No surprises so far, Carlssen and Anand in the top spots!

http://www.chessbase.com/newsdetail.asp?newsid=5919

Heh, I was just reading that page. Kosteniuk has gotten a point off Anand, Carlsen and Aronian. Not too shabby.

I feel a bit sorry for Karpov though. He started off with 9/12 for clear third place and has since then just scored 5 points in the last 16rounds.
Still a very good overall blitz performance for a player of his age.

Well, Karpov's style is actually well-suited to blitz, and the main disadvantage that comes with age (deteriorating calculative ability) isn't nearly as important in blitz. I'm not surprised that he did well.

Karpov seems to have done quite well for himself in blitz over the past several years. It's in the rapid games where he gets destroyed.

Nice to see a Judit Polgar appearance...

Another fantastic performance from Carlsen. 2894 performance rating. He won/drew every single minimatch he had including 2-0 versus Anand, and only gave up 6 draws in the 42 rounds. And he was probably still a bit sick. Can Kasparov really have had this big an influence?

I'm not especially surprised Carlsen won the event, he is clearly World #1 or as close as makes no difference and young players are known to be good at blitz.

two weird things I noticed about this year's event:

It was played 3 2 (Fischer time) instead of 5 0 - I think this is a first.
They used some messed-up round-robin pairings: for example according to Chessbase Carlsen had 11 whites and just 3 blacks on d2. Anyone know what was going on here?

http://tal.russiachess.org/results/2009/blitz/

according to the round by round pairings on the official site that is not correct.

btw, he beat the players placed 2-5 (Anand, Karjakin, Kramnik and Grischuk) all with 2:0, impressive.

Not sure what you mean by on d2, but I'm drunk I guess.

You can download the games from chessbase. Every single player played one white and one black against each other player. Glad to hear about 3 2. As I've been saying forever, non increment chess games are just stupid.

And not to be argumentative just for the sake of arguing but 39 year old Anand at #2 isn't exactly a fresh blossom, nor is second to last place 24 year old Naiditsch or last place 25 year old Kosteniuk exactly geriatric. I don't really think age has much to do with blitz ability.

d2 I meant day 2 (it was a 3-day event), it looks like Chessbase just got that wrong and everyone played according to a normal schedule.

I'd like to see a study of results in competitive blitz games with age, corrected for ELO. I think you'd find at least some correlation. You'd only need a couple of tournaments of this size to get a good sample. If anything, Kosteniuk is actually a good example for young players overperforming, she is rated more than 200 points below her average opponents and got a respectable score.

So I graphed it for this tournament only and does not look like there is much correlation. I could try and explain it away by saying "well, the outliers like Karpov are invited because they are especially good at blitz, the other 60-year old 2600s would have done much worse" but instead I'll just say more research is needed. I can post the graph if anyone is interested.

Going over some of the games from this event it's really interesting how so incredibly dynamic the top level games are. I remember reading somewhere a joke about Tal's style of play where two checks were worth at least a pawn, three checks a piece. It really seems to be coming back to that. Slow games, blitz games, all games. Material is taking a very distant second stage to initiative.

Just kind of funny. We all learn at first that the goal is to checkmate the opponent's king and want to sac everything we have to force a brilliant victory. Then we all become sophisticated and eventually learn a pawn or even a bad piece is more than enough to win a game. Only to go full circle in the end and learn that checkmate is what wins the fastest!

Kind of reminds me of poker. At first, we learn to not pay off with worse than top pair. Then you learn that you must fold top pair when getting too much action. And then only much later you learn once again that top pair really is the nuts.

Don't let good work go to waste! Share.

here's the graph, I can't get it to display the labels like I would like but it is a start and better than no labels at all, I think.

I think my methodology in deciding "expected score" is better than sites like Chessbase: I have calculated the expected score against each opponent and summed them instead of taking an average rating. After all, your expectation against a 2600 and a 2400 is not necessarily equal to your expectation against two 2500s.

http://www.chessbase.com/news/2009/t...es/blitz02.htm

A really interesting feature from chessbase. They've collected videos of a number of the games, and offering a java board to replay the game besides the video. It's really neat to see the different styles of blitz play in action. Watching Carlsen - Svidler, for example, Carlsen seems to be extremely deliberate and careful with each move, while Svidler just pounds out his moves often without second consideration. You can also see visibly when Svidler realizes his position is lost and starts moving at light speed (something alot of people do when losing). It's also interesting since you can see Magnus's selection of candidate moves based on his eyes and hands. Like on move 40 you see Magnus preparing to play Kc7 which swaps over to h4 at the last moment.

Just kind of neat, not to mention entertaining, to see a little deeper into the games this way.