GM Evgeny Miroshnichenko comments on the games of the 3rd round

Detailed analysis of the FIDE Candidates tournament third round games by GM Evgeny Miroshnichenko.

Shakhriyar Mamedyarov – Vishy Anand
Slav Defense D11

1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Qc2 dxc4 5.Qxc4 Bg4. 5…Bf5 is much more popular, and Anand also employed it, but the text-move creates more concrete problems for White.
6.Nbd2. The most popular continuation, which, however, totally lacks ambition. Earlier Shakhriyar tested 6.Nc3 Nbd7 7.e4 Bxf3 8.gxf3 e5 9.Be3 exd4 10.Bxd4 Bd6 11.0–0–0 Qc7 12.Ne2 0–0 13.Qc2 with success, and perhaps such manner of play suits his creative style better (Mamedyarov-Inarkiev, Rogaska Slatina 2011).

6…Nbd7 7.g3 e6 8.Bg2 Be7 9.Ne5 Bh5 10.Nxd7 Nxd7 11.0–0 0–0 12.Nb3 a5 13.a4 Bb4 14.e4.


14…e5. A novelty from Vishy – Black is better prepared for opening the game! Ivanchuk-Vallejo (Istambul 2012) saw 14…Qe7 15.Be3 Rfd8 16.f4 Kh8 17.Rf2 f6 18.Bd2 e5 19.Bxb4 axb4 20.fxe5 fxe5 21.d5 Rac8 22.Rc1 Nb6 23.Qc5 Qxc5 24.Nxc5 cxd5 25.Bh3 Ra8 26.a5 Nc4 27.Nxb7 Rf8 28.Rxf8+ Rxf8 29.exd5 Nxb2 30.d6 Bf3 31.Rc8 Rxc8 32.Bxc8 Bc6 33.d7 Bxd7 34.Bxd7 Nc4 35.Be6 b3 36.Bxc4 b2 37.Ba2, and White won convincingly.

15.Be3 exd4 16.Bxd4 Kh8! Black prepares f7-f6 and, possibly, Bh5-f7.

17.e5? This attempt to prevent f7-f6 is way too optimistic. Already on the next move White will have to compromise his kingside in order to protect the pawn. Instead of that he could dry the position out a bit by 17.Bc3 Bxc3 18.Qxc3. Black has several options, for example, 18…f6 19.Bf3 Bf7 20.Rfd1 Bxb3 21.Qxb3 Nc5 22.Qe3 Qe7 with an approximately even game.

17…Re8! 18.f4.


18…f6! 19.exf6 Nxf6. Black’s bishops are preventing White’s rooks from entering the game, and due to the …Bе2 threat White’s next move is almost forced.

20.Bf3 Bxf3 21.Rxf3 Re4?! The only inaccuracy of the ex-World champion! Much more ruthless is 21…c5, and after 22.Bc3 even 22…Re4 works – 23.Bxf6 Rxc4 24.Bxd8 Rxd8 25.Rc1 Rxc1+ 26.Nxc1 c4, and White’s defensive task is very difficult due to a horribly misplaced knight.

Also promising is the unexpected 22…Bxc3 23.bxc3 b6. White has problems with his horse on b3, for example, 24.Rd3 Qe7 25.Nd2 Rad8 26.Rxd8 Qe3+! 27.Kg2 Rxd8, forcing the ugly-looking 28.Nf1, and now 28…Qe7 solidifies Black’s comfortable advantage.


22.Re3?! Mamedyarov misses 22.Qd3!? Of course, Black is better here as well: after 22…Qe8 23.Re3 Rxe3 24.Qxe3 Qh5 he threatens …Ng4, and 25.Bxf6 gxf6 does not solve White’s main problem with his poor fellow on b3, who cannot hope to become a real knight.

22…Rxe3 23.Bxe3 Qe8 24.Bb6. More tenacious is the immediate 24.Bd4, although White’s position after 24…Rd8 remains difficult, because he cannot consolidate his pieces easily. On 25.Rc1 Black has 25…c5! 26.Nxc5 Bd2 27.Bxf6 (27.Rc2? Rxd4 28.Qxd4 Be3+) 27…gxf6, and White is helpless.

24…Qh5! A signal for action! White’s pieces got stuck on the queenside, and Black launches a direct attack.

25.Bd4 Re8 26.Rf1 Ng4 27.Qc2.


27…c5! 28.Nxc5. Other moves allowed a fork on e3, e. g., 28.Bxc5 Bxc5+ 29.Nxc5 Ne3–+.

28…Rc8 29.Rd1. This attempt of indirectly protecting on с5 does not work because white pieces are seriously overloaded.

29…Bxc5. A very practical decision. One could “fall into the trap” – 29…b6 30.Bxg7+ Kxg7 31.Ne6+ Kf6!, and White can resign.

30.Bxc5 h6. White is helpless! Shakhriyar played 31.Kh1, but realized that his last trap is way too naïve, resigned before Anand could reply. Almost all knight moves win for Black – 31…Nf2+, 31…Ne3, 31…Nxh2, 31…Nf6, but, of course, not 31…Qxc5?? 32.Rd8++-.


Veselin Topalov – Levon Aronian
Ruy Lopez C88

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0–0 Be7 6.Re1 b5 7.Bb3 0–0. Despite losing to Anand in the first round, Aronian sticks to his guns, inviting White to the Marshall attack.

8.a4. The most popular way of avoiding 8.c3 d5 9.exd5 Nxd5 – even such a renowned theorist as Topalov recognizes Aronian’s superiority in the Marshall.

8…b4 9.d4 d6 10.dxe5 dxe5.


11.Nbd2. Levon never faced this rare move before. Here are the most recent examples from his practice:

11.Qxd8 Rxd8 12.Nbd2 h6 13.a5 Bc5 14.Bc4 Ng4 15.Re2 Be6 16.Bxe6 fxe6 17.h3 Nf6 18.Re1 Rab8 19.Nc4 Rb5 20.b3 Bd4 21.Bb2 Rc5 22.Ra2 Bxb2 23.Rxb2 Ne8 24.Ra2 Nd6, and Black is already slightly better, although the game ended in White’s favor, Carlsen-Aronian, St. Louis 2013;

11.Bg5 Qxd1 12.Rxd1 Nxe4 13.Bxe7 Nxe7 14.Nxe5 Nc5 15.Nd2 Nxb3 16.Nxb3 f6 17.Nd7 Re8 18.Na5 Nf5 19.Nc5 Nd6 20.Nd3 with an advantage to White, Giri-Aronian, Antalya 2013.

11…Bc5 12.h3 Bb7 13.Qe2 Nd4 14.Nxd4 Bxd4.


15.Bc4. This move is a novelty. Curiously, the only game that achieved the position above, Jansa-Razuvayev 1986, ended in a draw immediately after 15.Qf3 Kh8. White transfers the bishop to d3 in order to protect the е4-pawn, finally freeing the d2-knight.

15…a5 16.Bd3 Nd7. Levon’s decision is very concrete – Black is ready to part with the dark-squared bishop, but wants to utilize the opponent’s underdevelopment in return. On 16…Qe7 White probably planned 17.Nf3 Bb6 18.Nh4!?, and Black has to weaken his position by …g6.

17.Nf3 Nc5 18.Nxd4 exd4 19.Bf4 Re8. Calling for trouble! Topalov was skeptical about this move at the press-conference and recommended capturing on d3. Indeed, after 19…Nxd3 20.cxd3 c5 21.Rac1 Rc8 followed by …Bа6 Black is just fine, but Aronian’s move is also quite good, according to the computer.


20.Qh5!? Qe7. Here 20…Nxd3 is wrong – 21.cxd3 Ba6 22.Qxa5 Bxd3 23.Qxb4, and White is much better.

An alternative to the text is a careless recapture on е4. Black can survive it, but it requires a lot of tightrope skill: 20…Nxe4 21.f3!? g6 (after 21…Nf6 22.Qc5 White regains a pawn and retains pressure) 22.Qh6 Nf6 (22…Nc5?! 23.Bg5 Qd6 24.Bc4! Ne6 25.Bf6 Qf8 26.Qh4, and Black is in a world of hurt) 23.Qh4!? Rxe1+ 24.Rxe1 Nd5 25.Bg5 f6 26.Bh6 – defending as Black is rather troublesome.

21.Bb5. Topalov decides to lure the pawn to с6, where it restricts the bishop, but it seems 21.Bc4 immediately is stronger. The path to equality for Black is very narrow – 21…Nxe4 22.f3 Qf6 23.Qg4 Nd6 (23…Bc8 is bad due to 24.Qg3! Ba6 25.Rxe4 Rxe4 26.fxe4 Bxc4 27.Bxc7, and White has a big advantage due to Bе5 threat and Black’s queenside pawn weaknesses) 24.Bd3 h6! 25.Qg3 Nf5 26.Bxf5 Qxf5 27.Bxc7 with equality.

21…c6 22.Bc4 Ba6!? It doesn’t seem like Levon Aronian overlooked the opponent’s reply, which leads me to believe that this was provocation. He could easily insert 22…g6, and after 23.Qf3 Ne6! 24.Bh2 Ba6 I don’t see how White can utilize Black’s kingside weaknesses.


23.Bd6!? Topalov accepts the challenge! The position after 23.Bxa6 Rxa6 seems promising for Black, who simply puts a knight to е6 and advances the c-pawn.

White can try 23.Ba2 d3 24.Bd6 Qxd6 25.Bxf7+ Kf8 26.Bxe8 Rxe8 27.cxd3, however, Black gets rid of the pin on d-file after 27…Re5 28.Qxh7 Bxd3 29.Rad1 by 29…Qh6!, and probably has better chances after 30.Qxh6 gxh6.

23…Qxd6 24.Bxf7+ Kf8 25.e5 Qd7 26.e6 Nxe6 27.Bxe8 Rxe8.


28.Qxh7. A draw looks logical after this move, with only one road bump (but more of it later). 28.Qxa5 is a serious alternative to the text-move. After the basically forced sequence 28…Bb7 29.Qxb4+ c5 it looks like Black has sufficient counterplay, but 30.Qb6! stops 30…Qd5 31.f3 Nf4 due to 32.Rxe8+ Kxe8 33.b4! Nxg2 34.Qxc5! Qxc5 35.bxc5, and White’s advantage in the endgame is almost decisive.

The computer insists on 30…Qc6 31.Qxc6 Bxc6, but here White’s chances are higher as well, at least in a game between humans.

Another move suggested by the machine – 30…Kg8, is hard to explain, but somehow it seems to be the best practical decision – after 31.a5 Bc6 32.a6 Qd5 33.f3 Rf8 Black’s threats should not be treated lightly.

28…Qd5! 29.f4! d3 30.f5 d2 31.Qh8+.


31…Kf7. Black had an amazing way to continue playing: 31…Ke7 32.Qh4+ g5!! 33.Qxg5+ Kd7 34.Rxe6! (34.fxe6+ Kc8 35.Qe3 dxe1Q+ 36.Rxe1 Bb7 gives Black good winning chances) 34…d1Q+ 35.Rxd1 Qxd1+ 36.Kh2 Rxe6 37.fxe6+ Kxe6, and here White holds only by 38.Qg6+! Kd7 39.Qf5+, forcing the perpetual, as after 39…Kd6 40.Qxa5 White picks up a b4-pawn as well.

32.Qh5+ Kf8 33.Qh8+ Kf7 34.Qh5+ Kf8 35.Qh8+. Draw.


Dmitry Andreikin – Sergey Karjakin
Ruy Lopez C65

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3. This modest move quickly gains popularity, as White fails to show up with anything in the main line.

Andreikin’s main Berlin triumph occurred in Nizhny Novgorod 2013: 4.0–0 Nxe4 5.d4 Nd6 6.dxe5 Nxb5 7.a4 Nbd4 8.Nxd4 Nxd4 9.Qxd4 d5 10.exd6 Qxd6 11.Qe3+ Be6 12.Nc3 a6 13.Rd1 Qc6 14.Rd3 Rc8 15.Ne2 Bc5 16.Qg3 f6 17.Be3 Bd6 18.Bf4 Bxf4 19.Nxf4 0–0 20.Rc3 Qd6 21.Nxe6 Qxe6 22.Rxc7 Rxc7 23.Qxc7, and he won a pawn and then a game against Vladimir Kramnik.

4…Bc5 5.c3 0–0 6.0–0 d6 7.Nbd2 Ne7 8.d4 exd4 9.cxd4 Bb6 10.h3. Karjakin never faced this move before. Earlier he confronted 10.Re1 Bg4 11.h3 Bh5 12.Qb3 d5 13.e5 Nd7, and White is somewhat better, Svidler-Karjakin, Loo 2013, and 10.Bd3 Bg4 11.h3 Bh5 12.Nb3 d5 13.e5 Ne4 14.Qc2 Nc6 15.Bxe4 dxe4 16.Qxe4 Bxf3 17.Qxf3 Nxd4 18.Qxb7 with a very sharp game, Vallejo-Karjakin, Sao Paulo/Bilbao 2012.

10…d5 11.e5 Ne4 12.Bd3 Bf5 13.Qe2 Nc6 14.Rd1.


14…Bxd4. A novelty. Svidler-Carlsen, Gjovik 2009 saw 14…Nxd4 15.Nxd4 Bxd4 16.Bxe4 dxe4 17.Nb3 c5 18.Be3 Qe7 19.Nxd4 cxd4 20.Bxd4 with a draw.

15.Nxe4 dxe4 16.Bxe4 Bxe4 17.Nxd4 Qd5 18.Nxc6 Qxc6. Compared to the game above, this ending is slightly less drawish, but neither side managed to gain anything real.

19.f3 Bd5 20.b3 a5 21.Ba3 Rfd8 22.Rac1 Qb6+.


23.Kh1. The computer insists on 23.Qf2 Qxf2+ 24.Kxf2, with certain initiative for White, although after 24…Rd7 25.Bd6!? c6! followed by f6 Black should equialize.

23…Be6. A silent draw offer. Once can understand a peaceful mood of the players, recalling their results in the previous round.

24.Rxd8+ Rxd8 25.Rd1 Rxd1+ 26.Qxd1 h6 27.Qd8+ Kh7 28.Qd3+ Kg8 29.Qd8+ Kh7 30.Qd3+ Kg8. Draw.


Peter Svidler – Vladimir Kramnik
English opening A35

A highly tense battle between the two Russian players! Peter Svidler missed a real chance to jump to the joint first place right before a day off. Let us skip the opening part, because the main events of this game took place around the 40th move.

1.c4 c5 2.Nf3 Nf6 3.Nc3 Nc6 4.g3 g6 5.d4 cxd4 6.Nxd4 Bg7 7.Bg2 0–0 8.0–0 Nxd4 9.Qxd4 d6 10.Qd3 Be6. At least four World Champions preferred 10…a6 – Carlsen, Kasparov, Anand and Karpov.


11.Bd2. After 11.Bxb7 Rb8 12.Bg2 Black rejects the materialistic 12…Qa5 13.b3 Rxb3 14.axb3 Qxa1 (White’s position after 15.Bd2 Qa6 16.Nb5 is much better – his bishop goes to с3, his knight – do d4) and chooses 12…Rc8 13.Nd5 Ng4!? with compensation instead.

11…Qc8 12.b3 Bh3 13.Rac1 Bxg2 14.Kxg2 Qc6+. This natural move is a novelty. Earlier Black tested 14…e6, which also doesn’t promise him an easy life after 15.Rfd1 Rd8 16.Bg5 Rd7 17.Qf3 Ne8 18.Rd3, and White exerts unpleasant pressure, Neverov-Dvoirys, St. Petersburg 1995.

15.f3 e6. Black has to create himself a weakness on d6 in order to control the d5-square. The computer suggests waiting moves like 15…Rac8 16.Rfd1 Rfe8 17.Be3 Qa6, but White still has a lasting advantage due to his spatial gains.

16.Rfd1 Rad8 17.Bf4 Rd7.


18.Qe3! A very strong maneuver, vacating a square for the rook. One can already assess the position: Black has no counterplay and must defend stoically.

18…b6 19.Rd3 Rc8 20.Qd2 Ne8. Black has no other options. The desired 20…d5 fails here – on 21.cxd5 Black must recapture with a pawn, as 21…Nxd5 22.Nxd5 Qxc1 23.Ne7+! Rxe7 24.Qxc1 Rxc1 25.Rd8+ Bf8 26.Bh6 leads to a mate in three.

After 21…exd5 22.Nxd5 Qxc1 23.Qxc1 Rxc1 24.Nxf6+ Bxf6 25.Rxd7 Rc2 26.Rd2 White is just a pawn up.

21.e4 a6 22.e5! h6!? 23.h4! White continues to ask tough questions. Also unpleasant for Black is 23.Bxh6 Bxe5 24.Bf4, however the text-move is more in Peter Svidler’s style, who likes to add pressure and never shies away from complications.

23…Rcd8 24.Rd1. Due to Nе4 threat Black has to search for some counterplay.



25.c5! Once again the best move. 25.cxb5 axb5 26.Ne4 (after 26.Bxh6 Bxe5 27.h5 Black trades queens 27…Rc8 28.Ne4 Qc2 with good drawing chances) 26…d5 27.Rc3 Qb6 28.Nc5 Rc7 29.b4 Ra8 gives White nothing.

25…Qxc5. 25…d5 doesn’t deserve much attention – after 26.b4 a5 27.a3 axb4 28.axb4 Black’s position is too passive, and seizing the a-file by 28…Ra8 does not make his life easier – 29.Bxh6 Bxe5 30.h5 Ng7 31.hxg6 fxg6 32.g4! with a significant advantage for White.

26.Ne4 Qb6 27.Nxd6?! Peter Svidler mentioned a stronger alternative – 27.Be3! White utilizes a pin on the d-file. After 27…Qb8 28.Nc5 Bxe5 29.Nxd7 Rxd7 30.Bxh6 the game enters the technical stage. White’s victory may be far from obvious, but after the text-move the evaluation almost becomes unclear!



28.h5?! Not an optimal choice, but it’s easy to notice that White has too many tempting opportunities.

One of the computer recommendations – 28.Nc4 Rxd3 29.Nxb6 Rxd2+ 30.Rxd2 Rxd2+ 31.Bxd2 – looks almost craven, however after the white knight arrives to d7, Black will have serious problems with regrouping his pieces, especially the king.

The crafty 28.Be3 Qb8 29.Bxh6 Nxd6 30.Bg5!! is almost impossible to find when there are any alternatives. White has an advantage after 30…Ne4 31.Rxd7 Nxd2 32.Rxd8 Qxe5 33.R8xd2, but the struggle continues.

After the immediate 28.Bxh6 Nxd6 29.Bg5 Ne4 Black suddenly wins!

28…Nxd6! At least getting rid of the passive е8-knight! Naturally, Black does not fall for 28…g5 29.Bxg5 hxg5 30.Qxg5+ Ng7 31.Ne4!! with an inevitable mate.

The next dozen of moves were made under growing time pressure.

29.exd6 g5 30.Be5 Rc8. The machine recommends passive approach: 30…Bg7!? 31.Bxg7 Kxg7, however, accepting a major-piece ending with passive rooks is something Black does not want, especially considering the time trouble.

31.Rc1 Rxc1 32.Qxc1. Exchanging a couple of rooks adds strength to the d-pawn – Black’s blockade is not that firm.

32…Qb7 33.g4 b4 34.Qc4 Bg7 35.Bg3 Qb5 36.Be1?! White misses a greater part of his advantage, if not all of it. Black had no threats, therefore it was advisable to reach the time control by, for example, 36.Rd2, setting up small traps along the way (36…Bc3?? loses at once by 37.Qc8+ and а4).

Transposing to an endgame by 36.Qxb5 axb5 37.f4 is interesting. After 37…gxf4 38.Bxf4 Bf8 39.Be5! Black is completely paralyzed. Possibly Peter rejected it because he wanted more.


36…Qe5! The queen suddenly becomes very active. Now Black should not lose, if he doesn’t make more mistakes.

37.Bg3. Of course not 37.Qc8+ Bf8 38.Qxd7? Qe2+ 39.Bf2 Qxd3, and White is a pawn down.

37…Qe2+ 38.Bf2 Bf8 39.Qxa6 e5? An attempt to compromise White’s kingside structure by е5-е4 is objectively wrong. 39…Qe5 is stronger – Black wins the d6-pawn, while his b4-pawn negates White’s queenside majority. After 40.Qa8 Rxd6 41.Qb8 Rd5 42.Qxe5 Rxe5 43.Rd8 Ra5 the game becomes even.

40.Qc4 Qxa2 41.Qc6!? It is hard to blame this move, especially since Black’s spectacular defense is hard to spot, but the strongest move here is 41.Qxb4. After 41…Qa6 42.Rd5 Rxd6 43.Rxe5 White keeps an extra pawn and some winning chances.


41…e4! 42.fxe4 Qe2. Black finally begins to create threats to the enemy king, however, after 43.Rf3 Rxd6 44.Qe8 f6 45.e5 his position looks lost.


Having spent about 15 minutes, Vladimir Kramnik uncorked a phenomenal defense.

45…f5!! The naïve 45…Qxe5 loses nicely after 46.Qxe5 fxe5 47.Rxf8+ Kxf8 48.Bc5 Ke7 49.Bxb4, and the pawn ending is easily won for White.

46.gxf5 Rf6!! And it turns out that White cannot win!

47.Kg3 (47.Qb8? Rxf5 48.Rxf5 Qg4+) 47…Qe4 48.Bc5 Qe1+ 49.Bf2. After 49.Kg2 Qe4!? White has a choice: it’ either move repetition or perpetual after 50.Bxf8 Qe2+ 51.Rf2 Qg4+.

49…Qe4 50.Bc5 Qe1+ 51.Bf2. Draw.