You may be familiar with the following graphic from the NBA. It impressively illustrates how the game in the NBA has changed over the course of time between 2001 and 2020, thanks in part to data analysts.
Dirk Goldsberry processed and analysed countless shots for his graphic and then visualized the findings in a simple way. The visualization and the message is a real eye-opener.
Today we want to try something similar. We're looking at long balls and the duels for second balls that arise from them.
Second balls have been little researched in the analytics community to date. The idea for this blog was born in an exchange with our esteemed colleague Aritra Majumdar (@themachineball). We have been friends for a long time and inspire each other.
We want to find out which zones the long balls are played into and in which zones the balls and second balls are recovered most successfully. And, of course, which following actions the teams use to win the balls.
Selection criteria for long ball
As starting action, we consider long balls from all players out of the penalty area. From a goal kick or open-play. Classic actions are goalkeeper kicks, long balls after short passing combinations under pressure from the opponent or long balls from the goalkeeper after back passes.
We also only take into account balls that are at least hit into the opponent's half. In this way, we want to exclude the shorter and less risky build-up passes.
We have also excluded clearances, as these are defensive actions.
Dataset
We use the data set of the top 5 leagues in Europe and additionally the Super League 2023-24. The data set consists of over 1,500 matches and over 2 million events. In total, it contains around 34,000 passes that meet our criteria and were hit into the opponent's half from inside the penalty area.
Long balls - Distribution per zone
Which zones are the long balls hit into?
The graphic shows the distribution of long balls into the different zones of the pitch.
We have divided the pitch into 11 zones and the percentage values indicate how often long balls were played from the penalty area into these zones.
It is clear that the zones are not exactly the same size. Nevertheless, the data shows that the passes are most often hit into the central zones. This is where the forwards try to tie up the balls and the team tries to win the second balls with good collective positional play.
The central zones 3 and 8 have the highest volume in their band. The band directly after the center line is used much more than the rear band. The distribution in both bands shows that the volume decreases the closer to the sideline. The zones on the outer lines have the lowest volume in both bands.
Balls into the end zone 11 are often wide balls that roll out and are only touched by a player in this zone. The pass end coordinates are registered where the first touch occurs.
Change the game: Rethinking second balls
The data providers rate the pass or long ball as successful as soon as a player of the same team touches the ball first. Regardless of whether the ball can then be held or controlled or not. Touch = successful pass. This is not meaningful enough for us in the analysis.
A long ball often results in a "Second ball". The term "second ball" refers to situations in which a ball resulting from a tackle, a header duel or an uncontrolled touch cannot be controlled directly and is therefore still freely available to both teams.
Teams that are able to systematically win these second balls can often gain a strategic advantage as they are more often able to retain possession and control the game.
In order to show the success rates of second balls, we must first define the success criteria. The goal of the team in possession is to maintain possession in the opponent's half.
Success criteria:
A team wins the long ball or the “second ball” by carrying out the first controlled or successful action and thus gaining control of the ball.
Duels and headers are not controlled actions and the ball is not considered won. In the case of passes and headers without duels, we only consider the ball to have been won when the action has been successful and the ball has reached a teammate.
It can take a few seconds due to headers, duels, uncontrolled touches of the ball or misplaced passes until a team has won the ball and regained control of the game.
Uncontrolled actions are:
Aerial Duel, Header, Defensive Duel, BallTouch, Clearance and Tackle.
We consider the ball won for the following actions:
Pass Successful (pass or header must be successful to secure possession)
Ball Recovery Successful (ball is recovered and secured)
Dribbling Successful or Unsuccessful (player has ball under control before dribbling)
Interception Successful (ball is secured)
Foul by opponent (ball is secured by free kick)
Throw-in for team in possession (ball is secured by a throw-in)
Scenario A:
Goal kick Team A > Header Team A > Missed pass Team B > Successful pass Team A.
Result: Team A has won the second ball
Scenario B:
Kick off Team A > Header duel near touchline > Team B wins header but ball goes out of bounds > Throw-in for Team A
Result: Team A has won the second ball
We subjected our logic to the eye test, validated it on the basis of over 50 games and over 500 actions and it fits very well. We then programmed the logic and applied the algorithm to all 1,500 games.
Success rates of second balls per zone
The graph shows the success rates of long balls and second balls in different zones of the pitch. The success rate indicates how often a ball was won in the respective zone.
The data shows that the first band after the midline has higher success rates than the second.
In addition, the lateral zones in both zone bands have the highest success rates, which are around 5% higher. In summary, it can be said that the more laterally the ball is played, the more successful the ball capture is.
An increase of 5% may seem small at first glance. However, in a highly competitive environment, a 5% increase will give you a competitive advantage and can make the difference between success and failure. And I am sure that there is a lot more potential with a conscious application of the strategy.
Since the vast majority of teams still use our strategy unconsciously, we estimate the potential to be much higher than 5% if it is used well.
The result is all the more surprising because we have seen in the distribution that most balls are hit to the center. So let's go one level deeper and analyze the follow-up actions that are responsible for winning the ball or second ball.
Top 3 following actions after winning the second ball
The following graphic shows the three most frequent follow-up actions per zone that were responsible for winning the ball. Or the first controlled or successful follow-up action with which the second ball was won.
The zones show different priorities when winning the ball, with a first successful pass or header (counts as a pass) being by far the most common reason in all zones with an average of 66%.
Throw-ins are in second place with an average of 18%. Yes, throw-ins. The further to the side, the more relevant throw-ins become. In the outermost zones of the first band, 25% of second balls are won through throw-ins! In the second band even 29 to 34%.
Classic actions for winning the ball through throw-ins are header duels near the touchline or a deliberate pass into out by a defender under pressure.
If a long or second ball results in a throw-in for our team, this is a won ball. This fact is often not taken into account in the analysis and statistics. Provoked throw-ins are an important element for winning the ball in the opponent's half.
This is followed by fouls and dribbling with an average of 7%. Fouls are slightly higher the more central the zone. Dribbles often occur after a missed pass or header by the opponent and are slightly more frequent in the outer zones than in the central ones.
Isn't it surprising that the data analysis shows that zones with lower chances of success are used more frequently and that the distribution strategy is therefore not optimal?
The teams prefer long balls into the center, probably because that's where they have the most offensive players and usually also the strongest headers, and they still want to have options on both sides. However, the lateral zones have a better chance of success. Why?
Causal hypotheses
Possible reasons for the higher success rates in the lateral zones are:
Additional throw-ins are created in the lateral zones, which are not generated centrally.
Winning throw-ins increases the success rate and makes the difference.
Lateral zones are less densely occupied defensively, as defenders focus more on protecting the central areas.
Full-backs are less strong on headers than central defenders
Long balls into lateral zones often have a sharper angle, which makes defending more difficult for the defender, especially in the header duel.
What does this mean in practice?
First of all, throw-ins and fouls must be included in the statistics for second balls won.
Teams and goalkeepers can use long balls to play the lateral zones more consciously and strongly and increase the success rate of winning the ball. New strategies can also be developed or existing ones adapted to provoke even more throw-ins after lateral balls.
Defensive teams in particular, who (have to) operate more often with long balls, can benefit more from this and thus increase their ball possession percentage.
The position of the ball at the kick-off can be positioned on the 5m line in such a way that it favors lateral balls and good angles. If you already know where the ball is going, you can also play the zone more deliberately. You can also deliberately physically disrupt the opponent full-back's header.
More and more teams are bringing in an expert as a set-play coach. Collective and coordinated movements could be used to systematically improve throw-ins and get more out of provoked throw-ins.
In the last minutes of the game with a lead, the focus is less on the attack and more on keep possession. As with a short retain possession corners. Here, teams could also choose strategies to provoke throw-ins or fouls in the side zones to gain possession and also add around 40 seconds to the clock.
We are of the opinion that you can get a lot more than the 5% with a conscious application. As always, however, it is not a question of either/or, black or white. The solutions should be varied in order to remain flexible and unpredictable. I recommend coding a body language sign with the goalkeeper to indicate the variation early on.
Best throw-in ball recovery teams
We have analyzed the teams in the top 5 leagues and the teams that produce the most of their ball recoveries through throw-ins are: RB Leipzig 23.3%, Sassuolo 23.3%,
West Ham 22.4% and Valencia 21.9%. They win over 20% of their second balls won through throw-ins.
The fact that the weakest teams achieve a rate of only 5% shows how high the potential for improvement of this strategy is. .
The long balls won from throw-ins are shown in orange.
In the Super League 23-24, Lucerne won 24.7% and Zurich 23.4% of their won
second balls won through throw-ins.
Team comparision in different leagues
Check out Liverpool's long balls to the right side. We noticed a strong accumulation of second balls won with throw-ins. In the lateral zone, over 55% of the balls won were generated from throw-ins. This highlights the potential of this strategy when applied (likely) consciously.
Incidentally, one of the tactical improvements I suggested to the coaching staff of a Super League team in 2023 was to hit more long balls into the lateral zones.
It may well be that after reading this article you will be more attentive to second balls or even see evidence of this long ball strategy in games that you would not have noticed before.
You only see what you know. Data Analytics can help to bring hidden or unconscious things to light.
The real adventure is not in discovering new landscapes, but in seeing the world with new eyes.” Marcel Proust
The smart and successful use of data goes far beyond the mere purchase of standardized tools. We have many years of experience in the areas of football coaching, match development, match analysis and data scouting. We also bring expertise in business innovation, design thinking and product development to the table.
Did you like the quality of our article? Then reward our passion with your "credits" and like and share this post within your social network. Thank you very much.
You can subscribe to the blog to make sure you never miss a post.
footballytics – change the game
We support clubs, coaches, agencies and players with analysis and consulting services for the use and interpretation of data. To make better decisions in scouting, match analysis and on the pitch.
Blog von www.footballytics.ch Share this post