The stat is more powerful than the sword

While at Fulham for half a season last year, professional centre back and source of hilarity Phillipe Senderos completed more assists (1) than Ross Barkley managed all season at Everton (0). Therefore Senderos would make a much better attacking midfielder than Barkley, who must himself be worth very little to any potential suitors. Now that I’ve ascertained this using nothing less than cold hard fact, let’s move onto the topic at hand; the merits and value of statistics in football.

Let’s start by asking what a football statistic is. For the most part they are just the result of counting something – passes, shots, tackles – then attributing them to a player or side. You can’t remember every time Andros Townsend takes a shot from outside the box because you don’t have a super-computer for a brain, so instead Opta do it for you.

Back in the olden days when teams like Nottingham Forest were good and teams like Wimbledon existed, TV coverage just showed you the score and you had to use your own stupid brain to try to work out which team was making the better go of it. Now however, it has been commonplace for years for TV networks to include possession and shot totals after and even during games, and this has evolved into punditry keen to show smaller, more precise stats like pass completion rates.

Before I embark on why I think stats sometimes get a hard time, I’d like to bring you some things I hate about stats. Some people can commit stats related faux pas so bad they make me want to let Andy Reid loose in their kitchens just before Christmas dinner.

[linequote]TV coverage just showed you the score and you had to use your own stupid brain to try to work out which team was making the better go of it[/linequote]

I hate that guy who uses the aforementioned possession and shot stats to justify anything at all. He might try to find the better team in a draw, make a thrashing seem less bad than it was or compare two entirely separate thrashings to find a more dominant thrasher. “Well you did beat us 6-0 but we only had 8% less possession than you and the shots were pretty similar” he might say. His side conceded six goals and failed to score a single one.

Remember, when Scott Parker decides against a risky pass, pivots and plays a short, attack killing pass to the left back; his side keep possession. When Andros Townsend ignores the run of a striker, opting instead to concuss a poor supporter 30 rows back; Tottenham’s shot count rises by 1.

I also hate that guy who uses statistics to decide how good a player is, and thereby rank and compare him with other players. I’ll touch on this more later, but it’s stupid and you should stop.

Remember that some finishers play with better creators and some creators play with better finishers. Remember that football is about more than goals and assists. Remember that different clubs employ different tactics and the usefulness of certain attributes and tendencies varies club to club and opposition to opposition. Please don’t tell me it’s staggering that big clubs are not interested in some 26 year old, uncapped winger in Serie A because he completes 4 dribbles and 3.2 key passes per 90 minutes.

I hate that Fantasy Football freak who’s dead keen to find out who got the assist whenever a goal is scored. You’re doing football wrong.

You might now be thinking that I’m of the view that there’s very little value in statistics but that’s not really true. You’ll note that each of my previous gripes is not with any actual statistic but with how it’s used.

There’s an old adage that goes something like ‘there’s no such thing as a bad statistic, just bad statisticians” and that’s kind of right. The stat isn’t trying to argue or imply anything, it’s just a number. Only the person using that stat with a particular bias in an argument is wrong.

[linequote]A statistic’s real strength comes when you disassociate it from ability and rather see it as an indicator of tendency and behaviour[/linequote]

I think there are a lot of straw man arguments made against them that don’t really do them justice, such as:

“There’s no statistic that can represent the beauty of a perfectly executed piece of skill“.

Well no, there isn’t. But there is no claim that there is, and the mere existence of statistics shouldn’t diminish anyone’s appreciation of such pieces of skill. No-one thinks that they should.

In my opinion, the true usefulness in statistics requires not seeing them as any sort of measure of ability. Because tactics and gameplans vary so often, it is difficult to decide whether or not having, say, a player shooting a lot is a good thing or not. It might be because the player is excellent at getting into scoring positions, it might be because the player just fires goalwards constantly even when there are better options to pass to, or it might just imply that the player’s teammates are so good that he’s constantly given shooting opportunities.

A statistic’s real strength comes when you disassociate it from ability and rather see it as an indicator of tendency and behaviour. Having a player that has lots of shots isn’t a good thing and it isn’t a bad thing, all you can gain from that stat is that he shoots a lot. If you want to know more about a player you’ve seen little of, you can use his stats to form a picture of what kind of player he is and what his role appears to be.

Does he play a lot of passes? Are they attacking passes? Does he thread through balls? Shoot a lot? Tend to dribble past players?

What begins to emerge is a picture of what a player does, rather than how good he is.

And let’s be frank, it doesn’t matter how many games you’ve been to, or how infrequently you miss matches on TV or how many sleepless Saturday nights you’ve spent obsessively watching streams of the Colombian second division while your mates are out getting laid, you can’t have examined every great detail of every player on the pitch every time. It’s not possible.

[linequote]TV has really tried to jump on the statistics bandwagon in a lazy attempt to appear modern and ahead of the curve[/linequote]

It’s not even possible to know everything about every player in the Premier League. What does Morgan Schneiderlin do? You don’t need to act like you have a season ticket at St Mary’s (unless you do in which case I’m so sorry for you) and are an expert on him. I can tell you at a glance he makes lots of tackles and interceptions, he passes neatly and isn’t dispossessed all that often.

In sports where stats are much more predominant, like baseball or cricket, stats can be used to show ability much more easily because everything is normalized and ‘turn based’. Ball is pitched/bowled, reset, repeat. All batters can be judged against all others because their tasks are nearly always the same – hit the ball. In football, 22 players move unpredictably around the field for 45+ minutes. Their jobs cannot be standardized because they’re so often changing, and can depend vehemently on what the manager wants the team to do.

TV has really tried to jump on the statistics bandwagon in a lazy attempt to appear modern and ahead of the curve, in what sadly ends up resembling a politician telling an interviewer how much he likes Dizzee Rascal.

Television has done the progress of stats more harm than good. Imagine your interest being spending huge amounts of time on statistical analysis and endlessly searching for ways to perfect it and its uses, only for the media to present your passion to the public through Robbie Savage looking at an Antonio Valencia heat map and pointing out he largely played on the right.

As for playing Phillipe Senderos being a better attacking midfielder than Ross Barkley? Well, that’s another thing stats are quite good for – finding quirky little anomalies and using them to wind up people who really care about how their players are perceived. Who really cares about stupid sh*t like that?

Woolwich fans. Did you know that Nacer Chadli created more chances and key passes per 90 minutes than Alexis Sanchez last year?

Spread the word.