Tuesday 14 December 2010

Home and Away

In the calendar year of 2010, Liverpool Football Club have won just five of their 28 away games in all competitions. At home, they have won 20 out of 27 fixtures. Despite one of their most wretched periods in recent years, Liverpool have lost just twice in normal time at Anfield (one of the defeats being the loss Chelsea in May, by which point Chelsea were still chasing the title but Liverpool’s season was over). Liverpool’s points average at Anfield in 2010 is 2.43 points per game. In the history of the Premiership, only Chelsea have sustained a higher points average over a 38 game season, playing half their games away. The point isn’t that Liverpool’s home form is particularly good (plenty of teams have had better home records – Chelsea were unbeaten at Stamford Bridge for 86 matches), but that even a decidedly mediocre Liverpool team looks world-beating at home.
It is a phenomenon Liverpool manager Roy Hodgson is familiar with, forging a reasonably successful managerial career despite winning only 13 of the 106 away games he has managed in the English league. These examples are extreme, but it’s a general and universally acknowledged trend. Of last season’s 380 Premiership games, 193 were won by the home team, just 91 by the away team.
This has always struck me as bizarre. In a game of football between well-honed professional teams, the better should generally win. Of course, part of the joy of football is that brute luck and circumstance ensure that the best team doesn’t always win. But why should brute luck and circumstance always favour the home team? And even if we accept that they do, why should they affect the game so much? Some rudimentary maths suggests that the result of something like 20-25% of matches is influenced by where the game is held.
In this era where football teams spend thousands on understanding every determinant of victory, it should be no surprise that sports scientists have given a lot of attention to this crucial one. Numerous explanations have been advanced, but none seem capable of explaining the sheer scale of the phenomenon.
Intuitively, the most obvious explanation of home advantage is the direct motivational effect of playing in front of a supportive crowd. Footballers have to run kilometre after kilometre, and the extra encouragement of all those around them willing them to succeed could account for the home side putting in the little bit more effort necessary to win.
The economists Mark Koyama and J. James Reade have come up with an inventive alternative account of how supporters influence players’ motivation. They model a classic principal-agent scenario, where supporters seek to ensure maximal effort from their team’s players, and players try to get away with putting in as little effort as possible. Since football is a team game, and results are only indirectly influenced by any given player’s effort, it is possible for players to ‘shirk’ (play at less than 100% - or should that be 110%?) without anyone realising. Koyama and Reade suggest that teams play better at home because their effort is observed by more supporters, so they are less likely to get away with shirking without being detected. They cite the declining proportion of home teams winning in the Premiership as evidence of this, arguing that with more games available on the TV and internet, it is easier for fans to monitor the effort of their team’s players and so players have less of an incentive to shirk during away games than before.
The problem with both these accounts is that they offer peculiar accounts of player motivation. Principal-agent models are usually applied to workers who don’t like their jobs. But many football players claim that they would do it for free. Sportsmen are notoriously competitive – but Koyama and Reade imply that footballers do not care whether they win or lose. As much as anything, football players are motivated by success, winning matches and trophies. This is what makes it to so surprising that home advantage is so influential – when there are promotions and titles on the line, why should the background noise be so important?
It seems clear that the noise does affect referees. Anecdotally, the rarity of penalties for visitors Old Trafford or the suspicious amounts of stoppage time in games Manchester United are losing there imply that officials are cowed by spectators. Indeed, it is hard to blame referees for doing what they can to avoid abuse from tens of thousands of people. Of course, as retired referee Jeff Winter observes, referees may go the other way and give decisions against their tormentors, or they may identify potential bias and overcorrect for it: “If as a referee you walk out to a hostile crowd, especially if you've got history with the club, by human nature you're not exactly going to bend over backwards to help them. Poor little Bolton Wanderers will get the 50-50s because the nasty Man United fans are singing songs about me”. More subtly, the human limitations of referees make it likely that they are not certain about every decision. In these circumstances, the reaction of the crowd may tip them one way or another. For example, the same tackle may provoke howls of outrage if performed on a home player, but attract little attention if carried out by a home player. Thus the referee, who may not properly have seen the incident, may be influenced by the reactions of the spectators.
Comprehensive statistical analyses show that refereeing decisions do in fact favour home teams, granting them more penalties and fewer red cards. Nevill et al carried out an experiment, showing two groups of referees the same match footage, but one set got to hear the partisan crowd noise while the others watched the game without it. As expected, the referees exposed to the noise were more lenient to the home team.
Another apparent benefit of playing at home is ‘local knowledge’. This is clearly important in cricket, where the actual playing surface plays such an important part in the game. Home cricket teams are likely to include players suited to the pitch, for instance spinners on dryer wickets. Further, many cricket teams prepare their pitches to suit their playing style. But football pitches are far more standardised and less significant to the game. There are small variations – larger pitches encourage more expansive play. Yet it seems unlikely that away teams are particularly disadvantaged on others’ turf.
More plausible is the theory that away players are discomfited by travel. This helps account for the falling significance of playing at home, as long and arduous coach rides have been replaced by rapid plane journeys. Last week’s round of fixtures in the Spanish La Liga offers further evidence that difficult journeys undermine away teams: with all Spanish aeroplanes grounded by strikes, teams were unable to fly. Result: of the away teams, only Barcelona won (and they’re so much better than anybody else they could probably have walked from Barcelona and not been fazed). Similarly, away teams seemed to struggle in European competition during the Icelandic volcano disruption earlier this year.

Yet while this clearly suggests that uncomfortable coach trips inhibit performance (though admittedly I’ve only given a small amount evidence here), it still doesn’t account for the effect on players flying and travelling well in advance. Sure sleeping in a strange bed in unfamiliar surroundings might make players less relaxed than usual, but enough to give home teams such a boost?
The psychologists Nick Neave and Sandy Wolfson have posited that physiological factors can help explain home advantage. They see it as a version of ‘territoriality’, the phenomenon by which animals are more vigorous and motivated in defending what they see as their own space. Territoriality is generally believed to result from higher testosterone levels in animals who perceive their home too be under threat. Testing footballers’ saliva for the hormone, they found that home players produced far more testosterone. This difference was not reflected in self-reported questionnaires on the players’ mood, suggesting that the process is not a conscious one. This finding is extremely interesting, but comes from a small-scale study which does not appear to have been repeated or scaled-up. In any case, the link between testosterone and on-field performance isn’t clear-cut. For example, excess testosterone could lead to over-aggression: home teams being ill-disciplined and losing focus.
Of course, it is highly possible that home advantage is greatly psychosomatic. Perhaps away teams play worse because they expect to play worse. They may be more satisfied with worse results because the concept of home advantage is so ingrained. Thus when big teams go to play, say, Birmingham City (who have a very strong home record), they will be more likely to settle for a draw instead of exerting themselves for the win they would expect at home.

This is especially apparent in teams’ tactics. It is generally expected in most games that the home team will have more possession and be more attacking. So ingrained are these tactical mindsets that ‘to play like an away team’ is often used as shorthand for a defensive outlook. It is highly probable that a significant reason for away teams’ lack of success is this concession of the tactical initiative, their diminished expectations from the start.
If my time at university has had any effect on me, it is to instil an inveterate dislike of the ‘list a load of points with pros and cons and conclude the truth lies somewhere in the middle’ approach to issues. However, on the question of home advantage, I’m not sure there’s much else I can do. There are a number of plausible mechanisms, but none of them are entirely persuasive and none of them seem to exclude the others. In any case, the sheer magnitude of the effect seems to suggest there are multiple factors in home advantage. So I still can’t say I know why Liverpool aren’t top of the table.