Friday, May 9, 2008

"My sh*t doesn't work in Twenty20 Test Matches!"

A couple of weeks ago the Guardian ran a story about Billy Beane's interest in applying sabermetric principles to soccer. Now the Guardian's Sport Blog asks whether cricket wouldn't make a better fit, and points to a blog devoted to cricket stats called Pappus' Plane. There's even a cricket Moneyball (sorta) called Best of the Best, in which author Charles Davis devotes some space to debunking cricket's versions of clutch hitting, team chemistry and all of that:

It is fascinating reading (for a cricket fan). Amongst other things, Davis objectively proves that using a nightwatchman is fundamentally flawed (you can read his analysis here). Ultimately though the book led me to think that there is a third major factor hindering cricket sabermetrics. Cricket is excessively obsessed with its past, and the majority of Davis's book is spent comparing players from different eras and trying to determine who is best. Which is all good fun, but it means that the statistical innovations he makes - such as the calculation of an 'under-pressure average' for batsmen - are squandered on pub-table debate. What Billy Beane did - by contrast - was to take such stats and actually apply them to team training and selection.
I guess the advantage to cricket's version of sabermetrics starting in the pubs instead of in mothers' basements is that when one of their statgeeks actually starts making decisions about teams -- which is inevitable -- it won't be met with as much resistance as a result of general familiarity.

Neat sporting stuff if you're into cricket. Neat sociological stuff if you're not.

(Thanks to Laurence Davison for the link)

No comments: