When letting players know about analytics is a good thing
When is it appropriate for a club to let footballers in on the team’s analytics research?
This question came up in a conference here in Toronto last March in which I was a panelist. Other speakers included the director of analytics for the NBA Jason Rosenfeld and assistant GM for the New York Giants Kevin Abrams—in other words, people who generally knew whereof they spoke. Someone on the panel raised the familiar problem of players “running up the numbers”: by informing players of favourable metrics—sprint speed perhaps, or shot volume—you run the risk they will attempt to artificially inflate them to improve their prospects.
There are some delightful examples of this in football. Blake Wooster, co-founder of 21st Club, recalled to me the story of David Ginola’s time at Aston Villa, a player who, in the early days of Prozone, was scolded by his then manager John Gregory on his failure to run higher distances (though Ginola was wise enough to know that intensity, not quantity, was the key in sprints). “Even David James, our GK, runs more than you!” Gregory apparently told Ginola. So, naturally, whilst the ball was on the other side of the pitch, the keeper James would jog back and forth along the touchline to ensure Ginola would get another earful from the gaffer!
There are other issues with letting players in on their stats which are even more problematic. Some teams—perhaps to justify hefty investment in the latest “cutting edge” player-tracking technology—may simply hand over stats sheets to players without any additional information over whether the numbers actually reflect areas in need of improvement. The player may therefore decide for themselves what they need to work on and, in doing so, try to fix what isn’t broken.
As the saying goes, a little knowledge is a dangerous thing.
At the time I broadly agreed with the panel—one should exercise extreme caution when informing players on any statistical trends. However, after some reading this past week I’ve come to believe there are a few instances when letting players in on club analytics might be very beneficial.
Consider, for example, the Major League Soccer club Toronto FC. On account of renovations to their home stadium in Canada, the team is currently halfway through an eight game road trip to start the season. Perhaps predictably, the club is struggling. Despite some high profile off-season additions including US international Jozy Altidore and former Juventus and Italy forward Sebastian Giovinco, the club has so far won one match and lost three.
For someone steeped in football analytics, these very early results shouldn’t be cause for alarm. Though Toronto’s TSR is very low (second lowest in the Eastern Conference), this has to be considered in light of a difficult away schedule. This effect is statistically significant; this past week at StatsBomb for example, James Yorke wrote on the away games in the Premier League, concluding that “…before a ball is even kicked, an away team can probably expect to have to generate around two shots extra per game to create a par outcome.”
As for the temptation to start into discussions over whether the team is “meshing” or not, it’s still far too early to tell. Finally, one doesn’t need to be an analyst to know that in MLS, losing eight consecutive matches well into the season won’t necessarily exclude you from a Cup final berth.
It’s possible that some TFC players are aware of this, if not explicitly then certainly through experience. Chances are however that most footballers read about themselves and their team in tabloids, not stats blogs, and are therefore more inclined to think like beat reporters for whom three consecutive losses definitely signifies a worrying long term trend.
For some footballers, this may not be a bad thing. A little introspection and doubt can lead to a healthy atmosphere of accountability and self-improvement. For others however—and Toronto FC has a history here—constant worry over results it can lead to factionalism, in-fighting and panic. For both footballers and first team coaches, it can lead to solutions in search of a problem.
A little education here might therefore go a long way, and allow players to focus more on overall process than living or dying on every single game. This isn’t to say that stats should be used as an excuse for poor results, or worse, a reason to not work as hard when on the road. If anything, the data on away games could be a challenge to players to work harder to overcome what is essentially a statistically stacked deck. It can help teams avoid needless bouts of low confidence which, if left unchecked, can start to affect team morale.
There are other areas where knowledge of statistical trends can be useful, too. For example, though dismissed by some analysts as spurious, knowledge of team behaviours in various game states (whether the team is ahead or behind a goal) can help teams chasing or defending a goal to be more self-aware.
Ironically, the field which allegedly discounts the importance of grit and self-confidence in football may be a critical tool for shoring up both.