clock menu more-arrow no yes mobile

Filed under:

Westfalenstats: Has Finishing Improved in Empty Stadiums?

Looking at xG Overperformance in the Premier League and the Rest of Europe

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.


While sports in some countries are beginning to open up to fans, the top leagues in Europe have only opened stadiums partially, if at all. The Premier League is one of those leagues that is still playing behind closed doors, which is unfortunate, given the utter madness that has been taking place on the pitch. There has been a glut of goals scored, and some major upsets, including the the reigning champions Liverpool being thumped 7-2 by Aston Villa.

So far this season, the Premier League is averaging 3.29 goals per game, compared to 2.72 per game last season. If this average is maintained until the end of the season, this would result in around 1250 goals, which would be a little over 200 goals more than in the previous campaign. So what is going on?

The Premier League’s Sharp Shooters

The Telegraph recently published an article that posited that Premier League players are “more relaxed” when playing in empty stadiums, resulting in better finishing and a higher quantity of goals across the league. As evidence of this improved finishing, Daniel Zeqiri compared the total goals scored by teams in the Premier League with their expected goals totals, and demonstrated that there are currently 17 out of 20 teams currently overperforming xG.

xG Overperformance in the Premier League
The Telegraph

I think the logic underpinning this theory is relatively clear. The idea is that the empty stadiums reduce the pressure a player experiences when taking a shot, and that this lack of pressure allows them to relax and play to the best of their abilities, meaning that they score more than they would in front of fans.

However, I’m not sure the theory holds under closer scrutiny. Not only should pressure impact people differently, I don’t think the theory aligns with our understanding of xG, which suggests that the qualities that define the best goalscorers is less to do with finishing and more to do with getting high probability shots on goal. Most of all, the sample size is simply way too small to draw any conclusions from. The Premier League season is about 6 games deep (5 at the time this article was published in The Telegraph). Not only is it difficult to say anything about the number of goals scored in such a small sample, it is also possible that we see a regression to the mean, and the next 6 games are boring, low-scoring nonsense.

While the article states that the number of teams running ahead of their xG totals is likely to decrease as the season progresses, I think it’s possible to test their hypothesis using the data available to us now.

Comparing Finishing Across Europe

In order to examine xG overperformance in a little more detail, it’s necessary to increase the sample size. One way we can do this is to compare this xG overperformance in the Premier League with other leagues in Europe.

The difference between team non-penalty goal (npG) and their non-penalty expected goal (npxG) tallies in the big five leagues doesn’t suggest any overperformance. However, the rules regarding attendance vary across the big five leagues. While La Liga and the Premier League are still playing behind closed doors, a limited number of fans have been allowed to attend games in the Bundesliga, Serie A, and Ligue 1. Therefore the above figure may capture the over/underperformance of npxG so far this season, but it doesn’t tell us about finishing in empty stadiums.

Instead, we can look at differences between each league and compare those that are playing in entirely empty stadiums versus those that are playing in front of fans.

On average, across all five leagues, teams are overperforming by about 0.28. Specifically, Bundesliga teams are overperforming by 0.1, and Serie A teams by 0.85. At the same time, La Liga and Ligue 1 teams are underperforming, on average, by -0.485 and -0.39 respectively. Finally, Premier League teams are currently averaging an impressive overperformance of their npxG of 1.32. The difference between team over/underperformance in the leagues with fans and those with no fans is not statistically significant, and when we visualize team npxG performance, you would be hard pressed to identify those that are playing behind closed doors.

There are some outliers in both directions, such as Lyon’s -9.3 underperformance or Napoli (6.1) and Bayern’s (5) overperformance, but there is no obvious difference between the leagues themselves.


We’re seeing a lot of weirdness going on in football this season. There have been upsets all over Europe and a lot of goals. With the increased injury risk posed by the packed schedule for top-flight teams, especially those playing in Europe, and the potential for coronavirus exposure and players having to go into quarantine, it’s likely that the weirdness will continue throughout the season.

However, all this is not limited to the Premier League, nor to those leagues that are still playing in empty stadiums. While the Bundesliga is averaging close to the same amount of goals per game as last season, and La Liga has seen a decrease, Ligue 1 is experiencing an increase of similar proportions, and Serie A has seen an even greater increase than the Premier League. Therefore it’s unlikely that any of this can be explained by the empty stadiums. It is also too early to say whether the number of goals scored is any more than a short-term blip. We should probably wait a little longer before trying to draw any conclusions!