Friday 6 June 2008

The Glicko Bounce

When we lose rating points, the search for a culprit is often the most important thing. Sometimes we might accept that we have just played badly, but often it is due to under rated juniors, or illness, or the ratings system itself. Of course when we gain ratings points it is mainly due to our improved knowledge of the game, or finally playing to our real strength. But rarely the ratings system itself.
Now realising that the plural of anecdote is NOT data, I have one example where the rating system is responsible for some rating shifts, and not just in a bad way.
A chess playing friend of mine who I'll call M (he reads this blog BTW), made a comeback to long time control chess towards the end of last year. He had a rating in the mid 1400's but his first tournament (6 games) was a bit of a shocker (2/6) and his rating dropped over 500 points, to the low 900's. Now given the comments I've seen posted by others elsewhere this is grounds for (a) never playing chess again and (b) having the ratings officer charged with crimes against humanity. Of course player M did neither, but simply kept playing chess. And in the latest rating period he played another 5 games, performed a lot better, and saw his rating jump up 500 points, so that he is only 40 points short of where he was before his first tournament back.
Of course he is still on the list with a ? (unreliable rating) next to his name, and his rating may bounce around a bit more, but at least in this case it demonstrates something important.
If you want your rating to reflect the level at which you play chess, you actually have to play chess.

No comments: