I am firmly in the camp that believes the top player's ratings are inflated because they tend to avoid playing in Opens (Kramnik played in his first open in something like 20 years just within the last year [I believe it was at last year's edition of the Qatar Masters. - ed.]), and this is an example of what can happen when the top dogs move to the shallow end of the pool.I meant to disagree (mildly) at the time, but forgot with all of the holiday hullabaloo. I do think the top players would probably get nicked sometimes, but I also think the top players are the top players because they really do perform that well. It hasn't been unknown for a player to soar to elite ratings status only to get knocked back down immediately upon getting into elite events.
These results are rare, but when the opportunity is present, the top guys will get nicked enough (a draw is a rating loss) to keep rating at a more accurate level.
Mostly, I think the top players would have to adjust to the style of an open Swiss tournament by learning to take more chances. This would increase their variance, but it would do so both ways, and ultimately I would expect them to end up in about the same positions on the ELO scale. As evidence, I give you the top five players, by rating, from Qatar, along with their results:
This is one tournament, and thus a painfully small sample size, but I think these results would hold up over time.
Thus endeth a minor disputation over a particular inconsequentiality.
ReplyDeleteI am not so sure about the ending. There were nine players in the FIDE top 20 who played. In terms of ratings gains and losses, 6 went up and 3 went down. None were all that big in terms of change, but Mamedyarov took a bit of a hit.
It's also telling that, of the nine, most ended up lower in the standings than they were in the pre-tournament seedings (two went up, 6 went down, and one- Carlsen- was even.)
It is interesting to note that the Chinese players tended to get some big ratings bumps. I attribute this to the problems with separate or isolated ratings groups and the resulting disparities (which this is all about), which would be reduced if these players played in the same groups more. The trend is in that direction, so there is hope.
Here is a much better-explained piece on the general inflation that has afflicted ratings for some time. http://members.shaw.ca/redwards1/
As a side note, I think the recent spate of top players blowing basic technical endings is a sign of where they might compare to the 1960s-1980's players in spite of the higher ratings, but that is just my subjective feeling,and I doubt it would survive the cold light of chess truth!
Okay, so not over for now!
ReplyDeleteI think the recent spate of top players blowing basic technical endings is a sign of where they might compare to the 1960s-1980's players in spite of the higher ratings, but that is just my subjective feeling,and I doubt it would survive the cold light of chess truth!
I think that probably has more to do with the lack of adjournments, which allow for both some rest and for reviewing the relevant material.
On this particular issue I also keep thinking of Nunn's work in one of his chess puzzle books (although that does NOT cover ratings), and Ken Regan's work.
As for the other issue of whether or not playing in the closed tournaments insulates higher rated players, we'll leave that alone for now for lack of evidence.
For the impact of closed systems on ratings, it is worth googling Claude Bloodgood- he is from my hometown, unfortunately. He is one of the reasons the USCF changed the rating system.
ReplyDeleteWhen I first started playing chess, a friend of mine, Gary MacGowan, told me that his club used to go to the prison to have matches with the inmates, because their ratings were inflated, and his club members got easy rating points off them. Gary told me he once received almost 200 points from a match, but I suspect that he exaggerated a bit!