Caruana: Do you want to lose?
Shirov: Sure, why not?
Caruana: Ok, let’s play the Najdorf.
Shirov: Cool, that’s where I got some new ideas.
Caruana: I bet you do…

This wasn’t the conversation before the game, but it could have been. I am kidding of course. Shirov probably wanted to prove that he can still slug it out with the big boys. What follows is the worst piece of preparation in modern history. Shirovs comes up with a super-refined moveorder, Caruana makes all the standard moves and look what happened. To his credit, Shirov almost managed to turn it around, but ultimately threw it away again.

On a sidenote: Upcoming norwegian drawmeister Christiansen managed to catch Leko in a forced draw. Very promising!

Here is what happens when you take risks and miss your chances:

Note: Keymer plays a pure strategy. Against g6 it’s always g3.

The following game shows that nowadays being a top player is all about winning with white. With black it’s simply impossible to beat a Drawmeister, at least in forced variations like the Sveshnikov, the Bg5-Najdorf or the Marshall Attack. Actually I am quite puzzled that Carlsen didn’t go for his usual strategy of playing second rate openings where he gives his opponent something like + 0.30 for free in order to get him out of book. On the other hand, the path of freedom in the opening (and everything else for that matter) becomes narrower every day. Maybe Carlsen was hoping for 7. Nd5.

Playing the Sveshnikov is basically a draw offer these days.

Pretty funny to watch the some of the underdogs wheeling out the Drawmeister and getting away with it. Amongst them german supertalent and rating-flatliner Keymer. Notable mention goes to Christiansen from Norway who even plays the recommended gambit in the Classical French with black. I told you guys, this works!

The following draw was more of a bailout than a Drawmeister though. Apparently Korobov dodged the theoretical discussion.

My man Radjabov won the event! Who would have guessed that (except me of course)? Did he get lucky? Of course he did. Let’s go through his openings.

White (7 wins 7 draws 0 losses):

• KID/Grünfeld with g3
• QGD with e3
• QGD with Nf3
• Caro Kann Two Knights
• English Symmetrical with Nf3 and Nc3
• Italian with d3
• Grünfeld with Nf3 and Rb1
• Grünfeld witb Bb5
• English Symmetrical with Nf3 and Nc3 again
• Marshall Attack with d3
• again
• Semi Tarrasch transposing to QGA
• Semi Tarrasch transposing to Panow Attack
• Sicilian Rauzer with Be2

Black (2 wins 11 draws 1 loss):

• d3-Italian with h6, a6 and Be6
• Bf4-QGD with Nbd7 and Nh5
• Catalan mainline with dxc4
• Nimzo Indian
• Berlin Defense
• QGD allowing the Exchange Variation
• again
• Sicilian Rossolimo with g6
• English Mikenas
• Reversed Dragon with Bc5
• again
• again
• Catalan mainline with dxc4

What I find interesting is that he played the d3-Marshall in a must-win situation and won. That’s actually quite amazing. He got lucky when Ding forced him to play dynamically in the Panow structure. Theoretically important are the two games in the QGD-Exchange against Mamedyarov. That may have had something to do with the fact that they played on the Azeri Team and shared some analysis.

In later rounds he avoided any further discussion, even though Ding clearly gave him the chance to repeat this line.

You can’t blame the guy for playing “solid”, because it is correct. In a game with complete information it is game theory optimal to play the best move, even if it leads to a dull position. If your opponent forgets his analysis or blunders, good for you. Patience is required.

Radjabov is putting on a clinic! He is doing nothing while his opponents are blowing up themselves. It’s quite similar to Nakamura’s strategy, except Radjabov didn’t forget his analysis. Kasparov and Karpov could keep drawing games forever, because there was no armageddon. This armageddon game where you choose black of course if you can, makes all the difference in the World.

It was quite surprising to watch Nakamura scoring just one draw out of four games. Obviously it had something to do with his opponent Aronjan, but they are currently just 13 Elo points apart in classical chess, while Nakamura is the favorite in shorter time controls.

My best explantion is this: Nakamura is the champion of pattern recognition which shows for instance in Puzzle Rush. He is lacking a chess education and was basically trained like a AlphaZero by playing blitz on the internet. In 960 there are no patterns, at least not in the early stage of the game. That seems to put him at a massive disadvantage. Once the positions start to look familiar, they are already lost. Ok, small samplesize. I could be wrong.

Note: It’s the exact opposite with Svidler. Svidler calculates concrete variations like a classical engine. He doesn’t need patterns.

MVL constantly seems to forget theory in this tournament, but not always to his disadvantage. He got away with it against Karjakin and now he even manages to score the full point with black against Nepo, who produces high variance for a reason.

Btw, am I the only one who has a problem with Rabiega’s commentary at chess24? He certainly calculates better than the average FM, but he acts like he is totally clueless about modern theory. This feels kinda familiar to me, because I remember such guys from the Oberliga. Well, this is probably typical for a veteran west-german GM, because there was no chess education. They just played chess in clubs and in tournaments. 100% learning by doing. Last night they mentioned Christopher Lutz, who made an impact on the german scene with his strong engine preparation. The funny thing is that Rabiega connects his decline with his time-out from chess. He missed a few easy engine-victories, but what he missed most is the bus. His approach to chess is simply outdated.

In the old days in order to become a strong player you had to have the talent and you had to be critical towards published theory. Then came the engines and all of a sudden everyone had access to the “truth”. Clearly the “truth” back then was relative to the strength of the engines, but it was clearly better than the Fake News that you got from published theory. Veterans who were too slow to make the transition got killed over the board. Nowadays everyone is working with strong engines, so you got to do your homework before you are even allowed to show your talent. The key is not to know what everyone knows, but to know what nobody else knows. It’s similar to Leko’s approach and slightly paranoid.

Chessbomb is a very nice plattform because you get the moves in colors. Blue means very close to the computer top candidate, red means a massive deviation (= blunder). Here is a game that is blue from the start to the very end. The question remains how deep the preparation went for both sides, especially since none if this was forced. Carlsen goes for lines that leave many options.

Obviously I should be happy because this shows once again that chess is a draw, even though it’s not always forced. On the other hand this tournament is a strong argument for inviting a few underdogs to make it interesting for the crowd. I am not exactly a fan of Rapport or Topalov, but you need a few players who play for three results. Unfortunately Van Wely retired, because he was a fighter.

This year there is no real target in the lineup, so there won’t be many decisive games. Since Nakamura became rock solid, Nepo is pretty much the only random factor left. He can beat everyone and lose to everyone, but his fequency of decisive results is significantly higher than any one in the field. To prove this theory he already dropped half a point to Anand with white. To be fair, Anand played a nice game.