Sure Barack Obama won the Presidential election last year. But he wasn't the only big winner. Nate Silver, the number-cruncher behind FiveThirtyEight.com was another victor, having predicted the result with uncanny accuracy. Silver is a sabermetrician, which is to say that he began his public life as an analyst for the brilliant Baseball Prospectus years before he brought his statistical nous to politics.
Using regression analyses, among other tools, to predict political outcomes is one thing; trying to create a predictive rankings system for international football is quite another. By "quite another" I mean vastly more difficult. Nonetheless, Silver has attempted this. So, as a rival to FIFA's rankings (which are, as he admits, less useless than was once the case) Silver has come up with his own football league table. Unlike other such enterprises, however, they are also designed to be predictive as well as reflective.
According to Silver, England are the third best team in the world. His top ten:
Silver explains the project here and, in more detail, his methodolgy here. (Warning, there are some head-wrecking numbers at that latter link.) Right now, in other words, Silver thinks England good enough to reach a World Cup semi-final.
But there seem a few obvious problems with this. For one thing, an algorithm that a) values home advantage and b) also takes account of margin of victory might seem ill-suited to a competition such as the World Cup Finals in which 31 of 32 teams play one another on neutral territory and in which, most of the time, winning 1-0 is as useful as winning by three or four goals. To be fair, Silver acknowleges somthing of this and even though he rates Italy as just the 12th best team, I doubt he'd really pretend that Chile are better than the Italians. Or, to put it differently, Chile's recent performances may be more impressive than Italy's but that does not mean they're more likely to reach the WCF quarter-finals than Italy.
And there's a big missing ingredient in Silver's methodology anyway: the Manager. You can make a case, unpersuasively in my view, that managers don't much matter in club football since resources - ie, money - are the single most determining factor in club football. Have enough money and buy enough obviously good playes and, QED, success is yours. Well, sometimes. But international footbll isn;t like that.
So while Argentina obviously have enormous resources of talent at their disposal, how that talent is used matters enormously. There's little sign that Maradona really has a clue. So, while Argentina might be a top five side in terms of talent they're nowhere near that as a team right now.
Managerial influence is most obvious in the international game. Think of Bora Milutinovic or Guus Hiddink or Luiz Scolari if you doubt that. For that matter, I doubt that any ranking system would have predicted that either Denmark or Greece might become Champions of Europe.
This isn't a knock on Nate Silver, rather a suspicion that, useful though statistical analysis is, there are limits to it. In oter words, his new venture is interesting and useful but hardly the last word.... (Not that I think he thinks it could or should be.)
Still, I want to know what Danny Finkelstein has to say about all this...
[Hat-tip: League of Ordinary Gentlemen]