Michael Aigner sent me a message via Facebook that explains the conundrum of Paul Richter’s early rating spike to 2211. There were two pieces of information that I didn’t know.
- Paul was not only undefeated (4-0) in his second tournament, he was also undefeated (5-0) in his first tournament. So he started his chess career with a nine-game winning streak. Wow!
- When rating an undefeated person, the USCF doesn’t use the usual formula — for roughly the reason that I sketched out in my last entry. Instead they just give him/her the rating of the highest person he/she has beaten, plus 400 points. In Paul’s case, he had beaten an 1811 player, so he was given a 2211 rating. The earlier rating of 1373 came about the same way — in his first tournament the highest player he had beaten was rated 973.
Now we know how he got that rating. The question still remains whether it makes sense to rate a provisional player that way. I’m sure that the USCF rating committee had many discussions before instituting the policy, and considered every scenario. I still think that they made the wrong choice. It’s inconsistent. They are drawing a distinction between someone who starts their career with an 8½-½ record and someone who starts with a 9-0 record, and the difference is HUGE. As I discussed in my last post, Paul’s rating would have been in the 1600s if computed the normal way, but with this top-victim-plus-400 way, his rating was in the 2200’s.
In the long run, of course, it doesn’t matter. Paul’s rating eventually found the right level. But it must be a little bit weird to “make master” at age sixteen… but really you’re just coming back to where your rating was at age ten. Did he get a master certificate when he was ten? Will he get one now, or will the computer say he was already rated 2211 earlier, so he doesn’t get one?
On a side note, I want to point out how extremely impressive Paul’s progress has been this year. In his last six tournaments, dating back to February, he has gone 2066 -> 2074 -> 2082 -> 2146 -> 2187 -> 2191 -> 2271. Six straight improvements, some of them huge, adding up to more than a 200-point jump in the last seven months. He probably would have gone over 2200 in his last tournament if I hadn’t beaten him, but that just set him up to make his master rating in epic fashion at the CalChess State Championship.
Addendum (4 hours later): Michael provided this link that officially confirms that Paul Richter’s rating leap of 838 points in one tournament in 2008 is the largest in USCF records dating back to 2004. The second-largest leap any player has had in one tournament was “only” 508 points.
Nevertheless, you can probably tell from what I’ve written that I consider Paul’s 80-point jump this weekend and his 200-point jump over the course of six tournaments this year to be a more remarkable feat than his record 800-point jump. That’s because his jumps this year are not an artifact of the rating system, but represent bona fide improvement.
