I admit it. I'm a math geek...and a sports geek. This can be a dangerous combination.
Lately, I became interested in calculating team-by-team home field advantage in the EPL.
English football is full of venues that are rich in both history and atmosphere. But which home sides enjoy the greatest advantage? I decided to crunch some numbers to find out.
Let me be clear, the question is not which venue is most likely to see a victory by the home side. Not precisely, anyway. It is still possible for weaker teams to gain an advantage at home. The trick is to separate the home team from the home venue.
Think of it this way: Is it tough to win at Old Trafford because it is Old Trafford, or because that's where Manchester United plays?
So how to do it? Well fixtures like the EPL, which follows the home-away double round robin format, offer a clean way to sort it all out (at least for games within the league). Simply put, it hinges on calculating how much better a team does at home than on the road.
Over the course of a season, the average disadvantage of the road games will equal the average home-field advantage for the league. So, by removing the the 'away-disadvantage', we can isolate the advantage gained on the home pitch. (If you would like an explanation of the detailed math, you may email me.)
Using data from 2000/2001 through 2006/2007 (sorry, full seasons only). I determined the average home field advantage for every team that played in the EPL for at least four of the seasons covered. The results are given below. The values are given in goals per game. Average home-field advantage for the EPL over the period was 0.4 goals per game.
Team | HFA
Portsmouth | 1.06
Fulham | 0.68
West Ham | 0.65
Arsenal | 0.61
Tottenham | 0.59
Everton | 0.58
Newcastle | 0.51
Bolton | 0.44
Liverpool | 0.43
Chelsea | 0.39
Southampton | 0.39
Manchester United | 0.37
Aston Villa | 0.36
Birmingham | 0.35
Middlesbrough | 0.32
Charlton | 0.25
Blackburn | 0.21
Manchester City | 0.13
Leeds | -0.04
My first instinct with a table like this is to look at the top and bottom rows to see if they check out. Unfortunately, both jump out for discussion.
First, how could Elland Road be that bad? Surely the numbers presented aren't in agreement with its reputation as a tough venue. Perhaps Leeds draws enough of its own fans when playing away to effectively take home-field advantage away from other teams. That would skew the measure of their home field advantage downward.
Even though these are my numbers, I'm not convinced that Leeds would be just as well to play its home games in an empty stadium (which is what the calculated home field advantage suggests.)
Secondly, Portsmouth's value is very high. Looking at their performance this year, the term 'regression to the mean' comes to mind. When I recalculate this at the end of the current season, the advantage measured at their pitch should be more reasonable.
But aside from the two outliers, the table seems pretty fair.
And now for some honorable mentions for teams that (while not making the cut of four seasons at the premier level) show very good home-field advantage.
Norwich, Sheffield United and Wolves all had advantages of 0.9 or better. Bradford and Crystal Palace were also strong.
On the other end of the spectrum, dishonorable mentions go to Sunderland, Coventry Wigan, and Leicester (with Sunderland being the worst). These were the only four teams aside from Leeds to show a negative home field advantage.
And to answer a question I asked back toward the beginning. It appears that it's tough to win at Old Trafford because that's where Manchester United play.
Mark Myers is the founder and primary developer at ChanceBot.com, a sports simulation and analysis website. He can be reached at firstname.lastname@example.org