Much to my surprise, quite a few people were surprised when King’s XI Punjab beat the Mumbai Indians yesterday in the IPL. They shouldn’t have been. At the halfway stage of the tournament, I predicted to a friend that Mumbai, then leading the league, would do worse in the second half than in the first, and Punjab, then at the bottom, would do better. And so it’s turned out. My reason for believing this had nothing to do with any deep cricketing insight, but with a simple statistical phenomenon called “regression to the mean.”
The teams in the IPL are more or less evenly matched: they have a similar mix of overseas players, national stars and domestic players. (The salary caps ensure that this will continue.) And the format, being just 20 overs a side, that ensures that chance events play a much greater role than in other formats. For these reasons, I don’t believe that any team can ever truly dominate the league—unless they have a phenomenally lucky season, which will even out in the long run—or be too far below the rest. While in the short run the game is unpredictable, in the long run everyone’s going to be bunched around the mean.
So while I’m wary of predictions about specific results in the IPL, I’ll be glad to make a general wager on IPL 4. I’m willing to bet that the team that tops the league at the end of the first half will do worse in the second; and the other way around for the team that comes last. I have absolutely no idea, of course, which those teams might be.
As it happens, I would not make a similar bet for the EPL, where neither of my two conditions apply. (ie, teams are not evenly matched, and there is a far greater premium on sheer skill.) Is that a good thing or a bad thing for the IPL?
Regression to the mean also explains the Sports Illustrated Cover Jinx, by the way.
I don’t mean to say that matches at the IPL are decided purely by chance. There is immense skill involved, and I love the contest between bat and ball that we get to see every day. But the skill is so evenly distributed among the teams that in the long run, it evens out. The X factor that a captain like Shane Warne brings to the game does count for something—but while the Rajasthan Royals won IPL 1 under him resoundingly, in IPL 2 they won six out of 14 games. Such it goes.