Hectic xmas schedule gives some teams an artificial advantage
Earlier this week West Brom made an official complaint to the Premier League regarding the scheduling of their match against West Ham, which took place only 48 hours days after their 1-1 draw against Arsenal. The root of their concern is not that their players had only one full day to recover, but that West Ham did not have the same issue. The Hammers' previous match was on Boxing Day, giving their players an extra 5 days of rest. Despite taking a first half lead West Brom went on to lose the match 2-1.
West Brom clearly feel that five additional days of rest may have given West Ham an edge in what was already an important fixture in the relegation battle. Is there hard evidence to support this? Using historical results, can we demonstrate statistically that one team is at a disadvantage when they play twice in a short period of time and their (second) opponent does not?
Is one full day of rest enough?
I searched for every professional league match in England in which one team were playing their second match in three days against an opponent that had enjoyed at least one extra day of rest. Since the 1995/96 there have been 482 such matches: 54 in the Premier League, 175 in the Championship (or the old Division 1), 137 in League 1 and 116 in League 2. Most of these matches took place either over the Christmas period or in April, as the season draws to a close.
To determine whether the teams were at a disadvantage, I compared their win percentage in those matches against their win percentage in the return fixture that season (which actually was normally earlier in the season). In the return fixture both sides had at least two full days of rest, so this comparison should isolate the impact of one team having very limited recovery time. Of course, home advantage could also be a factor, but that should average out over the entire sample.
Table 1 shows the results. The team that had only one full day of rest – labeled ‘Team A’ – won 29% of the matches, lost 39% and drew the remaining 31%. However, in the return fixture they won 36%, losing 36% and drawing 28%. So Team A won 7% fewer matches when they had had substantially less recovery time than their opponents. This difference may seem fairly small, but it’s highly significant (in a statistical sense): Team A were at a clear disadvantage.
What about longer recovery periods? I repeated the analysis looking at matches between teams that had had two full days of rest since their previous match (e.g, playing Sunday-Wednesday) against opponents that had had at least 3 days of rest (Saturday-Wednesday). This time I found no evidence of any disadvantage: there was no significant difference between the win percentage in that match and the return game the same season. This implies that two full days of rest is generally sufficient for players to maintain their performance levels in consecutive matches (although presumably that is not sustainable in the long-run).
I’ve provided evidence, using data, that when a team plays their second match in 48 hours against an opponent that had at least 24 hours more rest, they are at a small but significant disadvantage. If the Premier League and FA want to ensure that the fixture schedule does not favour some clubs they should either avoid making teams play twice in 48 hours, or – if it that is not possible – schedule the match so that both teams have exactly the same amount of time to recover from their previous match.
Of course, if the Premier League's priority to is to please the companies that pay to broadcast matches live, all this will be very much a secondary concern.