Soccer Analytics: Does Counterpressing Work?

counterpressing example

In soccer, there are legendary coaches who have asserted that upon losing the ball, teams that regain it within five, six, or even eight seconds have a higher chance of keeping the ball, and indeed, scoring. This is the foundation of what Jurgen Klopp called “Gegenpressing” and led to the rise of the RB Leipzig team in the Bundesliga, whose coach Ralf Rangnick stated that goals were most often scored within eight seconds of winning the ball from the opposition. This seems like an amazing statistic, but is it data driven or is it merely legend?

In the 2021 paper “Data-driven detection of counterpressing in professional football” (link) the authors, Pascal Bauer and Gabriel Anzer, describe a method for using supervised machine learning to detect counterpressing in video. If automated detection was possible then they hoped to be able to better evaluate some of these counterpressing rules of thumb.

History of the Research into Pressing Tactics

Much of the data behind the counterpressing strategies started with a man named Charles Reep. He was one of the first who studied the game for the purpose of collecting data that might be able to reveal new insights. He captured piles and piles of data — many of these in hand-written notes — to better understand the game. There is much that can be said about Charles, but this is too short of a post to discuss his successes and miscues. To our question about transition successes, however, in one paper that he authored in 1968 he found that 30% of the time that a team forced a transition and gained possession they were able to make a shot on goal and indeed, 25% of all goals came from regained possessions in the attacking quarter. This data wasn’t much used outside of Reep’s circle but in 1999 A.J. Grant collected data from the 1998 World Cup and confirmed these numbers. This relationship between transition “recaptures” and goals has been confirmed in papers from 2014 and 2018 as well. There have also been studies that learned that teams relied on counterpressing more often when behind in the score than when ahead. This would indicate to me that some teams know of the power of counterpressing, but don’t structure their main strategy around it, much like the press in basketball. Additionally, studies have discovered that teams that recover the ball more quickly after losing it tend to win more games. All of these things seem intuitive, but it’s helpful to see that there are measurements and data behind the notions.

The paper concludes a few things that I find valuable.

  1. First off, researchers have been able to discern counterpressing strategies using machine learning. This is very important, because it reduces the labor required to classify significant events and approaches in soccer.
  2. Using these automated detection methods, these same researchers also found that counterpressing is more likely be successful near the sidelines and that numerical superiority near the ball when it is turned over increases the chance of winning it back. Both of these, of course, makes good sense to me.
  3. Within the German Bundesliga, teams follow very different transition strategies and these differences could be detected by the machine learning. Each of these approaches had different levels of success regarding turnover recovery and goal scoring.
  4. Successful teams—measured against their final ranking—tend to use the counterpressing strategy more efficiently, providing credibility to the coaches that use it as a major offensive counter-attacking strategy.


Though there seems to be data that ties a fast recovery during transition with a higher probabiity of actually scoring, I was actually unable to find any data that actually quantified the number of seconds after a turnover where a transition was more likely to lead to a goal. Perhaps the number is somewhere in the mountains of tablets where Charles Reep recorded his data, or perhaps its just legend. But the data seems very clear that pursuing a counterpressing strategy with players who are highly fit and can fly all over the field (people like Tyler Adams??) allows teams to have a higher probability of scoring in games than teams who do not. Of course if Lionel Messi plays for the non-counterpressing team, all bets are off.

New Blog “Tag”. Soccer Analytics.

Arizona Youth Soccer, credit Tod Newman

I’ve been thinking about Soccer analytics for some time now. I coached a Middle School soccer team last season and decided to develop some simple measurements that might allow the team to see improvement. I selected shots, shots on goal (good shots), and turnovers (losing the ball for more than 3 seconds). As it turns out, without a focused team manager, it is difficult to collect these simple measures, even when carefully defined. Middle School attention spans are not long, everyone!

So in this light, I recently picked up a copy of Ryan O’Hanlon’s book “Net Gains” (link to Amazon) and was inspired to tune up my old COVID stats and visualizations (check out my COVID-19 tag if you really want to relive those times) for something much more interesting to me now. Since I haven’t seen much in the way of MLS analytics, I figured that might be a good place to start.

What do we Know about Soccer Analytics?

First, soccer is a highly unstructured game which typically low numbers of scores. Think about baseball… the players are often in set positions, both defensively and offensively. The batter stands in the batter’s box, the pitcher is on the mound, runners stay within the basepaths and stand on the bases. Defensive players tend to stand most of their time in the same spots. A Baseball diamond is a huge space that players will only occupy a small portion of throughout the game. It is rare that an official makes a single call that flips the outcome of the game.

Soccer Analytics Are Hard because Soccer has Low Structure!

Now think about soccer. There are many variants of soccer formations. Some clubs have traditionally used a 4x4x2 or a 4x4x3. But there are creative variants of these formations that could get adopted for special situations. There are very few spots on the pitch where players have a low probability of occupying during a game. This contributes to making soccer a very hard game to collect data on and analyze. This difficulty has also led to a lack of “killer” metrics that are indicative of team success. Indeed, in the book ‘Soccereconomics’, the authors Simon Kuper and Stefan Szymanski, find that in European leagues the amount spent on players’ wages is the most highly correlated measurable with team success that is known! And of course, with many games decided by one goal or tied, a single call from an official can reverse the outcome of the game. This is discouraging, to say the least, for anyone that wants to find any other signals hiding under that noise. Billy Beane, the former GM of the Oakland Athletics baseball team became famous for finding soft signals in the data that the high-spending teams hadn’t been paying attention to. These are hard to find in soccer.

One Early Metric I Like (And Think I can Collect)

One metric that I’m interested in is Expected Goals (both For and Against). This is a measure of (my words) the times when a team makes good decisions to put themselves into the position of taking a good shot. Most of the data indicates that in soccer, ten decent shots on goal will on aggregate score one real goal. So most shots have an Expected Goals score (xG) of 0.1. Some shots from better locations have a higher factor. Overall, it isn’t hard to count up the xG during a game. A team that has 3 xG but only 1 goal in a game could be thought to have fallen on the bad side of the luck that drives much of what happens in soccer. The xG for a team’s opponent can also be calculated. I use a feature called npxG that I find on the site (link to site) because it takes penalty shots out of the mix (I’m not a big fan of penalty shots, which seem highly subjective to me, and therefore unpredictable). Then the ratio of npxG “for” your team to npxG “against” your team is a very good ratio to measure with one number how your team performed.

Early Analysis on MLS Soccer

I collected data and did some data engineering on it to allow me to plot two things for each MLS Club. First, the annual salary for the club (in millions of dollars) and second the npxG ratio. The hypothesis is that when these are plotted for the teams in rank order by their number of points for the season, maybe we’ll see some trends.

2022 MLS Results

2022 MLS End of Season Results, comparing final points, npxG Ratio, and Team Annual Salary

This is a pretty satisfactory result ans shows a trend that correlates a high npxG ratio with success. Actually, the top npxG ratio of all goes to LAFC, who won the 2022 championship over Philadelphia (the second highest ratio). The trend is not linear down, reflecting the impact of chance on the results of individual games. Note however that there is no trend at all regarding team salaries and final results. I have seen papers that indicate that others haven’t found any trends with MLS salaries either, ostensibly due to the way the MLS implements a salary cap.

Can we Predict the 2023 MLS Championship yet?

2023 MLS Current Status, comparing current points standings, npxG Ratio, and Team Annual Salary

So as is obvious, the trend is non-existent after 15 or so games of the 2023 MLS season. My suspicion is that it’s too early in the season for “luck” to have filtered down to its normal level.

Plans for Future Analysis

I’m planning to evaluate more MLS seasons for this trend and incorporate a number of other metrics that are interesting and available (% possession is one that I tried to estimate for my Middle School soccer team, but an accurate % possession might have good correlation with performance. I’ll roll these kinds of articles out periodically. Please weigh in if you have interest and/or expertise to contribute!