With the resumption of most European football leagues after a hiatus due to the coronavirus pandemic, almost everywhere matches are played without spectators or with limited numbers. How empty stands affect the advantage of the receiving team, we will consider in our material.
Recently, a study was published showing the percentage of home wins in the German Bundesliga fell from 43% in pre-halftime games to 21% in 56 games played with empty stands. The Bundesliga season is over, but other championships are in full swing, and matches are also played without spectators. So now is the time to look at the changes that have taken place and see if bettors should consider the home side’s advantage when placing bets.
Reasons for home advantage in football matches
In the old days, teams won at home almost twice as often as away, although, of course, in different championships the indicators differed from each other. Despite the fact that the home field advantage has slightly decreased in recent years, teams still win at home stadiums in about half of their matches, and away – only slightly more than 25%.
Various studies have looked at the reasons why home play promises certain benefits for a team. A number of arguments have been put forward in favor of this hypothesis. These are, for example, the familiar lawn and the size of the “glade”, the fatigue of the visiting team from moving (and even more from a long-distance flight, as, for example, in Russia), the support of their fans, as well as the noise of the crowd pressing the referees.
The results of the derby taking place at a common stadium for the two teams, when the first two of the listed factors are absent, suggest that the pressure of the stands and subconscious refereeing in favor of the hosts may be the two most significant factors of influence. With regard to refereeing, usually the hosts are punished for fouls, including “mustard plasters” and sending off, less often than the guests.
Of course, the hosts feel the support of their fans, but this factor of influence can sometimes hinder, not benefit. In fact, the players of the host team are more eager not to lose than the guests, so as not to disappoint their loyal fans.
The psychological peculiarity of any person associated with the fear of losses was deduced by experts in the course of research on behavioral economics. The same is true in sports: the desire for a mandatory victory is less inherent in players than the desire to escape from defeat.
German Bundesliga statistics
This season’s Bundesliga has already ended, resuming earlier than other major European national leagues. A total of 83 matches were played without spectators, including one before the start of the official national lockout in Germany. Of these, 27 (32.5%) led to a home win. Obviously, there has been some mean regression since Reuters published its research.
However, this figure is still well below the dock frequency, which raises questions about whether it is statistically significant. At first glance, it seems not: with a probability of 11%, the observed home wins, draws and away wins in matches played behind closed doors happened by accident, assuming that the home advantage did not have an impact and the results of the games in the indoor stadiums were as expected.
This is not a statistically significant result. Draws and away wins reduce this probability to 5%. This may be enough to warrant mentioning data in a journal article, but it ultimately suggests more data is needed.
Germany’s second echelon, Bundesliga 2, which has resumed its season at the same time, shows a completely different picture. Of the 81 matches played behind closed doors, 35 (43.2%) were home wins, compared to 41.3% for the ancestral matches. This most likely means that the German elite division simply required a higher average regression.
Statistics from other European leagues
For now, a number of other major European leagues have resumed their 2019/20 seasons, including the English Premier League and Championship, the Italian Serie A and B, the Spanish La Liga and La Liga 2, and the top divisions in Portugal. Turkey and Greece.
At the time of this writing, 3426 matches have been played in all of these divisions this season, including the German Bundesliga and Bundesliga 2. Of these, 2,924 were played with fans and 502 without them. The combined home win rates for these two periods are 43.1% and 38.8%, respectively.
This drop is not as noticeable as for the German top division alone, but with increasing sample size it is still statistically significant (6%). However, as far as the Bundesliga data is concerned, there is a lack of evidence to suggest that something is actually happening in terms of home field advantage.
Factors of influence of empty stands on football matches
Taking a closer look at the match data reveals interesting things. As noted, research has shown a disproportionately low number of home fouls. As the latest data show, fouls for which the referees punished the hosts increased from 13.4 in front of spectators to 14.3 with empty stands.
According to these figures, the probability of such a random change is 1 in 250,000. Guests broke the rules a little more often, but the difference is not statistically significant. A high significant increase was observed in the 2nd Bundesliga (there were 12.1 fouls for the home teams and now 13.6, the probability of an accidental increase is 1 in 1000).
There are two other significant changes to match data related to goals scored. First, the total number of shots on goal taken by the hosts decreased from an average of 13.1 to 12.0. The chance of such a random change is 1 in 350,000.
On the contrary, this practically did not affect the number of shots on goal by the guests. Although the receiving teams in the 2nd Bundesliga have scored more goals on average when the stands are empty and therefore won proportionally more matches, their number of shots on goal has dropped significantly, from 15.0 to 12.8. The odds of this random change are 1 in 10,000. Looking at the results for this small sample (81 matches), it might seem that teams are particularly fortunate and / or scoring skillfully.
Secondly, the number of corners earned by the home team has decreased from 5.62 to 5.00 on average. The chance of such a random change is 1 in 350,000. As in the first example, almost nothing has changed for the guest teams.
How did these changes translate into goals scored? The average for the home teams decreased from 1.48 to 1.38, the probability of such an accidental change being approximately 5%. Of course, it takes a fair amount of luck to score a goal, so the correlation between corner kicks and shots on the one hand and goals scored on the other will be weak.
So, significant changes in the number of shots on goal and corners will not necessarily lead to a significant change in the number of goals scored, at least in samples of up to 500 matches. Expected goals (xG) theorists should be familiar with this weak correlation.
But if we take into account the high significance of the reduced number of corners and shots on goal by the receiving teams, we can expect that the more matches will be played with empty stands, the higher the statistical significance of the decrease in goals scored by the hosts and, accordingly, the proportion of meetings they won … As the sample grows, so will the signal-to-noise ratio.
Fewer goals and shots on goal in the German Bundesliga were reported in one study. It was also suggested that the players subconsciously prefer to make a pass rather than play in such a way that the fans will rage with delight. Other commentators have noted that matches without fans are becoming less stressful and more like training “fights”.
When it comes to fear of loss, home teams are now less afraid of being hit in the mud in front of their fans and are more inclined to focus on technical aspects of the game than risky attacks and unprepared shots on goal.
Impact on the betting market
A number of experts noted that the drop in home win rates allows bettors to find value in a draw or an away team win, if the markets have not yet shown such a trend (although this does not seem to be the case). Excluding the betting margin from Pinnacle’s Pro Close Line odds, we see that the average expected home win probability of 44.7% for ancestor games has dropped to 41.7% for post-like matches.
While this is significantly higher than the actual 38.8% home win rate, it should be noted that home teams were already underperforming relative to market expectations for the season prior to early March.
Aside from the occasional influences, it would seem that BC Pinnacle was aware of the impact of losing fans on home teams, or at least quickly adapted to the new reality it presents. As a result, BC experts have calculated that the difference in the probability of a home win is not as significant as it seems at first glance.
Impact of lack of spectators on league standings
In other materials, it was noted that luck began to turn less often to face the receiving teams, and this was noticeably reflected in the final league tables. For example, they suggested that English Premier League clubs should not play on neutral fields (as it eventually happened) due to the erroneous assumption that the home field advantage will not disappear if some teams play at their stadiums.
So, while the impact on some metrics is significant for the receiving teams, the odds of winning little change because of this. Perhaps more importantly, if the home teams have lost some advantage, it follows that the guests have gained it.
To test the effect of completely removing a home field advantage, we applied the calculations to the entire 2018/19 Premier League season and performed Monte Carlo simulations analyzing expected league points and league rankings based on BC Pinnacle’s final odds. The expected probability of winning was calculated as follows: first, the margin was removed, and then it was assumed that the probability of winning at home, drawing and winning on the road was 36%, 28% and 36%, respectively.
The result was practically the same, within the margin of error, as in the Monte Carlo simulation with the home field advantage. All 20 teams were in the same places and scored practically the same number of points. They could have won fewer matches at home but more away.
Of course, if after the start of the pandemic some teams had more matches left at home than away, they lost their advantage. But given that the odds of winning a home game have changed by only about 3% on average, this decrease in advantage is unlikely to have a big impact on the standings.
The Image below shows a comparison of the final results for the Premier League teams, first on March 14 (just before the season was interrupted) and then on June 17 (just before the restart):
The results of the matches resumed after the pause of the European football leagues confirm the hypothesis that in the absence of spectators, the hosts play worse than always. At the same time, the referees more often punish them for fouls, and the owners themselves are less likely to shoot on goal and earn fewer corner kicks.
However, this practically does not affect the number of goals scored and the probability of victories, according to statistics. Still, some home field advantage remains, and whatever the impact, bookmakers can identify and exploit this trend before bettors. So you can hardly find an edge over the betting market here.