Chess 2026 Candidates Second Half Recap

Crossposted on Substack

Congratulations to Javokhir Sindarov for winning the 2026 Candidates! At just 20 years old, he had one of the great tournament performances in recent history. I’m very excited to see how his career progresses.

The second half of the Candidates may not have been as captivating from a viewing perspective as the first, with how Sindarov was running away from the tournament. But there are still lots of lessons we can learn and apply to our own games.

Last time, I reviewed every decisive game from the first half and identified what I thought was the key mistake leading to each loss. Today I’m going to do the same thing.

What may be surprising is that even though these are the best players in the world, their mistakes are often not so different from ours. As a lowly National Master, I learned a lot from analyzing these games, and I believe you can too.

Protect the seventh rank - Nakamura vs. Caruana (Round 8)

Rooks on the seventh rank are deadly. What’s especially interesting is that in this game, both sides got their rooks onto the seventh/second rank; white was just a bit faster!

This is the critical position. Both sides already have one rook on the seventh/second rank. White has just created an escape square for his king on h2, making sure he won’t be checkmated on the back rank. Black could double on the second rank right now, but the problem is that then Rd1 would threaten checkmate.

To avoid the checkmate, Fabi mirrored Nakamura’s move with h5, creating an escape square for his king on h7. However, this innocent-looking move turned out to be the losing mistake. White played Rd1 anyway, preparing to infiltrate with his second rook. Because the f2-pawn still blocks the second rank while black’s f-pawn has advanced, white’s rooks are more menacing than black’s.

They soon reached this position:

All seventh rank rooks are not created equal! White’s king is protected by the g2-pawn, which is invulnerable thanks to the protection of the rook. Black’s king, meanwhile, is wide open. Black is lost, and went down without much of a fight.

Going back to the critical position, what should black have played instead of h5? Rab2!

The idea of this move is to play Rb7, guarding the seventh rank. Yes, you will lose the a5-pawn, but now black will have an active rook on the second while white will not. White is only slightly better according to the computer. This is far better than what happened in the game, where material was equal but he was positionally lost.

Dominance of the seventh rank is often more important than material.

Don’t make yourself walk a tightrope - Giri vs. Praggnanandhaa (Round 8)

Just because an opening is objectively sound doesn’t mean that it’s a good idea to actually play.

This position was reached from the Vienna variation of the Queen’s Gambit Declined. The Vienna, of course, is a very strong opening. I would argue, however, that Pragg made a mistake by letting this particular position occur.

The major problem here is that black has many options, but white straightforwardly has a fine position in all lines. White is risking little here. Black, meanwhile, can easily go wrong with natural-looking moves. He played Qxa5 and was already seriously worse. While he played well after and the game was a serious fight, Giri maintained a strategic advantage the whole way.

The best move in the diagram position is apparently to play Bxc4 and then Nc6. But even there, the engine gives white a plus (he is still up a pawn after all), and the position is completely safe for him. Pragg took on a position where he could easily go wrong and make a mistake, but which had no such corresponding danger for white.

Don’t play openings that are only risky for you. Either play openings that are safe, or that make both players have to find difficult moves.

Miscalculation

Caruana vs. Giri (Round 9)

Chess is 99% tactics, right? Well, for club players, that might be true. At the Candidates level, not quite. But they still do miss tactics on occasion.

Giri had outplayed Caruana from an Italian, and gotten a much better endgame. White has double isolated e-pawns, with the e4-pawn being especially weak. With c5, however, Giri gave white an excellent chance to save the game.

Fabi could have played dxc6. Then after Qxc6 comes the strong move Nc2. If Bxe4, then Nxb4 hits black’s queen!

So satisfying when you can respond to a threat on your queen with a threat of your own! Black is still better, of course, but with the e4-pawn safely traded off, white’s biggest problems are solved and he will probably make a draw.

Instead, Fabi missed the opportunity and resigned after just six more moves.

Sindarov vs. Praggnanandhaa (Round 10)

Here is the critical moment. In his usual daring style, Sindarov has sacrificed a piece to open up lines against the enemy’s king. Black’s queenside pieces desperately need to get into the game and help with the defense. Developing the bishop is the only real option. The question is: d7 or e6?

Be6 stops Rf7. White will play Rf6, increasing the pressure. Then black has the important move Qh7, coordinating his pieces for the defense!

Black is not going to get checkmated. The game will likely end in a draw.

I’m sure Pragg saw this. But he went for Bd7. Why?

One is that Rf6 now is much less strong, because there won’t be an attack on an e6-bishop. Two is that he miscalculated the results of Rf7.

Black is actually just lost here. Trading the queen for the rook is forced; otherwise white’s attack is overwhelming. And after a couple of checks, white picks up the d7-bishop.

White has too many extra, passed pawns. Sindarov went on to win easily.

Seeing such a great player as Pragg blunder like this is surprising. But it’s hard to calculate precisely when you’re being attacked with a wide-open king.

I won’t be the first person to tell you that improving tactics and calculation is the biggest thing most players can do to boost their chess.

Passed Pawns Must be Pushed - Wei Yi vs. Esipenko (Round 13)

In the last article, I was quite critical of Esipenko for the weakness in his strategic play. He is a very talented player (and the only one who came close to defeating Sindarov in this event) and had a better second half. But he made another costly strategic error against Wei Yi.

Esipenko has played daringly, sacrificing a piece for an attack and extra pawns. Right now, he has triple (!) connected passed pawns in the center! While his attack has stalled, these remain a huge asset. Unfortunately, he hamstrings that asset with his next move.

Bc4 safeguards the d5-pawn and also blocks the c1-rook from targeting the c5-pawn. But it also prevents the pawns from advancing. The d-pawn is now stuck babysitting the bishop. This means Esipenko can’t use the biggest advantage in his position.

Let’s fast forward a few moves:

The d4-pawn still can’t advance, because the c4-bishop would drop off. Black is having serious problems even keeping the e3-pawn defended. In a few moves, white won this pawn, and black instantly resigned.

Instead of fixing his pawns in place with Bc4, black should’ve pushed forward with d4!

The c5-pawn actually can’t be taken for free, because of the check on f5. Black’s pawns are very menacing, and he’s certainly not just going to lose them as happened in the game.

Sometimes the ancient proverbs are right. Passed pawns must be pushed!

Don’t Sharpen the Play if Your Pieces Aren’t Ready - Bluebaum vs. Giri (Round 14)

After a tournament of drawing twelve games, losing one, and never sniffing a win, perhaps Bluebaum decided to go all out in the last round. However, the way he did it was not prudent.

We’re still in the opening phase here, both sides out of book. Best for white is Nce4, putting pressure on the black kingside and preparing to trade that knight and then safely retreat the g5-knight to e4 in its place.

Instead, he castled queenside, allowing black to immediately pry open the queenside with b4! White had no choice but to respond with d5, hoping for the mess of tactics to work out in his favor. Unfortunately for him, he’s already objectively lost.

Look at this position. White’s pieces are not harmonious. The biggest problem is the bishop on d2, which has no obvious prospects and currently does nothing but obstruct the d1-rook. Black’s king, despite white’s aggressive intentions, is completely safe, while white’s king has open lines all around it.

Anish won a piece in the next few moves, and Bluebaum resigned after move 28. When your position is not harmonious and your opponent has the safer king, sharp tactics do not tend to work out in your favor.

If You Play a Surprise Opening, Know How to Play It - Esipenko vs. Caruana (Round 14)

In the last post, I praised a surprise opening as the reason for three different wins! Clearly, it’s one of the best ways to have real winning chances at this level. This game adds a warning to that advice.

Esipenko went for Qd3 in the Najdorf. This is a nice surprise because it’s perfectly playable, but only the 15th most popular move after a6! So he should have a good shot at getting a playable position that he knows better than his opponent does.

Instead, after Fabi responded with the natural move g6, it seems that Esipenko was immediately out of prep. He spent 3.5 minutes to play Bg5, castled quickly, and then this position was reached:

Fabiano has done nothing but play some of the most natural developing moves after Qd3. Here, Esipenko thought for almost 12 minutes before playing Be2! Then, after Fabi played another natural move Nc6, he spent 12.5 minutes before playing Nb3.

After Nb3, black is already better. Esipenko proceeded to get quickly outplayed and was already lost before move 20.

Clearly, Esipenko was out of book early and didn’t know how to play the position. Of course, it is impossible to remember every line, and being out of book early is a normal part of modern chess. But after you uncork a major surprise, you should have an idea of how to play the position if your opponent continues naturally. Otherwise, you’re just willfully entering a position where you’re not comfortable.

It’s crucial to have an idea of the plans and ideas behind your openings, not just the moves. This is especially true in unusual lines like Qd3, where standard plans may be less applicable. And it’s important to know what to do against natural moves, even if they’re not the most popular or recommended by the engine. If you don’t know how to play against the normal-looking moves, then you don’t really understand the opening.

~

Now we’ve analyzed the key reason for every loss in the 2026 Candidates! I learned a lot looking through these games, and I hope you did too.

The way I’ve structured this analysis mirrors my general chess philosophy. While chess is an extremely complicated and concrete game, there is a logic that runs through it. I believe that when combined with a foundation of strong calculation and opening prep, following key principles will take you far.

Thanks for reading, and if you want me to do more chess articles like these then please let me know!