Understanding Chess Ratings

… And Understanding the Problem with Chess Ratings

Chess is often thought of as a sport, albeit a cognitive sport. It is a game, but a brutal game, with furious exercise of all mental faculties (and incidentally some physical stamina as well).

As with all competitions, we not only want to know who is best, we want to compare ourselves to our opponents, and watch as an underdog International Master takes on a Grandmaster at the first table.

In chess, the competition and comparison is all about the rating, or the Elo as it is often referred to.

I think this is a mistake, but more on that later. First let’s understand it.

I don’t discuss gender gaps in this article, but notice that the real disparity is in age, far more than actual rating distribution.

Elo is not an abbreviation. It was created by Arpad Elo, a Hungarian-American chess master. The central idea is that wins and losses alone are meaningless, because some wins are more impressive than others. But we can’t know which wins are more impressive based on some quantification of moves… or at least we couldn’t sixty years ago. (With computer supremacy and the quantification of every move and position, that might be possible today! But that’s a topic for another occasion.)

Elo, also a physics professor and competent mathematician, devised an elegant mathematical solution that was easy to implement, even before…