When looking for data online soccer data sources, I stumbled upon this great post by Jonas at Opisthokonta. He collected coordinates of soccer stadiums in Europe and used them to analyse the relationship of distance travelled by the away team and the home field advantage (HFA). He finds that there is no effect of distance on HFA, using data from two seasons and four leagues. In a later post, he was so kind to share the R code he used along with the stadium coordinates. Thanks again for sharing!
This got me thinking about home field advantage in soccer in general. So I started looking into it a little and quickly the post I wrote turned way too long, so I split it in two: In the first part, I’ll just show some descriptive stats on HFA to get a feel for the magnitude across different leagues and over time. In the second part, I’ll try to replicate Jonas approach and connect distance travelled to HFA, adding several leagues and seasons to the data.
A note on data sources: All results data is from http://www.football-data.co.uk, the stadium coordinates which I’ll use for the second part are either from Jonas or from my own research on Wikipedia or GoogleMaps. I use data from season 2009/2010 up to and including 2013/2014 and from the first two leagues in Germany, England, Italy, and Spain. That’s 16,140 games, if you’re counting.
Ok, after this prolonged throat clearing, let’s start: First, how do you define HFA in soccer? The topic is well researched, with several posts looking into it, for example here at Bleacher report in a general overview , here at Pinnacle Sports from a betting point-of-view, here again Pinnacle Sport for a detailed analysis of HFA in the Premier League, here from Chicago Booth (wow!) comparing different sports and giving some possible reasons, here the Wall Street Journal using Champions League and Europa League data, here Freakonomics comparing some sports and citing Dohmen’s great paper on HFA in Germany, here Soccer by the numbers on HFA by team in the Premier League.
Most of the time, HFA is measured by the difference in winning percentage at home and away. This is certainly useful, but of course ignores draws. For example, if team A has a 70% home winning percentage and draws every away game, this would show up as the same HFA as team B which also wins 70% at home but loses every away game. To capture draws as well (and because win percentage is hardly ever used in soccer anyway), I chose the difference between average points at home and average points away as the definition of HFA (Of course, to judge the role of HFA for a single game, you need a different measure, we’ll come to that in part II when we need it).
So how big is HFA? The following table depicts the difference between average points at home and away.
Well, pretty big and very consistent. To be honest, I was surprised at how consistent HFA is. On average, teams make between 0.6 and 1.05 more points at home than away, with the total HFA over all seasons and leagues being 0.81. There seems to be a slight trend towards a lower HFA in the last 5 years. In addition, HFA is higher in the first division than in the second division (with the exception of Germany and the 12/13 season in Spain).
Let’s have a closer look at this difference between the top flight and the second division:
First, the downward trend in HFA exists regardless of first or second league and second, the difference between divisions remains if we average over all countries. Especially the fact that HFA is higher in the first division is really surprising. I would have thought that most factors causing HFA are more pronounced in lower leagues…
A different way to look at HFA is by country, aggregating both divisions:
England and Italy follow a very similar pattern, and Germany is a little more erratic but in the same neighborhood. In Spain, however, teams seem to enjoy a consistently higher HFA than in other countries.
So far, so good. Now we can dig a little deeper and look at individual teams. Who has the highest HFA? Can we find some pattern where enthusiastic home fans translate to a higher HFA? I’ll show a table for each league or it gets really messy. The teams are ordered by descending HFA in each league, the colors show HFA over all teams, regardless of league or country (green=high HFA, red=low HFA).
We’ll start with Germany’s first league:
To be honest, some surprises right at the start: Leverkusen and Hanover with Top-5 home field advantage? That’s strange, given that they do not really have a reputation of fanatic fans, a big stadium or an out-of-the-way location. BVB, Schalke, Gladbach, Düsseldorf are fine regarding the reputation of their fans. In general, good teams seem also to enjoy a large HFA. Hannover and FC Köln with the two seasons with the overall highest HFA, though Köln is pretty much all over the place. Second German league:
Düsseldorf again on top, though their HFA has been steadily declining in the last 5 years, regardless of league. Here, the connection between being a good team in general and having a high HFA is not so clear (Of course, promotion messes up the comparison a lot more here.). And we have the first season where a team actually managed to be worse at home than away: Congrats, Ahlen! Ok, on to the Premier League:
Here we have the season with the highest overall HFA: ManU in 2010/2011: A ridiculous 1.84. If not for last season where they somehow lost their HFA, they would have a top HFA. In general, general quality seems to correlate with HFA again. (I don’t really know enough about Everton to comment, Burnley can be classified as an outlier, I think.) Similar to Germany, the worse the team, the more erratic the HFA pattern over time. England’s second division:
I don’t really have a comment. But -0.3 HFA, Brimingham, seriously? What’s going on there? Ok, halfway done, on to Italy’s Serie A:
Hmm, intruiging, the two sicilian clubs Catania and Palermo with an very high HFA. Travel distance after all? The other clubs don’t really show any consistent pattern, Juve and Milan the big name clubs which do not really have a high HFA. Serie B:
Wow, extremely high turnover, lots of teams only one or two seasons in there, so hard to really make out anything. As for the other countries, some big swings in HFA from season to season in the lower division. (Though Crotone, the club with the most consistent HFA, is in the very southern part of Italy…) Almost done, finally the country with the highest average HFA, Spain:
Ok, again the islands (Tenerife and Mallorca) with high HFA. Barca, Real, Athletic also with high HFA, confirming the pattern that good teams generally also have a high HFA. Second Spanish division:
Pretty large HFA at the top, though again probably not as telling as several clubs are only in the second league for one or two seasons.
Phew, that concludes the descriptive first part of the post. What have we learned?
- Home Field Advantage is real and significant in all big 4 European soccer countries
- It has been declining since 2009
- It is largest in Spain
- It is (with the exception of Germany) higher in the first division than in the second division
- There is high variance between teams w.r.t. their individual HFA
- Good teams tend to have a large HFA
These stylized facts are pretty interesting, I think. In the next post, I’ll look at possible reasons for the HFA.