30 November 2015

Sergey Karjakin: My Nerves Proved Stronger

The World Chess Cup winner recalls the upheavals of a lengthy and extremely challenging tournament in his interview to Vladimir Barsky.

The interview has been published in the “64 – Chess Review” Magazine № 10/2015

– Sergey, have you calculated the total number of your games played in this World Cup? 

– 16 classical games and five tie-breaks for sure, though I need to rack my brains to remember the exact number of games in these very tie-breaks.  By most conservative estimates the total volume of effort would equal to as if as I played as many as 21 classical games. It is same as playing in two super tournaments coupled with an incredible tension peculiar to the knockout system!


– And it all started with Ermes Espinosa…

– Well, I dealt with him in a more or less difficulty-free manner. I seized the initiative in game one and then he overlooked a thematic strike.

Karjakin – Espinosa

19.c4 bxa4 20.c5! dxc5 21.dxe5 axb3 22.Qxb3 c4 23.Bxc4…

I succeeded in winning this match in quite a confident fashion, but in round two all hell broke loose! 

Onischuk – Karjakin

The game didn’t go well for me as I might not have properly played into the tournament yet. Having emerged with a good position out of the opening, I opted for 21...Re8 in this position. It spoils nothing just yet, although I saw that there was a superb move 21…d3 at my disposal and I cannot really explain why I failed to go for it. White would need to be on the alert after that. 

22.Qd3 Bc3 23.Rd1

Now I needed to come up with 23…Ng4!, which is nothing that’s so self-suggestive at all. 

After 23…Nd7 24.Qf5! Bxc4 25.bxc4 White obtained a bishop pair advantage, whereas his strong bishop comes to occupy the d5-square. Onischuk overtook the initiative and went on to deliver the finishing blows very precisely. I never had any real chances of bailing out. 

– So, this was the first time when you found yourself in a must-win situation, wasn’t it? 

– Yes-yes-yes!

Karjakin – Onischuk

1. e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3 Bc5 5.Nc3 a6

I was surprized by this move since after 5…0-0 one of White’s main ideas is exchanging on с6 anyway.

6. Bxc6 dxc6



I just needed to go on with finishing my development in this position. However, Black’s setup is very solid, but at least I could try to capitalize on being a tempo up, whereas I was in the belief that I could punish Black for the 5…а6 move. I pinned my hopes on 7...Bxf2+, because after 8.Kxf2 Qd4+ 9.Be3 Qxe5 10.h3 White is slightly better. 

7…Qd4 8.Be3 Qxe5 9.d4 Qe7 10.dxc5 Nxe4 11.Qd4 Bf5

I was somehow confident that Black had no other choice but to retreat his knight to f6, when after 12.0-0-0 White is better. I completely overlooked that he could play 11…Bf5, since taking the g7-pawn with the queen is impossible in view of 12…0-0-0 with strong initiative for Black. So, Black actually equalized completely without any delay. One of the old games Carlsen – Alekseev in which castles was made instead of 5…а6 is associated with this topic. Carlsen took on e5 with his knight anyway, but that was a brave decision since it was initially evident that Black was completely OK there. Alekseev even ended up winning that game, although the opening finished in an equal position. 

12.0-0-0 0-0 13.Nxe4 Qxe4

Had he opted for 13...Bxe4, it would have been extremely difficult for me to strike up any sort of struggle because he would have played Rad8 next move, exchanging the rooks. Black’s position would have been equal and rather comfortable, objectively speaking. Onischuk, however, wanted to play as solid as possible, but I still managed to overtake the initiative in the end. 

14.Qxe4 Bxe4 15.f3 Bf5 16.Bf4 Rac8 17.Rhe1 Be6 18.Re3 Rfe8 19.Rd4 b6 20.Rb4 a5 21.Ra4 Re7 22.c4 Rd7 23.cxb6 cxb6 24.c5 bxc5 25.Rxa5 c4? 

Black should have striven for more active play via 25...Rcd8 with the idea to penetrate White’s camp with his rooks. Onischuk’s opinion was that 26.Re1 was best met by 26…h6, retaining all his threats. I was then very unlikely to be able to win this game.


White has come to feature serious practical chances because should he succeed in exchanging off a pair of rooks and his a-pawn will be free to start running fast towards the queening square. 

26…Rdd8 27.Ra7 h6 28.Rc7 Rxc7 29.Bxc7 Rd3

This is yet another error upon which White manages to successfully regroup his pieces so as to place his bishop on с3. 

30. Ra8+ Kh7 31.Ba5 h5 32.h4 Bf5 33.a4 Rd6 34.Bc3

White probably features quite a substantial advantage already, which I managed to convert in the final run.

34…c5 35.a5 Rg6 36.a6 Rxg2 37.Rf8 Be6 38.a7 Bd5 39.a8Q Bxa8 40.Rxa8 Rf2 41.Rc8 Black resigns.

I was anticipating the tie-break with certain optimism as I was established in the belief that it was all in my hands. Both 25-minute games ended in draws despite my having emerged with a large advantage out of the opening in the first of them, winning a pawn. It’s a pity that I couldn’t convert, but he came up with strong maneuvers and tenacious defence. In game two I also had an advantage, but the position turned out to be of a sharp nature.

Onischuk – Karjakin

Here I went greedy and took a pawn 24...Qxa4. Instead I had a strong move 24...Nc5!, forgetting that after  25.Qc4 White threatens nothing since he is pinned along the 4th rank and I could advance 25…e5! As a result the game ended in a nice draw. 

25.Qa6+ Kc7 26.Qa7+ Kd6 27.Rxb6+ Nxb6 28.Qxb6+ Ke7 29.Qc5+ Kf6 30.Qe5+ Kf7 31.Rxg5 Rh7 32.Rxf5+ exf5 33.Qxf5+ Ke8 34.Qxh7 Qa1+ 35.Nb1 Qb2 36.Kd1 a4 37.f4 a3 38.Qg8+ Ke7 39.Qg7+ Ke8 40.Qe5+ Kf7 41.Qc7+ Ke8 42.Qe5+ Kf7


– What a picturesque position!

– Ian Nepomniachtchi came up asking, "How have you managed to have played into this? Have you made any illegal moves in this game?" (Laughing)

I won the next game, although unlike the first two games I had no advantage at all. However, when it came to mutual time trouble, I decided to add some more oil into the fire… 

Karjakin – Onischuk

34...c5 35.b5 Nb6?! 36.Rc1! (Black missed this move) 36…Rxa5 37.Bd8 Nxc4 38.Bxa5 Nxa5 39.Rc3 d5? 

After 39…c4 followed by d5 the game might have still ended in a draw, although it seemed to me that White could go on playing for a win without any particular risk. Upon exchanges on d5 and playing Rg3 I would at some moment take a pawn followed by g4, whereas my king would approach to hold back the black pawns. I evaluated the position as slightly better for White, though it should be equal, objectively speaking. However, he blundered yet another time. 

40.Rxc5 dxe4 41.b6

The knight is hanging, whereas 41…Nc6 is decisively met by 42.Rxc6. 

41…Nb3 42.Rc7 Na5 43.Rxg7 Black resigned. 

In game two Alexander was in need of coming back at any cost, and he played out the Catalan Opening with a pawn sacrifice. Honestly speaking, I am happy with the way I performed in that game as it seems to me that I stood worse at no moment and then went on to gradually outplay  him. The game turned out to be a good one.

– Then it came to your favorite Chinese player Yu Yangyi, didn’t it?

– Yes, this is my favorite Chinese player indeed! I have studied his games pretty well: I remember playing him in the recent match against China and also prior to that at the last year’s Olympiad. Back then our game ended in a draw and I was not especially happy about it, but I was not yet a specialist in terms of playing the Chinese successfully! (Laughing).

Yu Yangyi resorted to the same system as was played at the Olympiad. Back then I transited into the Chelyabinsk Variation, whereas during the training session prior to the World Cup together with Yuri Dokhoian we have prepared a more sharp continuation. 

Karjakin – Yu Yangyi

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Nf6 6.Nxc6 bxc6 7.e5 Nd5 8.Ne4 Qc7 9.f4 Qb6 10.c4 Bb4+ 11.Ke2 f5 12.exf6 Nxf6 13.Be3 Qd8 14.Nd6+ Bxd6 15.Qxd6 Bb7 16.Rd1 Rc8 17.g4 c5 18.Rg1 Rf8 19.f5

Now correct is 19...Ne4, and we had some ideas after 20.Qe5. However, I took him by surprise and he opted for 19...Qb6.

This is a typical move although a weak one, since after 20.fxe6 dxe6 21.g5! no tactics work for Black. As a result White ended up with a bishop pair and a better pawn structure advantage. 

21…Nd5 22.Qxb6 Nxb6 23.Ke1 Nd7 24.Be2 Ke7 25.Rd3 Be4 26.Ra3 Rc7 27.Kd2 e5 28.Rf1 Rb8 29.Kc3 Rb6 30.Rd1 Rbc6 31.Bg4 Nb6 32.Ra5 g6 33.b3 Bf5 34.Be2 Nd7 35.a3 Be6 36.h4 Bf7 37.Bf3 Rb6

In this position my opponent allowed me to deliver a nice shot.

38.Rxd7+! Kxd7 39.Bxc5 e4 40.Bg4+ Black resigns. 

As he needed to come back in game two he decided to resort to the Reti Opening, although he is not used to playing like this. However, he failed to take me by surprise because I spent time analyzing this line and was aware that Black features excellent compensation for the missing pawn. I knew everything up to move 16, whereas after that we started playing on our own. He pushed it too hard somewhere, sacrificing the exchange, but was still standing not so bad objectively speaking so that I was prepared to go for the repetition of moves. Then he blundered again and in the final position Black could win in more than one way. He chose good timing to offer a draw on move 37 because if he allowed the game to go beyond the time control move the final result could have been very different. As I had less than a minute on my clock I decided to seal the overall match victory because back then I considered it more important than five additional rating points. 

– After that you were paired against Andreikin, to whom you had lost during the previous World Cup in Norway.

– Well, this match could not be but a difficult one. I was very well aware how dangerous Dima was and experienced no easy illusions whatsoever. In game one he tried to take me by surprise, but it was me who managed to surprise him after all: he admitted after the game that had not been analyzing the sequence of moves chosen by me and was not fond of the position that arose in the final run. I offered a draw as I was confident that he was going to spend some time analyzing and the game would continue. He, on the contrary to my expectations, went along with my proposal!

– Neither did you succeed in engaging the fight in game two, didn’t you? 

– As it turned out Andreikin used to play this line himself though it somehow escaped me. I still managed to put forward a definite number of challenges in front of him. This topic is associated with a recent game Shirov – Kramnik played in the Russian Team Championship, where despite White having won the game, Black emerged OK out of the opening. 

I did manage to transfer my bishop to f6 and generate an irrational setup, but he came up with precise manoeuvres. I cannot say that I simply gave up my white color, I tried but it just didn’t work out the way I wanted. 

In game one I made up my mind in favour of the anti-Berlin as early as move one and played 1.с4! The game evolved to a very complex type of play. I repeated moves in the opening so as to accumulate additional time, but he decided to go on fighting and went 13…е6. It seemed extremely dangerous to me as the bishop gets stuck on f5. 

Karjakin – Andreikin

14.h3 h5 15.f3

I should have probably exchanged on с6. Should he play 15…Rxc6, then 16.f3, while in the case of 15…bxc6 I could go on playing against his positional weaknesses by 16.Na4. As for me, I went greedy and took a decision to win his bishop.

15…Nxe5 16.dxe5 Nd7 17.g4 hxg4 18.hxg4 Bxg4 19.fxg4 Nxe5

In this position I suddenly realized that Black featured excellent compensation: the g2-bishop’s scope is restricted by the opponent’s pawn chain, the g4-pawn is hanging, the queen is about to land on h4, Nс4 is a threat in some lines… Black’s play is not only easier, but, objectively speaking, his position is good either. However, there is still a lot of fight in store. I was maintaining the tension, having successfully eliminated the a7-pawn. Improvements are possible to be found on each and every move, but it should also be born in mind that the real action took place in quick tempo.

20.Bd4 Qh4 21.Qe1 Qxg4 22.Rd1 Nf3+ 23.Rxf3 Bxd4+ 24.Kf1 Be5 25.Qd2 Rc4 26.Qe3 Bg7 27.Qxa7 Rb4 28.b3 Rc8 29.Rdd3 Qg5 30.Nd1 Rg4 31.Qxb7 Rf8

Here I needed to defend my bishop by 32.Rf2, and White is better, whereas I moved 32.Ne3, which seemed like a logical move but which ran into an unpleasant 32...Bd4! shot.

As the rook is untouchable in view of the mating threat, I had to part ways with the exchange. 

33.Rxd4 Rxd4 34.Qc7 Rd2 35.a4 Rb2 36.Nd1 Rb1 37.Rd3 Rc1 38.Qh2 Rfc8 39.Kf2 Qf6+ 40.Bf3 e5? 

There was no end of time trouble for both of us, but here it was my opponent who committed a blunder. 

41. Ne3 d4 42.Ng4 Qg5 43.Nxe5

Taking the pawn and lying in wait!


44.Qh8+! Kxh8 45.Nxf7+ Kg7 46.Nxg5 Rxd3 47.exd3 Rc3 48.a5 Rxd3 49.a6 Rd2+ 50.Ke1 Ra2 51.Ne6+ Kf6 52.Nc5 Ra5 53.b4 Ra2 54.b5 Black resigns.

The return game went off in approximately the same fashion as the game against Yu Yangyi. Andreikin failed to strike up a complex type of play and I simply had an excellent position in which White was objectively required to fight for equality. It resulted in an endgame with a huge advantage for Black arising on the board. I remember him offering me a draw two years ago in a similar situation. Back then I declined the offer and resigned a few moves later. Now it was my turn to offer him a draw, and he agreed. I was very likely to win that game, but enjoying an additional half point played absolutely no role for me at the time. 

Mamedyarov was my next opponent.  He caught me on the novelty as White, having improved on Radjabov’s game against Jakovenko. I had to allow the exchange sacrifice, in return for which White was enjoying an extremely powerful compensation.  It gave rise to a very sharp struggle, where improvements can be uncovered for both sides, but as I had no time to subject this game to a deep analysis it is difficult to pronounce any definite judgment just yet. The following moments branded on my memory during an initial quick review of the game.

Mamedyarov – Karjakin

According to the computer I could just ignore my en prise rook in favor of 24…Be5! followed by the subsequent transfer of my bishop over to с3. However, it had never even crossed my mind and I moved 24…Ra7 once my rook was attacked. In general, when the smoke cleared out White ended up having an edge. 

The resilient defence allowed me to end the game in a draw. The endgame looked like a difficult one to defend, but somehow I had a feeling that it was still holdable. My game against Anand from the Candidates Tournament is still fresh in mind in which the same material ratio of bishop plus knight versus rook and passed pawns occurred. This way all the core ideas were rather familiar to me. However, Shakhriyar posed problems up to the very end so that only upon having made the last move I realized that I was within danger no longer. 

He tired me out a great deal and I came exhausted for the next day game. After the following opening moves 1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 Nbd7 5.Bf4 dxc4 6.e3 he employed a new for me 6…b5 move, although later it turned out that it had already been tried by his second Arkadij Naiditsch. Therefore prior to the start of the tie-break games I spent more time studying the games by Naiditsch than those by Mamedyarov himself!

In that game I failed to find the improvement plan when it came to the final position and offered a draw for that reason. Later on the computed indicated the right path. In general, I was not attempting to dry the game out, but let Black go away somewhat too easy. I should not have acted like this, but was out of my usual stock of firepower that particular day. 

Game one saw the version of the Belin variation employed by Naiditsch on a regular basis. I still had some old ideas prepared for this occasion as we reviewed this setup with Kramnik prior to the start of the 2010 Chess Olympiad in Khanty-Mansiysk. As a result of a more or less forced play a rook ending arose in which I had an edge. 

Karjakin – Mamedyarov

Here I needed to have the 32.f6+ Ke8 (32…Ke6?? 33.Rd6#) moves included, although the 33.a4 Rxb2 34.a5 follow-up suggested by the engine is highly unlikely to be found by a human player. I didn’t play 32.f6+ and he immediately made use of it by announcing check and driving the king away from pawns and in the subsequent game I was in time to advance f5-f6 no longer. He chose a proper moment to push f7-f6 himself and the game ended in a draw. 

He was pressing in game two. Mamedyarov made an attempt to take me by surprise in the opening and opted not for the most ambitious line, to be honest, but rather for the one maintaining tension nonetheless.

Mamedyarov – Karjakin

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3 b6 5.Bd2 Bb7 6.g3 c5 7.dxc5 Bxc5 8.Bg2 0-0 9.0-0 d5 10.cxd5 Nxd5 11.Qc2


I was aware of the excellent move 11...Nf6!, but the reason I failed to go for it has so far remained a mystery to me, whereas after 11...Nd7 12.Nxd5 Bxd5 13.Ng5 I needed to move 13…g6, which I disliked in view of substantial weakening of the dark squares. The engine lines demonstrate the precise moves to be played, but in the game Shakhriyar transited into a better ending. 

14.Rad1 Bxg2 15.Kxg2 Be7 16.Ne4 Qc8 17.Qxc8 Rfxc8 18.Bg5 Nc5 19.Bxe7 Nxe4 20.f3 Rc7 21.Bd8 Rcc8 22.Bh4 Nc5


23.Bf6! was a way to secure substantial positional advantage. Although Rc7 followed by Nd7 seemed to me to enable me with enough defensive resources, the position would have remained highly unpleasant. However, after 23...Kg7 it was not all that bad. Starting from that moment I kept on defending as good as possible.

24.Rfd1 e5 25.Rd6 Ne6 26.Rxe6 fxe6 27.Rd7+ Kg8 28.Bf6 Rc2 29.Rg7+ Kf8 30.Rxh7 Rxe2+ 31.Kh3 e4 32.Kg4 exf3 33.Kxf3 Rc2 34.Kg4 Rc5 35.Rh8+ Kf7 36.Rxa8 Kxf6 37.Rxa7 Rc2 38.b4 Rc4+ Draw. 

The following rapid game saw yet another anti-Berlin according to Naiditsch. But then he should not have allowed me to keep my extremely strong bishop on d5, where he stands very much like a bone in Black’s throat. Meanwhile he kept pondering over his next move for a way too long a time, his time reserve having melted down to a single minute only. 

Karjakin – Mamedyarov

Had I managed to unearth 20.c5! with the idea of 20…Rfd8 21.a3 Bxd2 22.Nbxd2 Nxc5 23.Rac1, the game could have ended then and there. However, I went greedy yet another time, having been tempted by a pawn winning continuation. 

20.Bxc6 Bxc6 21.Bxb4 Qxb4 22.Nxe5

White obviously keeps enjoying substantial advantage in this position also, but the game is not over yet and I went on to demonstrate an indecisive type of play.

22…Qd6 23.Nxc6 Qxc6 24.Nd4 Qc5 25.Nb3 Qe5 26.Rad1 Rfe8 27.Re3 Rab8 28.Rd5 Qb2 29.Qxb2 Nxb2 30.c5 Nc4 31.Re2 a5 32.Rd4 a4 33.Rxc4 axb3 34.axb3 Rxb3 35.Rd2 Kf8 36.f3 Rb5

At this moment Black exceeded his time limit. Although White still retained his winning chances, the whole bulk of struggle was yet to come.

The return game turned out to be a lengthy one, although there was nothing unusual underway. At a certain moment of the game I sacrificed a pawn in return for some compensation, though I was not obliged to do so as the position was equal and there was no way for him to break through. Objectively speaking Black was OK and I was nowhere out of the realms of a draw. As for my winning the game it was a sheer luck, but had no special value. The main point was to win the match.  

– So, you made it into the semi-final! 

– Well, a semi-final and a final is a different story. It needs to be understood that at that moment I finally ran out of whatever firepower I still had prior to my matches against Andreikin and Mamedyarov.

– Was the day off of any help to you?

– A single day off is far from enough! Let’s recall, for example, that the Wijk aan Zee tournament features three rest days per 13 rounds, whereas the World Cup has its first day off scheduled only as early as the semifinals. Therefore, it was the fatigue that took its toll on both final matches. Thus, in the first game against Eljanov I made a completely insane move in the opening.

Eljanov – Karjakin


Here I spent time calculating only 18.Nc4 and other similar nonsense, having completely overlooked that he could simply move 18.Nf3 and after 18…d3 simply exchange off on d3 and on с8, while my е7-bishop is hanging: 19.exd3 Nxd3 20.Rxc8 Qxc8 21.Rxe7 Nxb2.

It turned out so that Eljanov obtained huge advantage out of the opening for no reason at all. Now he could have moved 22.Qd2 in order to win a pawn after 22…Nd3 23.Rxa7 and although the game is not yet over, Black is in for the most difficult of defenses. However, he moved 22.Qd7, upon which I managed to come up with quite a decent maneuver: 22...Qc1+ 23.Re1 Qc8! 

There arose an ending with equal material balance and had my knight stood on f6, a draw could be agreed without any further waste of time. In the game that followed Pavel posed problems in a very shrewd manner so that I failed to equalize comfortably. Although he won a healthy pawn after the end of the time trouble, I still retained drawing chances. The game balanced between his winning and my bailing out. At a certain moment I started believing that he let his advantage go, but then I committed a blunder and my position became very precarious. Game tension persisted till the very end, but I finally managed to make a draw nonetheless. 

A situation similar to that after game one against Mamedyarov arose when I was in the defensive for a long time and was drained out of energy for the upcoming game. Although I was not aiming at drying the game out, I still failed to pose problems for Black. I realized that it was high time that the match went into the tie-break. 

The first rapid game went off badly for me, though I liked Black’s position at a certain moment as I believed Black to be beyond any particular difficulties. But, as it had already happened to me more than once during that tournament, I went greedy because it seemed very logical for to me to grab the а4-pawn, whereas in reality it turned out to be a losing continuation.

Eljanov – Karjakin

24...Qb3 25.Be1! Qxa4 26.Bh4

This is a very powerful bishop maneuver. I shouldn’t have, of course, taken on а4 but rather opt for 25…Nc6 or look out for any other idea. I was in the belief, however, that 26…Rd6 allowed me to keep my position together, overlooking that after 27.Ra1 Qb5 28.c4 I was unable to move 28...Qb6 in view of 29.Bxf6 Rxf6 30.Nd7 (even worse was 29...gxf6 30.Qg4+ Bg7 31.Bd5 fxe5 32.Rg1 Rg6 33.Qxg6). I was reduced to playing 28…Qa6, although it was obviously a bad idea from the very beginning since the black pieces lacked coordination. 29.d4! sprang to life, upon which the rest of the game was an mere roundoff  in which Pavel’s performance was flawless at that.

29…Rxd4 30.Qc3 Ne4

If 30...b6, then 31.Bxf6 and White mates me along the g-file: 31…gxf6 32.Qg3+ Bg7 33.Bd5. I had to give up a piece which, however, was an agony already. 

31.Qxa5 Qxa5 32.Rxa5 Nd2 33.Rd1 Bd6 34.Bf2 Bxe5 35.fxe5 Nxc4 36.Bxd4 Nxa5 37.Bc3 Nc4 38.e6 Rxe6 39.Rd8+ Kh7 40.Bd5 Black resigns. 

Thus I found myself in a must-win situation for the second time in this World Cup. Up to the very last moment I was uncertain as to the setup to choose for the game. I voted in favor of 1.Nf3 literally a couple of seconds prior to the start of the game as I was unsure how to pose problems in front of my opponent. It turned out, however, that my guess was completely lucky. We went by the line in which Pavel had already had several games under his belt, including the one against Malakhov in particular.  In that game Malakhov reacted not especially well as he made a move that was not required and Pavel carried out the same manoeuvre as he attempted during our game: he placed the bishop on b4 and obtained an excellent position after having exchanged that bishop off. Meanwhile, I came up with an improvement: I considered that if I transferred my knight to d3, I should get a slight edge. 

Karjakin – Eljanov

1.Nf3 d5 2.g3 Bg4 3.Bg2 c6 4.c4 e6 5.cxd5 Bxf3 6.Bxf3 cxd5 7.0-0 Nc6 8.d4 Nf6 9.Nc3 Be7 10.e3 0-0 11.Bd2 Rc8 12.Rc1 Nd7 13.Ne2 Nb6 14.Nf4 Nc4 15.Bc3 Bb4 16.b3 Nb6 17.Bb2 Qe7 18.Nd3 Ba3 19.Bxa3 Qxa3 20.Nc5

– I knew everything up to this point.

– But this line is not forced at all! Has it been analyzed that deep?

– This is true indeed. My guess proved correct. 

20...Rc7 21.Nxb7

I remembered that the engine’s evaluation of this move was positive, otherwise Black would have played Rfc8, Nd7, obtaining a decent position, whereas now, despite multiple exchanges, the problems for Black are not completely over. 

21…Rxb7 22.Rxc6 Qxa2 23.Qd3 Nd7 24.Rb1 a5 25.Bd1 Nf6 26.Bc2 g6 27.f3

Instead I had a strong move 27.Qd1!, which I underestimated. White was very likely to end up winning a pawn as a result of that move. Black, however, was not without drawing chances, but a substantial advantage would remain with me nonetheless. The 27.f3 move is not so bad, but it gave rise to an objectively equal ending, although White still kept up a certain amount of initiative. 

27...Qa3 28.Qc3 Qb4 29.Qxb4 axb4 30.Ra1

30...h5 31.h4 Re8 32.Kf2 Ree7 33.Ke2 Rec7 34.Rxc7 Rxc7 35.Kd2 Rb7 36.Ra8+ Kg7 37.Bd3 Nd7

This is actually a key moment of the game, underestimated by a great number of commentators: he needed to play 37...Ng8! in order to transfer his knight to e7. Black gives additional protection to the d5-square, while the knight can be routed into the game via c6. Breaking through Black’s position after that would have been extremely challenging for me. Meanwhile, he did not prevent the center from getting opened, thus allowing my bishop into the game.

38.e4 dxe4 39.Bxe4 Rb6 40.Ra4 e5 41.Ke3


– The following line would have resulted in quite an entertaining draw: 41...exd4+ 42.Kxd4 Rd6+ 43.Kc4 Nb6+ 44.Kc5 Nxa4+ 45.Kxd6 Nc3 46.Kc5 Na2!, when 47.Bb1, runs into 47…Nc3.

– This line defies human imagination altogether! However, I could have decided in favor of 43.Ke3, that is I would still be carefully evaluating which continuation to choose. 37...Ng8 was a more solid concept anyway, whereas now Black needs to be extremely precise with his play. 

Further onslaught went off by itself as my bishop operates on both flanks and my king is centralized. The game ended in my favour without any particular adventures. 

42.dxe5+ Kxe5 43.Ra5+ Kd6 44.Kd4 Rb8 45.Bd5 f6 46.Ra6+ Ke7 47.f4 Nb6 48.Ra7+ Kd6 49.Bf7 Rc8 50.Ra6 Kc6 51.Bxg6 Kb7 52.Ra5 Rc3 53.Be4+ Kb8 54.Rb5 Ka7 55.Rxh5 Black resigns. 

The following game proved rather unpleasant for Pavel. I do not want to delve into the opening part of the game because it was not where the fate of the game was being sealed. I believe the position was approximately equal, but then I committed an error and he could have played 42.Qс6…

– The computer evaluated White’s position as substantially better.

– Yes, but he answered quickly while I was in time trouble; it might have been a special kind of a tactical strategy. 

Eljanov – Karjakin

42.h4? Bf6! 43.h5? 

It appears that Black could have still bailed out after 43.Rc6 Qb2 44.Qb4. Now Black is objectively winning, and I went on to win the game without any problems. 

43...Bh4 44.Kh3 Qxf2 45.Qd3 gxh5 46.gxh5 Bg5 47.Qg3 Qf1+ 48.Qg2 Qf4 49.Qg4 Qe3+ 50.Kg2 Rb8 51.Rc8+ Rxc8 52.Qxc8+ Kg7 53.Qc4 Qd2+ 54.Kh3 Qd1 55.Qd3 Qxh5+ 56.Kg2 Qg4+ 57.Kh1 Bh4 58.Bf5 Qg5 59.Qf3 Be1 60.Bc2 Bd2 61.c4 Qh4+ 62.Kg2 Qxc4 63.Qf5 Qxd5+ 64.Be4 Qe6 65.Qh7+ Kf8 White resigns.

– What kind of a mess was it that White created in the following game?

– White was simply attempting to make a draw!

– Those attempts somehow used to come out badly when you were in charge of white pieces...

– Well, after completion of this World Cup one can say that making a draw as White is a real heroic act indeed! I would like to add that had I even lost this game the match would have continued.  It so happened that he was unlucky to a greater extent than I was lucky. As I believe the game was seen by the majority, I would like to single out one moment in the opening only. 

Karjakin – Eljanov

1.Nf3 c5 2.c4 g6 3.d4 Bg7 4.d5 d6 5.Nc3 Bxc3+ 6.bxc3 f5 7.g3 Nf6 8.Bg2 Qa5 9.Qb3 Nbd7 10.Nd2 Nb6 11.0-0 Bd7 12.Re1 Qa4

Here I played 13.Bf1 hastily, upon which 13...Ne4 generated real problems for me. Meanwhile, I should have played 13.е4 with an excellent position for White: both my bishops are active, the е7-pawn is hanging. So, this was the key moment. From then on Black featured advantage throughout the entire game and Pavel was performing superb at that. It is clear that he could have prevailed more than once. In the final position Black is likely to be winning still, but the winning path is no longer as clear-cut as it used to be. However, he missed a threefold repetition of moves. 

14.e3 0-0-0 15.Bd3 Nxd2 16.Bxd2 e5 17.f4 e4 18.Be2 h6 19.Kf2 g5 20.h4 Rdg8 21.Rh1 gxf4 22.exf4 Rg7 23.h5 Rhg8 24.Rhg1 Qa6 25.a4 Be8 26.Be1 Nd7 27.Ke3 Nf6 28.Rh1 Qa5 29.Bf2 b6 30.Rh3 Kd8 31.Be1 Ke7 32.Bf2 Kf8 33.Be1 Re734.Rh1 Rgg7 35.Ra3 Kg8 36.Bf2 Kh8 37.Be1 Rg8 38.Bf2 Reg7 39.Rh3 Qa6 40.Qd1 Ng4+ 41.Bxg4 fxg4 42.Rh1 Qxc4 43.Qe2 Qxd5 44.Rd1 Qe6 45.c4 Rd7 46.Rd5 Bf7 47.Be1 Qe8 48.Rf5 Kh7 49.Bc3 Be6 50.Rf6 Rf7 51.Rg6 Rxg6 52.hxg6+ Kxg6 53.Ra1 h5 54.Rd1 Rd7 55.Qh2 Qd8 56.Ba1 d5 57.cxd5 Rxd5 58.Qb2 Rd3+ 59.Ke2 Qd4 60.Qxd4 cxd4 61.Rxd3 Bc4 62.Bxd4 exd3+ 63.Ke3 Kf5 64. Bc3 a5

65. Be1 Ke6 66.Bc3 Kf5 67.Be1 Kf6 68.Bd2 Kg6 69.Bc3 Kf5 Draw. 

– Were you absolutely sure about the position having been repeated three times? It was a rapid chess game after all. 

– I just remembered it all starting with him playing a7-a5 and me dropping my bishop e1 in order to defend against h5-h4 so that this key position had clearly branded on my memory. This is sort of a key setup, let’s call it like this. It was followed by him moving his king to f5 and me retreating my bishop to Be1 yet another time – this being the second time already. When he played Kf5 for the third time I realized that I was entitled to claiming the threefold repetition of moves. 

Dirk Jan ten Geuzendam and Sergey Karjakin

– That’s clear now. Well, we have come to the most interesting part of the action – the final!

– This part may well be interesting for the spectators, but not for the participants. It is essential that a qualification be made at this point: we both considered our main task completed as we had qualified into the Candidates Tournament. Extra motivation to keep up the struggle needed to be found.

I fell victim to a physiological error: my body relaxed as I was experiencing a feeling of as if the tournament was over, whereas a match against the belligerent Peter was still ahead of me. The first game came out badly for me. The only thing I would like to explain is why I agreed to multiple exchanges of pieces after 16.d4. 

Svidler – Karjakin

I was perfectly well aware that besides 15...Rf7 there were many other moves to choose from. However, I was confident that after 16.d4 bxc3 17.bxc3 cxd4 18.cxd4 Nxd4 19.Nxd4 exd4 he would play 20.Nc5, at which case having my rook on f7 would have made all the difference. In this case multiple exchanges would have logically resulted in a draw.

As he opted for 20.Qb3 I simply failed to grasp early enough the huge degree of compensation that White possessed for the missing pawn. I hastily played 20…Rb8, and after 21.Rb1! it turned out that Black was simply out of any move in a hopeless position. I still featured some micro-chances if after 21…Qd7 22.Rec1 I would have chosen 22…h6 rather than 22…Qe6. Nevertheless, I was very likely to lose the game anyway. 

23.Nc5 Bxc5 24.Rxc5 Rd8 25.Ba5 Rd6 26.Qc4 Nc3 27.Rxb7 Qe1+ 28.Bf1 Ne2+ 29.Qxe2 Black resigns. Well, I am to blame myself: Peter was maintaining pressure and I gave in.

Game two started with Svidler’s significant improvement over my game against Carlsen, where I had substantial advantage. As for the finesse of the game opening you had better ask Peter. It seemed to me that I was performing quite decently in general and found a way to maintain the fight, although the evaluation never deviated far from equality. 

Karjakin – Svidler

In this position I spent much time studying the move 34.g3. I believed 34…Nd7 to be the only idea for Black, which is then answered by 35.Bd5. As now bad is 35…Qc7 in view of 36.Bxf7+ Rxf7 37.Nxf7 Kxf7 38.Qd5+, he needs to go 35...Qb5 for that reason, while after 36.Bc6 even 36…Qxe5 is enough to make a draw, although the simplest is 36...Qc5. Whereas 36…Qc5! was demonstrated by the engine, over the board I thought that Black should take on е5 and the game would have resulted in a draw. I should have played like this anyway. Although I saw that objectively speaking there was nothing for me in this position, I came up with a “subtle” rook maneuver. 

34.Qd4 Nd7 35.Nxf7 Rxf7 36.Rb2 Qc6 37.Rb5? 

As I considered 37...Bc5 to be Black’s main idea, then my last move is a trap: 38.Rxc5! Qxc5 39.Bxf7+ or 38...Nxc5 39.Qd8+. It turned out, however, that I engineered this trap for myself. 

37…Kh8! 38.Rd5 Nb6 White resigns. 

It is only natural that this game strongly influenced the entire course of the match: had it ended in a draw, I would be quietly playing the following game as Black, which could have also ended in a draw, upon which I would have done my best to win as White. That is, the picture of the game would have been totally different towards being more logical, let’s put it like this. Meanwhile, what happened next was witnessed by everyone…

Svidler – Karjakin

I consider my play to be pretty decent up to this moment of the game. However, after the logical sequence of moves 22...d4 23.Qxf5 Rxf5 I evaluated the arising position as a drawish one. I need to introduce one qualification here: although Black can push playing for a win, I simply lacked any belief that I could win this position because it was an objective draw. Were it any kind of a routine match situation I would have acted exactly like this. And here I thought that I would better die like a man! 

22…Rf6 23.b4 Ne5 24.cxd5 Nd3 25.Qe3 Nxf2

My initial plan was to go 25…Nf4, but then I started to dislike 26.d6 Qg5 27.Qf3. Curiously enough, it looks like I still have some draw here, but no more than that.

26.Rf1 Qe4

I offer transition into an endgame that seems like equal on the one hand, but is not entirely devoid of struggle on the other hand as my king is likely to become an active piece.

27.Rbe1 exd5

As a matter of fact I got sight of the 28.Qc3 rejoinder, but I believed it essential that I could answer 28...Qf5 (28...Qf4 29.Nxd5), followed by 29.Rxe8 Nxh3+ 30.Qxh3 Qxf1+ 31.Kh2 Qf4+. However, very strong in fact is 30.Kh2! Qxf1 31.Nxd5. I missed that altogether as I tried to play as fast as possible. Had he managed to find this continuation that would have meant the end of it as he would have won the game and the match respectively. However, he was after a spectacular idea: 28.Rxf2, counting on 28... Qxe3 29.Rxe3 Rxe3 30.Rxf6 Kxf6 31.Nxd5+. However, 28...Qh4! came to the rescue. 

– Did you see this in your advanced calculations?

– You mean that he could take the f2-rook? No, I didn’t see that.

– This came as a sort of surprise, didn’t it?

 Well, this was yet another surprise. But I felt that there must be something out there as there was a great deal of tactical opportunities in the position and he might get confused. Not because he was not in good terms with tactics, but because there was too much of it. Here he could have bailed out, but in order to do so he would have had to find a very strong maneuver: 29.Qxe8 Qxf2+ 30.Kh2 Qxb6 31.Re7+ Kh6 32.Rd7! However, this opportunity is not so easy to uncover after all. I thought that I had definite chances in this position already.

Let him explain why he decided in favor of 29.Qd2. To my opinion he overlooked that 29…Rxf2 30.Qc3+ d4 31.Qc7+ can be met by 31…Rf7.

One can of course say that I was lucky, but on the other hand, I was not obliged to give away a half point in the previous game either. In general it should be added that had there been two draws, the course of the match would have been more or less logical.

I won the fourth game as I got substantial advantage out of the opening, although I was totally unprepared for this line, to be honest. Already the 10...h5 move is to be condemned as an error and the immediate 10...Kc8 was stronger. I will not claim this line to be entirely bad, because now you can have virtually any position worked out up to a draw. However, the game was clearly about two results only and I would never be fond of defending a position like that. In my opinion, this is a deteriorated version of Berlin; although it may well be that I simply misjudge the position.

Karjakin – Svidler

1.Nf3 d5 2.d4 c5 3.c4 cxd4 4.cxd5 Nf6 5.Qxd4 Qxd5 6.Nc3 Qxd4 7.Nxd4 Bd7 8.Ndb5 Kd8 9.Be3 Nc6 10.f3 

10...h5 11.0-0-0 Kc8 12.Bg5! g6 

12...a6 fails to 13.Na4!, although the move in the game took me by surprise. Was it that he missed my reply?

13.Nd6+! exd6 14.Bxf6 Rg8 15.e4 Be6 16.Kb1 Kd7 17.Nd5 Bg7 18.Bxg7

It seemed to me that the game was nearly over and I went on to let some of my advantage evaporate. It was essential that I moved 18.Bh4! in this position. That is I saw the move itself, but overlooked that 18…f5 runs into 19.Nf4! 

18…Rxg7 19.Bb5

I had the impression that I had all his pieces tied down very nicely: 19…a6 fails to 20.Nb6+ and 19…g5 bumps into 20.Nf6+ followed by Nxh5. While I failed to find any ideas for him, he came up with a very strong rejoinder 19…Kd8!


It appears that I should have better dropped my knight back to f4. I thought that I was keeping my advantage here as well, but it proved to be rather insignificant. 

20…Bxd5 21.Rxd5 Kc7

This is yet another strong move and it should be noted that he was defending very decently in general. My initial plan was to transit into a rook ending, but then I realized that this was selling it for too cheap a price.

22. Rc1 Re8 23.Rd4 Re5 24.Ba4

After 24.Bxc6 bxc6 25.Rdc4 c5 26.f4 Re7 27.e5 g5 28.exd6+ Kxd6 the resulting position was very close to a draw. I had only six minutes on my clock, but it had no significance since it was an endgame already. I realized that my last chance was to keep up the tension by having my bishop retreated, upon which the game was at least not over yet. This way I provide him with more opportunities to commit errors.

24…b5 25.Bb3 Rc5 26.Rd5 Rxc1+ 27.Kxc1 a6 28.Rd3



I saw the 28...f5! advance, but thought that I was still better, although it is obvious that I have allowed the lion’s share of my advantage disappear. However, after 28...g5 I already started to feature definite winning chances: the king marches to the center, and later the f3-f4 idea comes to the forefront. Black is going to have an unpleasant defence to deal with.

29.Kd2 h4 30.Rc3 Kb6 31.Rd3 Kc7 32.Ke3 f6 33.Rc3 Kb6 34.Rd3 Kc7 35.Rc3 Kb6 36.Bd5 Ne7 37.Kd4 Rh7 38.Be6 Rh8 39.a3 Rd8 40.Rc2 Rh8 41.Rf2 Ng6 42.Kd5 Rd8 43.Bf5 Nf4+ 44.Kd4 Re8? 

However, I needed to go 44...d5 45.e5 fxe5 46.Kxe5 d4. It was difficult to evaluate this position over the board, but I felt that White was better. On the other hand, the d4-pawn is strong. Naturally enough, I expected exactly this continuation to take place, whereas 44 ... Re8 was apparently linked to some calculation error. Now, after 45.g3 Ne6+ I already have a forced win, and I'm glad I was able to bring this advantage home with accurate play.

46.Bxe6 Rxe6 47.Kd5 Re5+ 48.Kxd6 hxg3 49.hxg3 g4 50.fxg4 Rxe4



This is a robust move, whereas after 51.Rxf6 Rxg4 there was still a lot of fight ahead. 

51…Re3 52.Rxf6 Rxg3 53.Ke5+ Kb7 54.Kf5 Rb3 55.g5 Rxb2 56.g6 Rg2 57.Ke6 Black resigns.

– So, you were back in the match by winning two games in a row, and the second game was won in a very convincing manner. What were your thoughts prior to the start of the tiebreak?

– I would like to clarify at once that I did not experience any easy illusions about the tie-break. I thought that the chances are 50/50 because Svidler is no rookie whom you can easily knock out of the saddle. He has accumulated a wealth of experience and is a seven-time Russian Champion. He has already lost enough games in his life, so that it was nothing tragic that had happened to him. That was, no doubt, frustrating for him, but the match went on, and I knew the tiebreak was going to be extremely challenging.

In game one I caught him in the opening in a certain way... No, it would be wrong to say like that, because it amounted to nothing for me in the end, but at least he spent a lot of time thinking. In principle, we were both out of our opening books quite early in the game. He overtook the initiative at one point, but I was able to profit from his lengthy thinking in the opening. Objectively speaking, his position was very good, although he was down a pawn, while I was keeping up the tension. I knew that if I somehow managed to plant my knight on g4, then it could become potentially dangerous for him. I have not analysed this game with my computer yet, but it seemed to me that I performed quite decent in pure terms of using all the practical chances available to me. In the end it all petered out to an opposite-coloured bishop ending with an extra pawn for White. At first I was not sure that it was drawish, but then I realized that the position held without any problems. Nevertheless, I managed to uncover an idea, which worked in the game.

Karjakin – Svidler

This was the last moment to play 78... f4!, whereas following any of my recaptures he had quite a simple type of fortress. Although I later read in the Internet that Black seemed to still be able to bail out, it was already crazy as it was necessary for him to flee his king over to the queenside...

78...Bd7? 79.Bf4 Be6 80.d5+! Bxd5 81.Kc8 Bb3 82.Kd8 Bc4 83.Ke7 Bb3 84.e6 Bc4 85.Kf6 Bb3 86.Bc1 Bc4 87.Ba3 Bb3 88.e7 Kd7 89.Kxg6 Black resigns. 

– You've won two games in a row, and now you had a match point already...

– Yes, and at this moment Peter pulled himself together, one couldn’t miss it from the look in his eyes. The defeat upset rather than surprised me.

In this game I sacrificed my a4-pawn and believed to have obtained a great compensation for it. But then I probably sold it cheap as it was necessary to maintain the tension. Meanwhile, I won the pawn back and thought that I would end up making a draw somehow. But he actually had a strong knight against my bishop with all pawns on the same flank – it appears that my position was unpleasant. Peter went on to win in a pretty confident fashion.

In tiebreak rapid games I had to immediately go for 1.e4 so as to play the line very well familiar to me. 1.Nf3 is a good move, but I do not feel it to such an extent as to trust myself in that I could find my way through unfamiliar situations especially being in the state of extreme fatigue. It would be somewhat overly optimistic. Svidler allowed d4-d5; I was superficially familiar with this type of position but mixed the plans altogether. I acted poorly, and lacked one tempo to finish development. And then I blundered to cap it all, although, frankly speaking, my position had already become unpleasant even without this blunder.

– On the other hand the situation was not unfamiliar for you since you once again needed to win as Black!

– Yes, yes! Oddly enough, I won quite confidently, and played a good quality game at that. I had prepared this line for the third classical game; although the associated plans are relatively simple, but he did not have time to place his pieces properly. He overdid it, and things went wrong for him. The game featured approximately the same scenario as in the previous games when I really messed it up as White. He lost a pawn, then another one, and finally gave up.

It was a key game. You need to understand our state of mind as we produced eight successful games, crazy sequence of games, crazy tension, and then it was followed by the blitz games. In this format, in my opinion, anyone would have generated errors no matter the player. Therefore, the last two games are difficult to comment in earnest. You can simply pronounce a "diagnosis" that my nerves proved stronger. It is because of this that I finally won the match.

Kirsan Ilyumzhinov, Sergey Karjakin and Peter Svidler


– What can you say about the tournament in general?

– This is one of my biggest successes not in terms of the quality of the games, but in terms of the number of times I managed to come back. It is always worth a lot! The victory in the World Cup is very prestigious indeed.

Speaking of revenge games I can recall winning the World Rapid Chess Cup in Odessa in 2010, when I found myself in a must-win situation four times. And besides that there also was an "Armageddon" in which I did not lose as Black. That actually makes it as many as five times. Yet it was the rapid chess that falls short in comparison in terms of inherent tension and range of emotions, whereas here I indeed go through the feelings of euphoria.

– It has been a long tournament, hasn’t it?

– Way too long, to be honest. I remember reaching the semi-finals in the 2009 World Cup, where I lost to Boris Gelfand, who eventually finished winning the tournament. Even back then I spoke at a press conference about the need in more days-off.

– Do you mean that the format of event featuring 128 participants is OK, just a few extra days-off need to be added into the schedule?

–Yes, they do need to be added.

The finalists together with the Russian Chess Federation President Andrey Filatov


– Recently Magnus Carlsen has come up with a proposal of contesting the world title in a knock-out tournament system. It would be interesting to know your opinion on this issue.

– I would of course never mind to be crowned World Champion upon completion of this tournament! Objectively speaking, however, this event fails to deserve to be awarded such an exalted status, because so much is decided during a tiebreak. When we contest the title of the classical World Chess Champion that means its fate should be decided in the classical duels. That means that a tie-break is possible, but, roughly speaking, following completion of twelve games rather than two. With all due respect to the World Cup and all my joy of victory, I believe that to be wrong.

– Do you mean to say that the current system featuring the Candidates Tournament and the World Championship match is reasonable?

– Yes, it is very reasonable. It functions and attracts sponsors and, honestly speaking, I see no reason to introduce changes into something that produces results.

– Who was out there supporting you during this lengthy challenging marathon whom you would like to express your gratitude to?

– Quite naturally it is Galiya who comes first. We have become many now, as you know; at the time of this magazine publishing there will be even more of us! So this victory is, of course, dedicated to her. It will be sort of our multiple-child allowance. (Laughing.)

Pictures by Eteri Kublashvili