What are all these numbers for anyway?
Anderson is co-author, along with David Sally, of The Numbers Game: Why Everything You Know About Football Is Wrong, and he was an obvious man to speak to about the acceptance — or not — of statistics in a game that doesn’t traditionally rely on that kind of data.
Soccer is his frame of reference, but the lessons are easily applicable to Gaelic football and hurling.
Take the most basic starting point: is agreement on what statistics are — what constitutes a tackle, a wide, a possession — the essential first step to processing data about sport?
“Yes and no,” said Anderson. “If you’re a club or a team then what you want is different to what you want as a media outlet.
“If you’re a media outlet you need to ensure you’re comparing apples and apples — you need a commonly agreed upon set of indicators, so you can say ‘team A did this so many times compared to team B’ and everyone knows what you mean by ‘this’.
“A club or team is different because its manager or coach may say, ‘these stats are what matters to us and this is how we define it’. He’s trying to give feedback to players and to find out why his team is winning or losing.
“The best way to approach it is to go from easy to difficult. In football we all know what a goal is: it’s either in or it’s out. We can count those, and we can count free-kicks and so forth. It gets complicated quickly. Counting shots can be difficult because sometimes there isn’t agreement on what constitutes a shot and what doesn’t.
“What both teams and outlets are basically trying to find out is why one team won and another lost — the why drives us, not the what.”
The fons et origo of all nerdish stat-hunting is Michael Lewis’ book Moneyball, of course, about bright baseball executives finding a key statistic which allows them to purchase underrated players.
That ‘Moneyball’ approach isn’t always applicable in team sports, says Anderson.
“The fundamental difference between baseball and a game like soccer is in the number of participants. You’re talking 11 versus 11, while in baseball it’s a one versus one situation, discrete, easily measurable series of events: someone throws the ball and somebody hits it. End of story.
“Football’s a different animal. In the average Premier League game Opta collect approximately 3,000 different events, and maybe 200 to 250 different types of event. That’s far more complex, obviously.
“However, that’s not to say that certain aspects of the game can’t be broken down to be a little bit more like a static game such as baseball. There are set-pieces which are more rehearsed and quantifiable.”
Anderson makes a persuasive case for consulting the numbers even if they’re not as immediately potent in open team games as in baseball.
“The fact that you can’t do everything doesn’t mean that you shouldn’t do anything. Doctors haven’t cured cancer but that doesn’t mean they don’t try to understand aspects of the body which lend themselves to healing.
“The suggestion that ‘football isn’t like baseball so we shouldn’t analyse it like that’ isn’t really on point. It’s also true, though, that baseball has a lengthy history of talking numbers, it’s around since the 1880s, and everybody’s at-bats, pitches and so on have been available for that length of time.
“In soccer you’ve gone from zero to 100 in terms of information on the ground. We used to have goals, cards, substitutions, etc, but now every second of every game is logged — where players are on the pitch, where they’re going, how fast they’re going, what distance they’ve run.
“That’s a huge amount of information and has come upon people in a very short period of time. It’s not surprising that people in football are still trying to figure that out, and they will, but that’s not the same as saying that football will be ‘solved’ in the next few years.”
Okay, if 11 versus 11 throws up 3,000 events under 200-250 different headings, then make what you want of two 15-on-15 games.
We’ll come back to this, but one of the significant conclusions to be drawn from Anderson’s discussion is that ‘solving’ the sport isn’t the end goal with the stats.
They can help, but there’s no magic bullet involved.
