FIDE Grand Prix Tashkent - Round 8

Jobava defeated Gelfand and joined Andreikin and Nakamura at the top.

FIDE Grand Prix Tashkent - Round 7

Caruana finally won a game. Nakamura and Andreikin continue to lead.

FIDE Grand Prix Tashkent - Round 6

Unexpected events continue. Andreikin, the 10th seed in the 12-player tournament, is now tied for first. Meanwhile, 2844-rated Fabiano Caruana remains without a single win.

FIDE Grand Prix Tashkent - Round 5

Nakamura leapfrogged Vachier-Lagrave and now is in clear first place. The model boldly predicts that almost anything can happen:

Performances can fluctuate quite dramatically; the winners of the last Grand Prix, Caruana and Gelfand, are currently near the bottom of the standings.

FIDE Grand Prix Tashkent - Round 4

All draws today

FIDE Grand Prix 2 (Tashkent)

While I was buried in work, the next Grand Prix tournament began. Surprisingly, Caruana starts off with just 1.0/3. Before the tournament started, he would have had a 52% chance of winning, but now that has been reduced substantially. Vachier-Lagrave is now the favorite due to his 2.5/3.

FIDE Grand Prix - Round 11

Unsurprisingly, Gelfand and Caruana tied for 1st.

FIDE Grand Prix - Round 10

Caruana and Gelfand reclaimed the lead today. However, Caruana is definitely the favorite. He is much higher rated than Gelfand, and he will face bottom seed Tomashevsky (2701) in the last round. The rating gap in Gelfand's last game will be much smaller; Gelfand (2748) plays Svidler (2732).

FIDE Grand Prix - Round 9

Both Gelfand and Caruana lost today. Now half of the players are tied for first with 5.0/9.

FIDE Grand Prix - Round 8

Svidler defeats Domingues; remaining games are drawn.

FIDE Grand Prix - Round 7

CARUANA LOSES! Yes, that can happen.

FIDE Grand Prix - Round 6

Caruana wins as usual

FIDE Grand Prix - Round 5

All the games were drawn, so there is very little change in the forecast.

FIDE Grand Prix - Round 4

Caruana wins another game and leads with Gelfand at 3.0/4

FIDE Grand Prix - Round 3

Gelfand defeated Grischuk and moved into sole first place. His chances have certainly improved, but Caruana remains the favorite due to his 2844 rating.

FIDE Grand Prix - Round 2

Svidler and Nakamura caught up with Gelfand and Caruana today. Now for the standard comment about how the probabilities in the first column add up to more than 100% because multiple players can tie for 1st, and then back to writing my thesis.

FIDE Grand Prix - Round 1

As usual, Caruana won his game, so his odds improved in the forecast.

FIDE Grand Prix

The Grand Prix starts soon, so it's time for another forecast. The favorite in the 12-player round robin is Fabiano Caruana; we will see if he can maintain his impressive 2844 rating. He is definitely the player to watch. But since there are only 11 rounds and such strong competition, it is actually more likely than not that someone else will claim first place.

As usual, the probabilities in the first column add up to more than 100% since more than one player can tie for first.

This is the part where I often lament that important events are decided by a small number of games. Most people who work with applied statistics would be uncomfortable drawing conclusions from a sample size of just 11 rounds. But the Grand Prix has a very good feature that should be included in more championships: not everything depends on one tournament. This is the first of four tournaments in the Grand Prix cycle. Each participant will play in 3 events. When the outcome is based on 33 games rather than just 11, there is much less scope for luck. My article "Are the World Champions Just Lucky?" showed that random variation can matter a lot in short matches. You would think that a 24-game match between two equal players would end with a very close score, but there is more than a 10% chance that one of them will prevail "decisively" by more than 12.5 - 8.5. That's how much luck can matter in a small sample. So it's great that the Grand Prix depends on multiple tournaments - now why can't we do something similar for the Candidates Tournament? The second most important event in the chess world is decided in just 14 games (it's an 8-player double round robin). If the tiebreak rules had been a little bit different, the 2013 World Championship would have been Kramnik vs. Anand while Carlsen would have had to wait another year. If the challenger to the world championship had to perform well in multiple Candidates Tournaments, then we could have more confidence that the result was truly due to skill rather than luck. Given that FIDE found sponsorship for the four Grand Prix tournaments, it shouldn't be impossible to find sponsors for a more important series of Candidates Tournaments. This will likely mean that the World Championship will be held less often than once a year, but that is quite acceptable. For a long time, it was once every three years, and ensuring that the challenger is the best qualified player and that the winner is truly the best is crucial. With the knockout world championship tournaments in the late 90s and early 00s, we have already seen what happens when a player is hastily crowned as champion without a good candidates cycle (did anyone believe that GM Khalifman was best in the world?).