Moneyball: Not Round Here, Mate

I was tempted to call this piece “Moneyball: moneybollocks!” I haven’t seen the film starring Brad Pitt, who I only mention for the purpose of search engines, but the concept is simple enough: beat your richer opposition by being smarter with your resources. Well I think we all agree that Arsène Wenger does that – or at least was doing that, when he was winning trophies. He did have the advantage of inheriting pretty much the best back five unit possibly ever seen in football, plus Dennis Bergkamp, a man who is perhaps unfortunate only to have been voted third best player in the world. But even allowing for that, AW has a deserved reputation as a manager who can pick up players (relatively) cheaply and make them great. Anelka of course stands out in terms of profit, but for me the brilliant Thierry Henry is the best example of turning a player others had lost faith in into a superstar. (I love Thierry.)

Some say the latest example of ‘Wengerball’ is Gervinho. In fact Arsenal fan @johncrossmirror has written twice this week about how The Forehead is taking the Premier League by storm, and is another example of Billy Beane’s brilliant theories brilliantly put into action by Wenger. John Cross writes that Gervinho’s stats are all important, and he has “played nine Premier League games so far, scored twice, made five assists and created a goalscoring chance every 39 minutes.” Well, yeah, but he’s also missed goalscoring chances more often than that. Like the one he tried to flick through his legs against Norwich, where luckily Van Persie saved his blushes by being right behind him. That kind of thing doesn’t often get counted in football stats because Arsenal scored anyway, but in baseball there’s no second batter standing behind, you’re on your own pal.

Gerv: Statistically good in France

John also wrote: “But Gervinho looks as if the ball is running away from his toe. He sometimes looks as if he will flatter to deceive with his quick and flashy twists and turns. Yet his finishing can be wasteful and his end product can seem inconsistent.” Well maybe he looks as if the ball is running away from his toe because the ball is running away from his toe. And his finishing is wasteful and his end product is inconsistent.

My point is that The Forehead is hardly an unqualified success as yet, so doesn’t make the best example of proving how good Billy Beane, Wenger, Moneyball or even Brad Pitt are. Having said that, I do have some faith in Wenger’s judgement of players and by all accounts Gerv did have a brilliant season last year. So he may well come good, as Koscielny certainly has so far this season.

But is Gervinho’s purchase anything to do with Moneyball theories? Baseball is a sport that has much more meaningful individual statistics than football, for the simple reason that there is far less randomness. One guy stands in one place on a mound, another guy stands several yards away and tries to hit a ball thrown by the first guy. They’re always in the same place, the ball comes in the same kind of direction and height, at the same kind of speed. No one tries to put in a sliding tackle while they’re doing it. Each play lasts a few seconds, so randomness barely enters the equation.

You can measure all kinds of thing in football, of course, and these days people do. Arteta has been criticised for lack of effectiveness, and passing sideways too much. There are probably stats to prove it. But maybe every time he passes sideways it’s because it’s the best ball to play at the time. You can’t tell until the play finishes and a whole series of essentially random or chaotic events have been played out by 22 people, any or all of whom can influence the outcome. Not so in baseball: one guy throws, one guy hits, maybe one guy catches. It’s all far more predictable than the Brownian motion of a football match.

Torres: Look at those stats!

Another example: Fernando Torres is so good that Chelsea paid £50m for him. A fantastic goalscorer. His stats at Liverpool were great. Trouble is, you put him in a different system with different players at Chelsea, and look at the stats now. Who would pay even £5m for him if the Chelsea form was all you had to go on?

So I don’t think Moneyball statistical theories really work in football. Sure we all want the player who can score 30 goals a season, but that isn’t guaranteed, as Torres proves. But in baseball a pitcher can carry on doing exactly the same thing in the same way in any team.

In that case what is it that Wenger is doing? Because he clearly does beat nearly all managers in terms of getting value for money, as evidenced by the miniscule Arsenal net transfer spend over his whole reign, not to mention that our record signings figures have not been in the same ballpark as our main rivals since we paid £333,333 for Malcolm MacDonald about 35 years ago. What Wenger’s good at is the other aspect of Moneyball: spotting potential in players before they reach maturity, so that you gain value by buying them. It’s very difficult to be sure that a brilliant 16 year old is going to be a brilliant 24 year old – not that it stops clubs like Chelsea paying millions for 16 year olds. If you’re clever enough, though, and have the right knowledge and instinct, you can pick the right ones at 20 or 21, when they’re still on the way up. Wenger does that better than any other manager. He also does it more often, and in recent years at the expense of signing anyone else at all. This would be fine if all the players reached maturity of course, but the buggers keep leaving and we end up where we started: fourth in the League and with a squad full of 21 year olds. (Until the panicky end of August, of course, when rational thought went out of the window on the last day of the, er, window.)

The other thing Wenger does is pick players who are reaching maturity but are still cheap. He does this mainly by buying them from France, where of course he has intimate knowledge of a league that is the fifth best in Europe but where prices are far lower than the biggest leagues (though PSG are doing a Citeh, so that might start to change). He still has to have the talent to work out which players will adapt to a higher standard and tougher games in the Premier League, as well as adapting to his team’s style, but that surely is nothing to do with Moneyball, that’s just being a good football manager.

So does Moneyball apply to football? Not really. Statistics are great in their place, but they don’t replace good judgement, man-management and the ability to make something that is greater than the sum of its parts.

About these ads

5 thoughts on “Moneyball: Not Round Here, Mate

  1. Give Gervinho time. Thierry, Dennis and Robert P didn’t score many early days at Arsenal. Gervinho has been putting in a shift, getting into positions to score and laying on goals. Support him.

  2. Well, if you read the book,
    “”””So does Moneyball apply to football? Not really. Statistics are great in their place, but they don’t replace good judgement, man-management and the ability to make something that is greater than the sum of its parts.””””
    is precisely what was suggested for baseball in the first place. (the movie abridges that part somewhat).

    It is simple really. The ideal statistics (or even the ideal statistical trends) for baseball, will not be the same as for football.
    If you read the book, you will realise that the ideal sabermetrics stats for baseball (Bill James’ etc) were languishing unloved for a while (mid-80′s to early ’00s when the Oakland A’s with Beane and DePodesto picked them up).
    Maybe these ideal stats for our game have not yet been built or looked at yet. Maybe there are better stats there than the goal assist KMs run passes made key contribution OptaStats, and theyhave not been found out yet. I don’t know of OxBridge math/stats geniuses who have looked at football stats, and found out new trends (like they did with baseball in the US)…. It does make us non-genius ‘good judgment’ ‘assessment’ guys a little insecure (like it made the Oakland A’s chief scouts, in the book) , but it does not make it wrong.

    • I just think statistics don’t have such a place in sports that feature essentially random and chaotic sequences of play that can last for several minutes. Even things like goal assists are meaningless unless you also measure where everyone else on the field was and what they were doing at the time. In baseball you know where everyone else was and what they were doing.

    • You’re right! Good spot – no one else has mentioned it. Maybe they couldn’t be bothered, or maybe you’re cleverer than them. I’ll change it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s