{{ activeLang }}
  • {{ lang }}

Expected goals prepare football for statistics

Posted on:
Reading time:

Night after night we are treated to football experts manning the talk show table. Very entertaining and wonderfully accessible due to the football humour. But also, often from the lower abdomen. If a player loses the ball three times in the summary, he suddenly can't do anything about it and the next bad buy is a fact. Because the gentlemen mainly look at each other a lot, it is quickly taken over by other football tables and it is a fact: this player can't do anything about it.

Nowadays, however, comprehensive statistics are available for almost all players. Mileage, pass accuracy, pass direction, runs, dribbles, front assists, assists. All statistics that can support the strength of a player to a greater or lesser extent. Yet the analysts hardly look at that. While to a large extent it objectifies a player's performance.

Now there is a relatively new statistic that says a lot about the true balance of power in a match: the expected goals. In just a few years, this statistic has taken football by storm. Expected goals is a value that indicates whether a goal attempt results in a goal. It does this by cracking data from thousands of historical shots and filtering them based on factors such as distance, type of shot, type of pass and the number of defenders between the shooter and the target. All this information is used to create an average percentage probability that a particular shot will go in, taking all of these factors into account. This statistic indicates whether the statement "Luuk de Jong really should have scored this one" is true.

This will objectify football on two levels: team level and individual level. The best example is Juventus in 2015-2016. They started the Serie A season badly. They only won three of their first 10 games, but their expected goals stats were excellent. They dominated most of their matches, creating many chances, but were not rewarded with victories. The Old Lady continued on the same path, her luck finally turned, and they won 15 games in a row. At the end of the season, they became champions with nine points. Most clubs would have panicked and fired the coach. However, it would clearly have been the wrong decision to sack coach Allegri and the expected goals indicated that.

In addition to comparing a team's expected goals, we can see which players hit the target. When a player exceeds his expected goals, it suggests that he is lucky or a great finisher. If a player exceeds his expected goals for a few games and is not known as a super striker (like Giakomakis at VVV), he is probably on a hot streak that won't last forever. But someone like Dušan Tadić, who scores more goals than he gets, is clearly an above average finisher.

However, analysts are having a hard time with this. For example, expected goal numbers for individual matches are useless without context. When a team scores a few early goals, which come from difficult chances, and then doesn't create anything more, it will negatively affect the stat. Suppose the opponent fires many low-quality long shots, that team then gets a higher expected goals statistic. However, that doesn't mean the opponent deserved to win the match. So you see, statistics don't say everything.

Sign up for the newsletter

Do you want to stay informed about the latest developments regarding video analysis and SPORTSVIEW? Register now and receive a monthly SPORTSVIEW update!