To accomplish that aim, I’ve done radars for three players from one tournament: the 2009 Tal Memorial, a round-robin also with nine rounds, similar to the Norway Blitz tournament. (Vladimir Kramnik (6/9,) Magnus Carlsen (5.5/9,) and Viswanathan Anand (5/9.))
Let’s now take a representative player from the Norway Chess blitz tournament. For a decent representative we want someone who had a good tournament, as we would expect the results to be generally better in longer chess than in blitz chess. However, we don’t want an outlier, such as Magnus Carlsen, who had an outstanding tournament. Let’s go with Levon Aronian, who had a good performance in Norway Blitz, but was still relatively middling in some metrics.
(A note: Since I’ve added three new players, for clarity when viewing these, and as I think it suits the data better, the outer line is for spots 1-3 on any metric, and the center circle for players 12-13.)
We can see that compared to his earlier radar here,
in several categories, Levon performs relatively similarly. For instance, his win percentage, despite the seeming improvement, is actually ranked the same. However, we can see a large drop-off in his mistakes+blunders rate and his average centipawn loss when compared to this new pool of players.
The cause for this is the dominance of these categories by Tal!Kramnik, Tal!Carlsen, and Tal!Anand. Each of these players (as shown below) posts huge numbers in these two categories, which shows the level of mistake avoidance possible with longer time controls.
Perhaps even more impressive, we can clearly see that each of these had extremely high levels of mistake avoidance and centipawn loss relative to the blitz players despite their varying performances. Even Tal!Anand, who did poorly in several categories, finished first and second in these categoried respectively.
Similarly, despite Magnus Carlsen producing a dominant performance in Norway, he scored 1.53 in M+B and 15.2 in % pawn loss, scores beaten by Tal!Anand with 1.29,12.4, Tal!Kramnik with 1.67 and 13.4, and Tal!Carlsen with 1.5 and 8.6. Clearly, players at similar levels are much more likely to avoid silly mistakes, and play better in general, when playing standard chess, even at the highest level.
In other categories, one by one:
Firstly, we can see that a given points total doesn’t necessarily depend on the time control, but that there is potentially more variance owing to the amount of mistakes in blitz chess. Wins track similarly, as there are many more draws (either grandmaster or earned) in standard chess.
Rating increase/decrease likely doesn’t translate over time controls, as these players have played many fewer blitz games, and so the blitz ratings are much more volatile. In addition, performance rating is unimportant, because the ratings were higher in Norway Blitz than in the Tal Memorial (although this could be remedied with more similar rating levels.)
Thanks for reading. I should be back in the next few weeks with another post, and I welcome feedback on anything either in this post, or on my blog.
Discussion thread: Thread 173746
P.S. I shouted this out last post, but I’m writing fiction at my other blog and I’d very much welcome feedback or comments there as well.