knowledge of
counter-pressing.

Finding 3 - Counter-pressing across 90 minutes decreased

As the literature suggests, implementing a counter-pressing philosophy requires frequent
high intensity sprints by the defending team. Due to this fact, it is expected that counter-pressing
may decrease during the 90 minutes of a regular season game if fitness levels are not at the level
required. Through video analysis, it was found that Miami University’s counter-pressing actions
did decrease on average during 90-minute season games. On average, Miami University pressed
18 times in the first 15 minutes of a game, and steadily declined to 10 presses in the last 15
minutes of the game (Table 5). Possession-adjusted counter-pressing per minute (PAPPM) also
decreased across the 90 minutes, from 2.24 PAPPM in the first 15 minutes to 1.24 PAPPM in the
final 15 minutes. Counter-pressing efficiency did decrease from 70.29% to 66.20% during
season games but varied throughout the 90 minutes rather than a steady decline. Following a
meeting with the coaching staff at Miami University, it was concluded that low fitness levels are
the clear explanation for the decreasing counter-pressing statistics during the 90 minutes.
Although a steady decline in counter-pressing was found across the 90 minutes, it would
be expected that counter-pressing would increase following the halftime break. With players
resting for 15 minutes during halftime, counter-pressing should increase beyond the last 15-
minute period of the first half. However, Table 5 shows that this was not the case. On average,
counter-pressing decreased from 14 before halftime and 13 after halftime. PAPPM also
decreased after halftime, with efficiency staying in the mid-60’s%. This slight decrease in
counter-pressing, rather than an increase, could be down to numerous reasons. However, the
most likely causes are either a drop in the player’s concentration during the halftime interval, or
a lack of motivation during the halftime interval team talk.
Despite counter-pressing actions decreasing throughout the season, the final three MAC
season games saw a large increase in counter-pressing. During the final three games of the
season Miami University averaged 91 presses per game, whereas the three games previous
averaged 72 presses per game (Table 6). Following a meeting with Miami University’s coach,
this increase in counter-pressing was discussed. It was hypothesized by the coaching staff that
the player’s mindset had changed before these final three games. The first increase in counter-
pressing occurred during MAC game nine, against Eastern Michigan. At this point in the MAC
season Miami University were near the bottom of the MAC, with very little chance of pushing
higher up the league table. Whereas during the previous eight season games Miami University’s
counter-pressing was more conservative to protect their defense, the players mindset changed
during the final three season games with increased total counter-pressing. A comparison of high-
central counter-pressing percentages across the MAC season debunks this hypothesis. Following
game five, counter-pressing in the high-central zone increased and averaged 16%, whereas
counter-pressing during the first five games occurred 10% of the time (Table 7). A change in
mindset may have potentially occurred following the first five games, however. With Miami
University 0-5 in the MAC, players may have been willing to risk counter-pressing higher to
improve their position in the MAC with seven games left to play.

11

 We recommend that high counter-pressing intensity levels be maintained for counter-
pressing to be effective. However, counter-pressing will require high levels of physical fitness. In
Miami University’s case, fitness physical of the players was a major constraint of implementing
a higher and more effective counter-pressing strategy. On a basic level, we recommend
increasing the fitness levels of all players. On a more advanced level, collegiate soccer provides
coaches with a unique opportunity at a high soccer level to maintain a high-intensity press during
an entire 90-minute game. The NCAA rules allowing substitutes who have left the field during
the first-half, to re-enter the field in the second-half, allows coaches to rotate counter-pressing
players throughout both halves to maintain counter-pressing levels. Not only would this increase
pressing intensity, this would also allow coaches of a team of low physical fitness to maintain
pressing by switching out these players before counter-pressing decreases. This recommendation
would require a large squad size to implement, but may be a way to maintain high counter-
pressing intensity

Conclusion of findings
Throughout the literature, it is suggested that counter-pressing is an indicator of success.
The findings of this research demonstrate that increased counter-pressing efficiency and
increased presses leading to shots led to a higher likelihood of a win for that team. However,
disciplined and cohesive counter-pressing is required for this success to occur. Alongside this,
high levels of player fitness are needed to maintain the counter-pressing intensity for a successful
outcome. The analysis of this research project found that Miami University’s counter-pressing
decreased during each 90-minute game. Miami University’s counter-pressing efficiency was also
lower than the majority of opponents, with Miami University’s most successful game outcome
occurring when they pressed more efficiently and centrally. Across the MAC season, counter-
pressing data for both Miami University and the opponent has indicated the likely outcome of the
game. The more successful counter-pressing team on each occasion had a successful game
outcome. It can therefore be determined that the counter-pressing philosophy is effective in
women’s collegiate soccer.

Discussion

Introduction
The objective of this paper was to examine the efficacy of counter-pressing as an
offensive-defensive philosophy. We also aimed to demonstrate how video analysis of player
positioning can be used to measure the efficacy of counter-pressing.

Summary of findings
The results of this study have suggested that counter-pressing statistics are an indicator of
game success. Total counter-pressing actions did not consistently correlate to game success
during Miami University’s MAC season. More efficient counter-pressing was linked to game
success, with the winning team in each game counter-pressing more efficiently than their
opponent. The most prominent indicator of success was counter-pressing that led to shots. A
larger amount of presses leading to shots seemed to indicate a larger score differential between
the winning and losing side. In each MAC season game, the team who pressed more frequently
to shots won the game.

12

 Team cohesion during the press was examined and found to be crucial for counter-
pressing success. Following the counter-pressing trigger, Miami University’s press often did not
activate, leading to a high likelihood of an unsuccessful press. The lack of activation also led to
the FWD to CDM gap occurring, which gave space for the opponent to exploit and overload.
When this gap was overloaded by Miami University’s counter-pressing players, it increased the
likelihood of a successful press.
Finally, counter-pressing during each 90-minute game decreased throughout the season.
Despite this decrease, it would be expected that counter-pressing would increase after the
halftime interval as players have a rest period. This was not the case, suggesting a lack of
concentration or motivation following the break. Further study is needed on this to examine the
cause of this counter-pressing decrease. Coaching staff at Miami University hypothesized that
decreasing counter-pressing levels were due to low player fitness levels, following a small
preseason period for conditioning.

Implications of findings
The data collected in during this study has numerous implications for practice for coaches
and analysts.
Firstly, this study has proved that counter-pressing decreases during a 90-minute games
in women’s collegiate soccer. Although this is a rudimentary finding, it is an important one for
practice and governance of women’s collegiate soccer. Currently, coaches of women’s collegiate
soccer are hampered by a short preseason and extremely busy game schedule. Adjustments to
philosophies such as counter-pressing are made difficult in this environment, where coaches
often have two games per week in the Fall and a very short preseason. There is a current push to
implement a two-semester season in NCAA men’s soccer, with games during the Fall and Spring
semester and often one game per week. If this system were to be implemented, it would allow
coaches more time during the season to make adjustments to training practice based upon
analytics, thus potentially improving performance. A one game per week system would also be
beneficial to the fitness levels of players, decreasing their chronic workloads. As ever though, it
seems that this rules change would be implemented into men’s collegiate soccer before women’s
collegiate soccer.
Secondly, an intense high central press is more likely to led to game success. To
implement this game strategy, counter-pressing levels must be maintained throughout the entire
90 minutes. With the current NCAA rules on substitutions and squad sizes, this strategy is
possible. Large player rotations every 30 minutes of the game would maintain the high intensity
press during the game. Current rules allow unlimited substitutions during each half, as long as
players do not reenter the field during a half that they have been previously substituted in.
Therefore, with a large gameday squad who have been conditioned for counter-pressing, this
strategy is possible. If women’s collegiate soccer maintains a one semester season with often two
games a week, then this strategy could help coaches main a high-intensity counter-pressing
philosophy.
Finally, it is clear that more funding is needed in women’s collegiate soccer to continue
the analytic revolution in the sport. The findings gained from analytics in women’s soccer can be
implemented in practices and gameday practices across the season, potentially giving the team a
competitive edge over their opponents. However, not all women’s collegiate soccer teams have
access to the level of funding required to implement analytics. In Miami University’s case, they
plan to use GPS technology for data points in their future analysis of game and practice

13

performance. This technology is costly and does not provide context to game situations without
video analysis. Therefore, teams using this technology effectively will not only purchase
expensive GPS technology, but will require an analyst to study game footage.

Limitations
The limitations of this study require further research. Firstly, analysis of counter-pressing
should ideally be completed over consecutive MAC seasons for comparisons. Due to time
constraints, only one MAC season for Miami University could be analyzed. A multi-season
comparison will give Miami University’s coaching staff evidence that coaching practices
implemented following analysis led to successful outcomes for the team. This is important to
give coaches the knowledge that analysis is an effective tool in women’s collegiate soccer to
improve training practice and game performance. Secondly, all MAC season games should be
analyzed to determine the most effective counter-pressing tactics within the conference. This
analysis would highlight vulnerabilities of teams against certain types of counter-pressing
intensities. For example, a weaker opponent may be more vulnerable to a high central press,
whereas a stronger opponent may be susceptible to a limited mid press. Future work here will
allow coaches to use analysis to determine game tactics for each opponent they face. Finally, the
sole use of video analysis to interpret counter-pressing means that the definition of what is and
isn’t a press within a game will be subjective to the researcher. For example, what may be
defined as a successful press by one researcher may be defined as unsuccessful or not a counter-
pressing action. However, due to the researcher’s experience within soccer, mistakes are
unlikely.

Pointers for further study

While this study examines how counter-pressing is an effective philosophy using video
analysis of one women’s collegiate soccer, a larger-scale and multi-season study will be required
to explore the effectiveness of this philosophy.
Furthermore, more studies using video analysis could be helpful to create an accurate
measure of advanced statistical models such as xG (expected goals) and the xG chain in soccer.
Analytical research on soccer using GPS analysis alongside video analysis is also needed. GPS
analysis is effective in accurately locating where game events occurred, such as a press. This
form of analysis will also provide researchers the knowledge that a high intensity run occurred,
which is an important aspect of counter-pressing. Using both technologies will allow researchers
to create accurate measures in soccer which provide context to what happened within the game
for that data point to occur.
Finally, studies which explore counter-pressing efficacy with teams who implement
recommendations from this study are required. For example, a study which measures the
intensity levels of counter-pressing from a team who regularly substitute pressing players, would
provide coaches of collegiate soccer important feedback on how to implement pressing more
effectively.

14

 References

Aquino, R., Viera, LH., Carling, C., Martins, G., Alves, I., & Puggina, E. (2017). Effects of
competitive standard, team formation, and playing position on match running
performance of Brazilian professional soccer players. International Journal of
Performance Analysis in Sport, 17(5), 695-705
Arnason, A., Sigurdsson, S., Gudmundsson, A., Holme, I., Engebretsen, L, & Bahr, R. (2004).
Physical fitness, injuries, and team performance in soccer. Medicine and Science in
Sports and Exercise, 36, 278-285.
Barnes, C., Archer, D.T., Hogg, B., Bush, M., & Bradley, P.S. (2014). The evolution of physical
and technical performance parameters in the English Premier League. International
Journal of Sports Medicine, 35, 1095-1100.
Barros, R., Misuta, M., Menezes, R., Figueroa, P., Moura, F., Cunha, S., Anido, R., & Leite, N.
(2007). Analysis of the distances covered by first division Brazilian soccer players
obtained with an automatic tracking method. Journal of Sports Science, 6(2), 233-242.
Bialkowski, A., Lucey, P., Carr, P., Yue, Y., Sridharan, S., & Matthews, I. (2014). Identifying
team style in soccer using formations learned from spatiotemporal tracking data. Data
Mining Workshop, 1, 9-14.
Borbely, L., Ganczner, P., Singer, A., & Hrebik, J. (2018). All about pressing in soccer.
Maidenhead: Meyer & Meyer Sport UK
Bradley, P.S., Sheldon, W., Wooster, B., Olsen, P., Boanas, P., & Krustrup, P. (2009). High-
intensity running in English FA Premier League soccer matches. Journal of Sports
Science, 27, 159-168.
Brooks, J., Kerr, M., & Guttag, J. (2016). Using machine learning to draw inferences from pass
location data in soccer. Statistical Analysis and Data Mining, 9(5), 338-349.
Brown, L. (2017, February 17th). How video analysis helped take Lincoln City to the top of the
league and in reaching distance of FA Cup history. Retrieved from
https://www.independent.co.uk/sport/football/fa-league-cups/how-lincoln-city-s-
revolutionary-approach-to-video-analysis-took-them-to-the-top-of-the-league-and-
a7586026.html
Bush, M., Barnes, C., Archer, D.T., Hogg, B., & Bradley, P.S. (2015). Evolution of match
performance parameters for various playing positions in the English Premier League.
Human Movement Science, 39, 1-11.
Carey, D., Smith, G., Smith, D., Shepherd, J., Skriver, J., Ord, L., & Rutland, A. (2001).
Footedness in world soccer: an analysis of France ’98. Journal of Sports Sciences,
19(11), 855-864.
Carling, C., Bloomfield, J., Nelsen, L., & Reilly, T. (2012). The role of motion analysis in elite
soccer. Sports Medicine, 38(10), 839-862.
De Silva, V., Caine, M., Skinner, J., Dogan, S., Kondoz, A., Peter, T., Axtell, E., Birnie, M., &
Smith, B. (2018). Player tracking data analytics as a tool for physical performance
management in football: A case study from Chelsea Football Club academy. Sports, 6(4),
130.
Ferrari, S. (2017). Performance analysis in soccer. Potentialities and challenges in the African
context. Journal of Physical Education and Sport, 17(1), 436-440.
Goff, S. (2018, March 1st). The beautiful game discovers that algorithms can be beautiful too.

15

 Retrieved from https://www.washingtonpost.com/news/soccer-
insider/wp/2018/03/01/the-beautiful-game-discovers-that-algorithms-can-be-beautiful-
too/?utm_term=.c3cd549c21dc
Gómez, M.A., Ortega, E, & Sampario, J. (2013). Situational variables. In McGarry, T. &
O’Donoghue, P. (Eds.) Routledge handbook of sports performance analysis (pp 259-
269). London: Routledge.
Gürpınar-Morgan, W. (2018, May 21st). How StatsBomb data helps measure counter-pressing.
Retrieved from https://statsbomb.com/2018/05/how-statsbomb-data-helps-
measure- counter-pressing/
Halson, S.L. (2014). Monitoring training load to understand fatigue in athletes. Sports Medicine,
44(2), 139-147.
Harkins, J. (2016, October, 11th). How we measure pressure. Retrieved from
https://www.optasportspro.com/news-analysis/blog-how-we-measure-pressure/
Hawkins, R., & Fuller, C. (1999). A prospective epidemiological study of injuries in four English
professional football clubs. British Journal of Sports Medicine, 33(3), 196-203.
Hewitt, A., Norton, K., & Lyons, K. (2014). Movement profiles of elite women soccer players
during international matches and the effect of opposition’s team ranking. Journal of
Sports Science, 32(20), 1874-1880.
Hughes, M., & Franks, I. (2007). Analysis of passing sequences, shots, and goals in soccer.
Journal of Sports Sciences, 23(5), 509-514.
Jang, J., Ankit, A., Kim, J, Jang, Y., Kim, H., Kim, J., & Xiong, S. (2018). A unified deep-
learning model for classifying the cross-country skiing techniques using wearable
gyroscope sensors. Sensors, 18(11), 3819-3844.
Kidd, R. (2018, November 19th). Soccer’s moneyball moment. Retrieved from
https://www.forbes.com/sites/robertkidd/2018/11/19/soccers-moneyball-moment-how-
enhanced-analytics-are-changing-the-game/#67117e4676b2
Kleinschmidt, J. (2019, July 25). Billy Beane describes how A’s ‘Moneyball’ revolution sped up
MLB analytics revolution. Retrieved from
https://www.nbcsports.com/bayarea/athletics/billy-beane-explains-how-moneyball-sped-
mlb-analytics-revolution
Lamas, L., Drezner, R., Otranto, G., & Barrera, J. (2018). Analytic method for evaluating
players’ decisions in team sports: applications to the soccer goalkeeper. PLoS ONE
13(2), 1-18
Larsson, P. (2003). Global positioning system and sport-specific testing. Sports Medicine, 33,
1093-1101.
Low, B., Vilas Boas, G., Meyer, L., Lizaso, E., Hoitz, F., Leite, N., & Gonçalves, B. (2018).
Exploring the effects of deep-defending vs high-press on footballers’ tactical behavior,
physical and physiological performance. SciElo Analytics, 24(2), 1-5.
Lozada, C. (2016, March 18). How data geeks are taking over basketball. Retrieved from
https://www.washingtonpost.com/news/book-party/wp/2016/03/18/how-data-geeks-are-
taking-over-basketball/?utm_term=.e3e24c1d0f9b

16

Marr, B. (2018, May 21st). How much data do we create every day? Retrieved from
https://www.forbes.com/sites/bernardmarr/2018/05/21/how-much-data-do-we-create-
every-day-the-mind-blowing-stats-everyone-should-read/#9f5ebdb60ba9
McKenzie, R. & Cushion, C. 2011. Performance analysis in football: A critical review and
implications for future research, Journal of Sport Sciences. 31(7): 639-676.
Memmert, D., Limmerk, K., & Sampario, J. (2017) Current approaches to tactical performance
analyses in soccer using positional data. Sports Medicine, 47(1), 1-10.
Mills, T. (2019, January 7th). Why big data and machine learning are important in our society.
Retrieved from https://www.forbes.com/sites/forbestechcouncil/2019/01/07/why-big-
data-and-machine-learning-are-important-in-our-society/#2b15b6f27aa2
Morales, N. (2018, June 27th). The offensive-defensive tactic that has defined the World Cup.
Retrieved from https://www.theringer.com/soccer/2018/6/27/17509246/2018-world-cup-
tactics-pressing-germany-senegal-argentina
Nalton, J. (2016, December 23rd). Jurgen Klopp’s football philosophy explained in 10 key
quotes. Retrieved from https://www.thisisanfield.com/2016/12/jurgen-klopp-quotes-
football-philosophy/
OPTA. (2019). The next frontier of sports data, developed by OPTA. Retrieved from
https://www.optasports.com/services/analytics/
Redwood-Brown, A. (2008). Passing patterns before and after goal scoring in FA premier league
soccer, International Journal of Performance Analytics, 8(3), 172 – 182.
Reep, C., & Benjamin, B. (1968). Skill and chance in association football. Journal of the Royal
Statistical Society, 131(4), 581-585.
Ricky, A. (2019, January 31st). How data analysis in sports is changing the game. Retrieved
from https://www.forbes.com/sites/forbestechcouncil/2019/01/31/how-data-analysis-in-
sports-is-changing-the-game/#1f3844b43f7b
Salvo, VD., Collins, A., McNeill, B., & Cardinale, M. (2006). Validation of Prozone: A new
video-based performance analysis system. International Journal of Performance
Analytics, 6(1),0 108–119.
Schoenfeld, B. (2019, May 22nd). How data brought Liverpool to the cusp of glory. Retrieved
from https://www.nytimes.com/2019/05/22/magazine/soccer-data-liverpool.html
Smpokos, E., Mourikis, C., & Linardakis, M. (2018). Seasonal physical performance of a
professional team’s football players in a national league and European matches. Journal
of Human Sport and Exercise, 13(4), 720-730.
Stein, M., Häubler, J., Jäckle, D., Janetzko, H., Schreck, T., & Keim, D. (2015). Visual soccer
analytics: Understanding the characteristics of collective team movement based on
feature-driven analysis and abstraction. International Journal of Geo-Information, 4(4),
2159-2184.
Timms, A. (2018, August 16th). Are super-nerds really ruining US sports? Retrieved from
https://www.theguardian.com/sport/2018/aug/16/sports-nerds-analytics-data
Thornton, JS., Frémont, P., Khan, K., Poirier, P., Fowles, J., Wells, GD., & Frankovich, RJ.
(2016). Physical activity prescription: a critical opportunity to address a modifiable risk
factor for the prevention and management of chronic disease. Clinical Journal of Sport
Medicine, 26, 259-265.
Trainor, C. (2014, July 30th). Defensive metrics: Measuring the intensity of a high press
. Retrieved from https://statsbomb.com/2014/07/defensive-metrics-measuring-the-
intensity-of-a-high-press/

17

Tursi, D., Napolitano, S., Polidoro, L., & Raiola, G. (2013). Video analysis as an instrument in
juvenile soccer training. Journal of Human Sport and Exercise, 8, 688-693.
Vescovi, JD., & Favero, TG. (2014). Motion characteristics of women’s college soccer matches:
female athletes in motion (FAiM) study. International Journal of Sports Physiology and
Performance, 9(3), 405-414
Vescovi, JD. (2012). Sprint profile of professional female soccer players during competitive
matches: Female athletes in motion (FAiM) study. Journal of Sports Science, 30(12),
1259-1265.
Whitehouse, M. (2014). The blueprint for soccer’s new era. Oakamoor, UK: Bennion Kearny
Ltd.
Wilson, J. (2014, May 1st). Is this the end for tiki-taka? Retrieved from
https://www.theguardian.com/football/blog/2014/may/01/the-question-is-this-
the-end-for-tiki-taka-football
Winner, D. (2012). Brilliant orange: The neurotic genius of Dutch football. London:
Bloomsbury Publishing

18

 Figures

Figure 1: Counter-pressing Action
Symbols












19



Figure 2: Individual/Team Counter-pressing
Template

20



Figure 3: Example Individual Counter-

pressing Map

















21



Figure 4: Example Team Counter-pressing Map


















22










High

Mid

Low

Figure 5: High, mid, and low counter-pressing zones


















23



Right

Central

Left





Figure 6: Right, central and left counter-pressing
zones






















24


Figure 7: 4-2-3-1 formation example





















25




























Figure 8: Miami Counter-pressing heatmap vs
Kent State

















26

























Figure 9: Miami Counter-pressing heatmap vs Northern
Illinois




















27



Figure 10: Example of the FWD to CDM
gap





















28

 Tables

Table 1: Total Counter-pressing Actions
Game Total Counter-pressing Total Counter-pressing
Actions Miami University Actions Opponent
vs Bowling Green 125 96
vs Toledo 101 106
vs Ohio 101 61
vs Kent State 47 113
vs Western Michigan 70 97
vs Northern Illinois 82 103
vs Buffalo 67 102
vs Akron 69 91
vs East Michigan 97 111
vs Central Michigan 88 108
vs Ball State 89 99

29

Table 2: Counter-pressing Efficiency
Game Miami University Efficiency Opponent Efficiency
vs Bowling Green 57.46% 77.08%
vs Toledo 66.22% 73.58%
vs Ohio 59.47% 70.49%
vs Kent State 70.21% 80.53%
vs Western Michigan 65.45% 74.23%
vs Northern Illinois 71.40% 60.19%
vs Buffalo 72.29% 79.41%
vs Akron 72.46% 75.82%
vs East Michigan 54.57% 80.18%
vs Central Michigan 66.92% 75.93%
vs Ball State 71.91% 75.76%

30

Table 3: Counter-pressing that led to shots
Game Miami University Press Lead to Opponent Press Lead to
Shot Shot
vs Bowling Green 0 5
vs Toledo 2 3
vs Ohio 2 3
vs Kent State 0 8
vs Western 4 5
Michigan
vs Northern Illinois 6 1
vs Buffalo 3 4
vs Akron 2 3
vs East Michigan 4 7
vs Central Michigan 0 5
vs Ball State 3 3

31



Table 4: Counter-pressing Zones
Counter-pressing Zone % of Total Presses
High 38.8%
Mid 54.8%
Low 6.4%

Right 32.6%
Central 39.7%
Left 27.7%

32



Table 5: Counter-pressing averages per 15 minutes
Minutes Counter-pressing Actions Efficiency Press Per Min PAPPM
0-14.59 18 70.29% 1.22 2.24
15-29.59 17 63.69% 1.14 2.10
30-HT 14 64.67% 0.90 1.66
45-59.59 13 66.53% 0.86 1.58
60-74.59 13 65.91% 0.86 1.58
75-FT 10 66.20% 0.67 1.24
OVERALL 85 66.22% 0.94 1.73


































33



Table 6: Counter-pressing action averages across final six games
Mac Game Game Counter-pressing Actions
6 vs Northern Illinois 82
7 vs Buffalo 67
8 vs Akron 69
3 game average 72

9 vs East Michigan 97
10 vs Central Michigan 88
11 vs Ball State 92
3 game average 92

34



Table 7: High-central counter-pressing across entire MAC season
Game High-Central Counter-pressing %
vs Bowling Green 14%
vs Toledo 17%
vs Ohio 12%
vs Kent State 4%
vs Western Michigan 6%
vs Northern Illinois 16%
vs Buffalo 18%
vs Akron 15%
vs East Michigan 16%
vs Central Michigan 14%
vs Ball State 17%

35

 Glossary of terms

Counter-pressing
A tactical philosophy where the team try to regain possession as high as possible into the
opponent’s half to regain possession.

Successful press

Press that led to goal

The defending team wins the ball through counter-pressing and scores within 10 seconds
of the turnover

Press that led to shot

The defending team wins the ball through counter-pressing and has a shot within 10
seconds of the turnover

Press that won possession

The defending team win possession of the ball through pressuring the opponent.
Possession is gained through directly turning over the ball on the field, or pressuring an
opponent to concede a set-piece (throw-in, free-kick, corner).

Press that pushed opposition back

The defending teams’ pressure forces the opponent to pass or dribble the ball
sideways or backwards.

Unsuccessful press

Failed press
The offensive team player either dribbles or passes the ball in a forward direction \
beyond the defensive counter-pressing player.

Failed press that concedes free-kick

Total counter-pressing actions

The total amount of counter-pressing actions completed during the game. Calculated by
adding together total successful counter-pressing actions and total unsuccessful actions

Presses per minute

The amount of presses actions completed by the individual or team during a game.
Calculated by dividing the total counter-pressing actions by 90. Dividing by 90 is due to
games being 90 minutes in length.

Counter-pressing efficiency
The percentage of total presses that are successful by an individual or team. Calculated
by dividing total successful counter-pressing actions by total counter-pressing actions.
36

Possession adjusted presses per minute
Similar to presses per minute except this statistic accounts for the length of opposition
possession of the ball during the game. The percentage of possession for each team has
already been collected by coaches on VidSwap. Possession percentage is calculated by
totaling the number of passes during a game by a specific team, then dividing this number
by the total passes during a game by both teams. This percentage of possession is based
upon a 90-minute game. The percentage of possession during a game is then converted
into a decimal format. For example, 37% would convert to 0.37. A numerical
representation of total possession time is then calculated by multiplying 90 (minutes in a
game) by this decimal. Finally, possession adjusted presses per minute can be calculated
by dividing total counter-pressing actions by the numerical representation of the
possession time.

An example;
• Analysts want to calculate the possession adjusted presses per minute of Team A
• Team A has 53% possession and Team B has 47% possession.
• Team A completed 60 counter-pressing actions during a 90-minute game.
• Team A 53% = 0.53
• 90 x 0.53 = 47.7 (numerical representation of possession time)
• Total counter-pressing actions / Numerical possession time = Possession adjusted presses
per minute
• 60 / 47.7 = 1.26
• Possession adjusted presses per minute for Team A = 1.26 per minute.

xG
A statistical model. This model takes the location of the shot and the location of the goalkeeper,
and calculates the likelihood of a goal occurring from the specific shot. This model requires all
shots from the specific league over the previous season to be recorded.

37