Blocking the goalkeeper's sightline can increase goal probability

Data from StatsBomb

Summary:

• StatsBomb studies have already shown how pressure on a shot reduces goal probability (xG). However, there may be an extra reduction if the pressing defender is positioned in a direct line between the shooter and goalkeeper.

• If a defender can’t get close enough to pressure a shot, but is positioned in a direct line between the shooter and goalkeeper, there’s a higher chance of the shot being scored.

• It seems that many defenders should adjust their decision making in this specific game moment, moving out of the way of the shot and leaving a clear view, if they can’t get close enough to pressure.

In the 100th minute of a dramatic encounter between Mazatlán and Tigres, Nicolás Ibáñez saved a point for his side. Mazatlán had played with ten men since the fourth minute, but thanks to an organized low block, Tigres errors, and an excellent display from goalkeeper Ricardo Gutiérrez, Los Cañoneros led 2-1.

But with Mazatlán on the verge of earning a historic victory, Ibáñez controlled a half-cleared ball, turned, and found the back of the net to level the match. It was a powerful strike from the Tigres forward, but it came from the edge of the box, and wasn’t hit right towards the corner of the goal. According to Fotmob, the shot had an xG value of 0.13, and a post-shot xG of 0.30. Ricardo Gutiérrez was somewhat at fault then, right?

Well, the replays tell a different story, adding extra context to the situation. The image below shows the moment that Ibáñez struck the ball. Two Mazatlán players are positioned directly between the ball and Gutiérrez, blocking the goalkeeper’s view. Gutiérrez is clearly trying to get a view of the ball, by moving his head towards the left.

With his view blocked, Gutiérrez’s reaction to the shot was slow. Here, with the ball traveling towards his goal, the Mazatlán keeper is still yet to move his feet, as he attempts to track the trajectory of the shot.

Gutiérrez almost reaches the ball, but after seeing the shot, and therefore reacting, late, he wasn’t able to prevent the Tigres equaliser.

Are the two defenders responsable for a 0.30 post-shot xG shot finding the back of the net then?

The team that tried to avoid blocking shots

An excellent article from Chris Summersell during 2021-22 pointed out an interesting observation. Jügen Klopp’s Liverpool had blocked a low percentage of their shots faced from outside the box.

The complete article:

In a few specific examples, we could see how Liverpool defenders allowed their opponents space to shoot from distance, rather than trying to block the shot. They maintained confidence in their goalkeeper’s abilities and wanted to give Allison the best chance of making a save, by giving him a clear view.

The Mazatlán example suggests this was a good idea, but over a larger sample size, does the data back up the decision or not?

Defender positioning impacts goal probability

Thanks to StatsBomb’s free data set, there’s a method of studying the impact of defender positioning when a shot is attempted. In the StatsBomb event data, every shot contains a “freeze frame”, showing the positioning of every player around the ball.

Whilst it isn’t possible with this data to know exactly whether or not the goalkeeper’s view was blocked or not, it can be used to generate an approximation.

In this study, I created a direct line between the ball and the goalkeeper when the shot was taken. Every player that crosses the line counts as someone that might be blocking the goalkeeper’s view. Also, I took into account the distance between the player crossing the view and the shot location, to consider the impact of pressure on the shot.

I filtered out the best chances (0.25 xG or more) and shots closer to the goal (only shots around the box and from distance), to assess situations where goalkeepers would normally have enough time to react. The study contains almost 30,000 shots.

The types of shots analyzed:

• Under pressure, with 1 player possibly blocking the view

• Not under pressure, with 1 player possibly blocking the view

• Not under pressure, with 2 or more players possibly blocking the view

• Far from the ball, with 2 or more players possibly blocking the view

The graphic shows a simplification of the shot types

It appears that defender positioning does impact goal probability, given the context of the StatsBomb xG model. The table shows the differences between goals and xG, within the shot types, plus a control group that looked at all of the shots studied. As expected, no difference in xG and goals was observed overall.

StatsBomb’s xG model already considers pressure impact, however, it seems that there may be an additional suppression on goal probability if the defender is located between the shot and the goal. This further proves the value of players being able to get direct pressure on shooting opportunities, and defenders don’t need to worry about blocking their keeper’s view, as long as they can get close enough to the ball.

However, if the defender doesn’t manage to get pressure on the ball (more than four meters away from the shot), blocking the goalkeeper’s view become a negative action.

This finding goes alongside previous comments from Ted Knutson, StatsBomb founder, who encouraged the usage of screens at free-kicks in order to block the goalkeeper’s view.

This impact seems to disappear if there’s more than one player crossing the direct line, but returns if neither of the players are within seven meters from the shot (“far from the ball”).

José Hernández pointed to his eye in the aftermath of León’s second goal versus Querétaro, suggesting to Diego Reyes that he had blocked his view of the shot.

Why do these differences show up in the data? It seems to be related to the percentage of shots that get blocked. The under pressure shots were blocked more than 52% of the time, with that percentage dropping to 27% for the shot types with higher goal probabilities.

A lesson for defenders?

Analyzing defenders has traditionally been highly challenging with event data, and there’s less clarity over which defensive actions are the most valuable to a team, compared to attacking actions. This study could be used to better analyze and scout defenders, for example, by seeing how often a player gets pressure on a shot, as a percentage of shots faced near to them, or how often they block the view without generating pressure.

It may also be possible to create an “expected blocks” model, and see which players are most effective at blocking shots, given their positioning when the shot is taken, or just before the shot, by using tracking data.

This study also seems to contain a lesson for defenders and their decision making. If a defender isn’t close enough to pressure the opponent when they shoot, it’s best to move towards one side and leave the goalkeeper with a clear view. Ideally, a defender will be positioned to block one side of the goal, limiting shot options, without blocking the keeper’s view.

Liverpool’s center-backs allowing a shot from outside the box, whilst leaving the goalkeeper with a clear view.

This analysis goes against the view point of many coaches. However, it seems that it may lead to fewer goals being conceded, and could’ve earned a historic victory for Mazatlán against Tigres.

Thank you for reading, if you’d like to contact me regarding any questions, comments or to consult my services, you can contact me on:

• Bluesky - https://bsky.app/profile/tomh36.bsky.social

• Linkedin - https://www.linkedin.com/in/thomas-harrison-a682a2175/

• Twitter/X - https://x.com/tomh_36

También, se puede leer este artículo en Español- https://cambiodejuego.beehiiv.com/