Chess rating system for other sports?

Form means nothing
By Daniel Finkelstein Special Correspondent
April 06, 2008

YOUR team has lost. Again. It’s a disaster. Or is it?

When you lose games, you lose points. That much is clear. But how seriously should you take defeat and how heartened should you be by victory?

That all depends on your view of form.

It’s impossible to escape talk of winning streaks and form in football. It’s on every commentator’s lips.

The idea is that when a striker scores in three games, he is more likely to score in the next game. He has what they call in basketball a ‘hot hand’.

Confusingly, there is an alternative view of form.

If a team wins a whole rash of games, isn’t it more likely to lose the next one as the luck ends?

The reason for this confusion isn’t too hard to discover.

There are copious numbers of academic papers that show that form is a myth. Winning one game makes you neither more or less likely to win the next one.

If a striker scores in a game that makes him neither more nor less likely to score in the next game.

The only sport that shows evidence of a ‘hot hand’ effect is ten-pin bowling. And even that disappears almost before you notice it.

So if form isn’t real how do you know whether you’re team is getting better or worse or whether you are just seeing luck play itself out?

Making use of the chess rating system designed by Hungarian American mathematician Arpad Elo, and Mark Glickman’s work on parameter estimation, we have re-engineered our computer model.

This allowed us to update ratings for each side only by knowing their own result, rather than, as in our full model, knowing the result of every team in Europe for the last five years.

It immediately became apparent how slowly each sides ratings changed.

Here is the full article.