<INTRODUCTION>

Anand vs Kramnik, 2008 is the <ninth game of the 2008 title match>.

Quantitative mapping of this game between these players follows. Figures in brackets immediately after each move are the corrected engine evaluations generated on the reverse slide that followed the initial forward slide originating from the first move of the game. Some evaluations are bolstered by analysis. This smoothed out many, but not all fluctuation in the engine’s evaluations, especially in the opening.

General methods used are described in the bio at the <Bridgeburner chessforum>.

The evaluation values in the opening come at the end of reverse slide back to the starting position from the last move of the game, and are included for completeness rather than in the interests of accuracy. Engine preferences are included throughout the game where they differ from players preferences except in the well trodden opening.

Some analysis is included to provide some idea of the reason for the engine preferences where they didn’t coincide with the moves played. Where the differences are minor, the accompanying analysis is brief and indicative, where the differences are pivotal, more extensive analysis is provided.

<Summary>

A must-win for Kramnik, he pressed hard with the Black pieces, but came up short, missing several moves that would have provided initiative and advantage. Kramnik however did not miss any winning moves as such, as there were none. Anand for his part, played a single <bad move> in time pressure, although it was not a <blunder> that would have objectively created a lost position for him. Although he played no <dubious moves> as such, he played a series of minor inaccuracies which allowed his game to drift into a difficult position on the several occasions Kramnik missed his cue.

This game is an example of excellent fighting chess. Despite time pressure in a number of complex positions, neither side stumbled seriously enough to upset the basic equilibrium of the game.

THE GAME

Comments have been lifted from three sites to accompany this game.

First is from http://www.uep-chess.com/cms_englis... where annotations and comments are supplied by the Cuban GM Amador Rodriguez-Cespedes.

GM Zsuzsa Polgar commented during the game from her blogsite at http://susanpolgar.blogspot.com/200...

IM Malcolm Pein also provided comments for chessbase.com at http://www.crackteam.org/2008/10/26...

The time on each player’s clock for each move of this game can be seen at http://www.crackteam.org/2008/10/26....

GAME MOVES 1-6

<1. d4> (=0.13) <1…d5> (=0.13)

<2. c4> (=0.13) <2…e6> (=0.13)

<3. Nf3> (=0.13) <3…Nf6> (=0.13)

<4. Nc3> (=0.10) <4…c6> (=0.24)

<5. Bg5> (=0.19) <5…h6> ( 0.31)

<6. Bh4> (=-0.11)

Pein: <Surprising as 6.Bxf6 Qxf6 7.e3 is the solid Moscow variation and more suited to the match situation. However Anand's strategy has been to play for complications most of the time.> NB: Anand played this move instantly.

<6. Bf4> was played in S Ovsejevitsch vs J Todorovic, 2003. White survived a poor opening to mount a mating attack, only to blunder a piece and then a mate in one, probably during a time scramble.

GAME MOVE 6:

<6…dxc4> (=-0.11)

Polgar: <Kramnik is trying to dictate this game with something new in this match. Let's see when he will uncork a novelty or will Anand beat him to the punch?>

Pein: <Kramnik has to accept the challenge and he enters the sharp Anti Moscow Gambit. He must have been delighted to get such an unbalanced position.>

GAME MOVE 7:

<7. e4> ( 0.35) <7…g5> ( 0.35)

Played 95% of the time. <7…Bb4> was played in Kasparov vs Korchnoi, 1995, which White won in 27 moves despite mishandling the opening.

GAME MOVES 8-9:

<8. Bg3> ( 0.35) <8…b5> ( 0.30)

<9. Be2> ( 0.30) <9…Bb7> ( 0.26)

GAME MOVE 10:

<10. Qc2> ( 0.38)

Rodriguez: <This came as a surprise, as certainly this move is not one of the most popular in the present position [ 10.e5 was the early continuation, not in fashion anymore Nh5 11.a4 a6 12.Nxg5 Nxg3 13.Nxf7 Kxf7 14.fxg3 Kg8÷ Kramnik vs Anand, 1997, Belgrade] [ 10.0-0 is the way to go for most games being played nowadays Nbd7 11.Ne5 Bg7 12.Nxf7!! ( 12.Nxd7 Nxd7 13.Bd6 a6 14.Bh5 Bf8 15.Bxf8 Rxf8 16.e5 Qb6 17.b3 0-0-0÷ Kramnik vs Anand, 2007, Mexico City) 12...Kxf7 13.e5 Nd5 14.Ne4 Ke7 15.Nd6 Qb6 16.Bg4 Raf8 17.Qc2 was the spectacular game Topalov vs Kramnik, 2008, Wijk aan Zee] [ 10.h4 is by far the most explored move ] [ 10.Ne5 is also a very popular continuation]>

Pein: <10.h4 is more cutting edge.>

Note: Also the most popular: Opening Explorer

<Main engine preference>: <10.h4> ( 0.26): <10…g4 11.Ne5 b4 12.Na4 Nxe4 13.0-0 Nd7 14.Nxc4 Nxg3 15.fxg3 h5 16.Qd2 Be7 17.a3 bxa3 18.b4 0-0 19.Nc5 Nxc5>

GAME MOVE 10:

<10…Nbd7> ( 0.26)

Rodriguez: <At the press conference Kramnik said that this time he wanted to avoid being behind in the clock as occurred in games 3 and 5. He reacted quickly saving time, and he played well indeed.>

Polgar: <The most popular response here is 11.Ne5. However, 11.Rd1 or 11.O-O are fine as well. Perhaps Anand will choose a less popular line.>

<10…h5?!> was played in C Praveen Kumar vs A Nasri, 2008, 2008 during the Dubai Open, and Black got the better of White’s timid response (<11. h3?!> instead of <11. Nxg5!>, only to blunder and lose in 24 moves.

GAME MOVE 11:

<11. Rd1> ( 0.34)

Polgar: <Anand chose a somewhat less popular continuation.>

<11. h4?!> was played in Tan Zhongyi vs R Lufei, 2001, leading to an eventful draw.

<Main engine preference>: <11.Ne5> ( 0.26): <11…Nxe5 12.Bxe5 Rg8 13.0-0-0 Nd7 14.Bg3 Qa5 15.d5 e5 16.Bg4 Nf6 17.Bf5 Qc7 18.Kb1 h5 19.dxc6 Bxc6 20.f3>. <11. Ne5> is also the most popular move in theory.

GAME MOVE 11:

<11…Bb4> ( 0.28)

Rodriguez: <This move was new, however it did not lead to a new position as now the computer showed some games where White had played 12.e5 and 12.Ne5, the road Anand follows next.>

<11…Be7> was successfully tried in the Armenian Championships in G Sargissian vs A Anastasian, 2005, where Black won a positional classic.

<11…Qb6> (which is also one of the engine’s preferred moves), was unsuccessfully used (through no fault of the move) during the Feldbach Open in A Chernin vs Pavasovic, 1997.

<Main engine preference>: <11…a6> ( 0.34): <12.Ne5 Bg7 13.0-0 Qe7 14.f4 gxf4 15.Bxf4 c5 16.d5 exd5 17.Nxd7 Qxd7 18.Be5 Qe7 19.Bxf6>

GAME MOVE 12:

<12. Ne5> ( /0.28)

Polgar: <This is definitely what Kramnik needs, a sharp game, to have a chance to make a comeback.>

GAME MOVE 12:

<12…Qe7> ( 0.28)

A novelty. At the 34th Olympiad in Istanbul, <12…Rg8> was unsuccessful in Van Wely vs Dao Thien Hai, 2000. Black had at least an even game until <13. 0-0 Qe7 14. a4 a6 15. Nxd7 Nxd7 16. d5>, but here instead of <16…Qf6!> with a good game that would allow Black to gradually increase pressure on the king side and in the centre, Black played <16…Nf6> allowing <17. d6>, throttling his game; although it is not quite a forced loss, it is extremely difficult to defend.

Polgar: <This seems to be a rare move. I believe that it was recommended in an earlier analysis by German GM Christopher Lutz (who had achieved a high rating of 2655 back in July 2002). I do not have access to that article now but you can probably search for it.>

Rodriguez: <Again Kramnik plays a new move and this time it can be called an absolute novelty, leading to a new position. The move had been recommended by German grandmaster Lutz who annotated an important game played 8 years ago by Dutch GM Van Wely who was precisely Kramnik's secondant in the WCC in Mexico City, so Kramnik might have studied this position before. [ 12...Rg8?! 13.0-0 Qe7 14.a4 a6 15.Nxd7 Nxd7 16.d5 Nf6 17.d6 Qd7 18.f4 Bxc3 19.bxc3 c5 20.fxg5 Bxe4 21.Qd2 Nd5 22.Bh5 Bg6 23.Bf3 Bd3 24.Bxd5 exd5 25.Rde1+ Kd8 26.Re7 1-0 Van Wely-Dao Thien Hai, Istanbul 2000]>

Pein: <Kramnik finally gets to play a strong novelty. 12...Rg8 and 12...Nxe5 have been played before. 12...c5 Is the move Black wants to play in general terms but after 13.0-0 Bxc3 14.Qxc3 Nxe4 15.Qe3 his position is very ropey and if 15...Nxg3 16.fxg3!>

GAME MOVE 13:

<13. 0-0> ( 0.43)

Polgar: <An option for Black is 13...Rg8 to be ready in case if White plays f4. I believe GM VanWely has played this line before. VanWely has worked with Kramnik before so I would not be surprised if Kramnik is very well prepared with this line. Other possible lines include 13...h5 and 13...Nxe5 which is not a bad idea.

I am doing the commentary from Los Angeles airport. Therefore, I do not have access to all my database and information.>

Rodriguez: <13. Nxd7 was suggested by Kramnik as a possible improvement after the game>

Pein: <Now the typical thrust 13.h4 is a waste of time after 13...0-0-0.>

<Main engine preferences>: <13. Nxd7> ( 0.28): <13…Nxd7 14.a4 a6 15.h4 Qf6 16.Kf1 Rg8 17.hxg5 hxg5 18.Kg1 Qe7 19.Na2 Ba5 20.Rh6 Qf8 21.Rh3 c5 22.d5>

GAME MOVES 13-14:

<13…Nxe5> ( 0.43)

<14.Bxe5> ( 0.43) <14…0-0> ( 0.43)

Rodriguez: <The position already looks OK for Black as it is not easy for White to organize his attack on the king side. For instance, after the normal break 15.f4 Nd7! the position becomes very suspicious.>

GAME MOVE 15:

<15. Bxf6> ( 0.70)

Pein: <Anand concedes the dark squares and now he cannot be better but if 15.f4 Nd7! Threatening f6 gives Black a big advantage.>

Polgar: <A very surprising choice of move. The idea is to push f4 to open up the Kingside to exploit Black's weak King.>

<Main engine preference>: <15. h3> ( 0.48): <15…a6 16.Bxf6 Qxf6 17.e5 Qg6 18.Qxg6+ fxg6 19.Bg4 Rae8 20.Ne4 h5 21.a3 Bxa3 22.bxa3 hxg4 23.hxg4 Re7 24.Nxg5 c5 25.dxc5 Rc8 26.f4 Rxc5>

GAME MOVES 15-16:

<15…Qxf6> ( 0.70)

<16. f4> ( 0.71)

Polgar: < Black should not take the pawn. He should move his Queen out of the way with Qg6. Here is a possible line: 16...Qg6 17.f5 exf5 18.exf5 Qf6 19.Ne4 Qe7 20.f6 Qe6 and the position is unclear. 16.Qg7 is also playable.>

<Main engine preference>: <16. e5> ( 0.70): <16…Qf5 17.Ne4 Be7 18.Bf3 g4 19.Nf6+ Bxf6 20.Be4 Qf4 21.exf6 Qxf6 22.f3 Qf4 23.fxg4 Qxg4 24.b3 cxb3 25.axb3 Rad8 26.b4 Kg7 27.Bf3 Qh4>

GAME MOVE 16:

<16…Qg7> (=-0.04)

Polgar: < This is possible 17.fxg5 hxg5 18.e5. I think the opening choice in this game clears up a few issues: 1. Kramnik is not going down quietly. He is still trying hard. 2. Anand is not coasting to the finish line. He is not afraid to fight hard all the way to protect his title. Anand is spending a lot of time here.>

Rodriguez: <While many grandmaster where calculating the consequences of 16...gxf4, Kramnik played quickly this move and as Anand admitted at the press conference, around here he started to feel that something had gone wrong with his previous play.>

<Engine’s preference> <16…Qg6> ( 0.81): <17. f5 exf5 18. exf5 Qg7>:

click for larger view

Black has a significant advantage in this position, mainly because White’s compensation for the pawn has more or less evaporated, particularly as Black has two strong two bishops.

<The evaluation jump of <<0.67>> attracts an error weighting of <<0.5>>, as it meets the Project definition of a <<dubious move>>.>

GAME MOVE 17:

<17. e5> (=-0.04)

Polgar: < I have an interesting thought with 17...f5 18.exf6 Qxf6 19 fxg5 Qxg5 with an unclear position. 17...Qh7 and 17...Be7 are also playable. White is down a pawn but he has compensation with a potentially strong attack on the Kingside. Therefore, he must try to open it up. He also has the threat of Ne4 - Nf6.

On the other hand, even though Black is up a pawn, he must find a way to coordinate his pieces. The g2 pawn may be a potential target for Black. One way is to open up the g file, move his King to the h file, and get his Rook(s) to the g file.>

Rodriguez: < A critical moment where Anand is ready to deploy his pieces in the Black's king direction and Black has to show a consistent plan. The computers were in favor of advancing the f-pawn to f6 or f5 to secure the king but weakening the position. Kramnik finds something much better.>

GAME MOVE 17:

<17…c5> (=0.07)

<Main engine preference>: <17...Be7> (=-0.04): <18. fxg5 hxg5>

Pein: <! If 17...Bxc3 18.bxc3 strengthens whites centre.; and 17...f5 18.exf6 Rxf6 19.fxg5 Qxg5 20.Rxf6 Qxf6 21.Ne4 gives White sufficient play for the pawn.>

Rodriguez: <! An energetic continuation, giving back his pawn for a strong initiative.>

Polgar: <A very peculiar move, definitely not one I had expected. It is obvious that Black wants to open the h1-a8 diagonal for his Bishop. A possible line is 18.Nxb5 gxf4 19.Bf3 Bxf3 20.Rxf3 a6 21.Nd6 cxd4 22.Rxf4 Qxe5 23.Rg4+ Kh8 24.Nxc4 with an exciting position.>

GAME MOVE 18:

<18. Nxb5> (=0.07) <18…cxd4> (=0.07)

Polgar: <This has become a very complicated position. White has a number of options such as 19.Qxc4, 19.Bf3, 19.Nxd4, etc. I think it is most safe to play 19.Qxc4 to get break up Black's pawn chain in the center.>

GAME MOVE 19:

<19. Qxc4> ( 0.52)

Rodriguez: <An alternative was 19.Bf3 but after Bxf3 20.Rxf3 Rac8 21.Nxd4 Rfd8 the position looks fine for Black>

Polgar: <A good line for Black is 19...a5 protecting the Bishop on b4 and making the a6 square available for some potential lines.>

<Main engine preference>: <19. Bf3> (=0.07): <19…Bxf3 20.Rxf3 Rac8 21.Nxd4 Rfd8 22.Kh1 gxf4 23.Rxf4 Qxe5 24.Rxf7 Kxf7 25.Qh7+ Ke8 26.Qg6+ Kf8 27.Qxh6+ Kg8 28.Qg6+ Kh8 29.Qh6+ Kg8 30.Qg6+ Kh8 31.Qh6+ Kg8>

GAME MOVE 19:

<19…a5> ( 0.52)

Polgar: <A line that popped up in my head is 20.Nd6 Ba6 21.Qc2 Bxd6 22.exd6 Rfd8 and Black is fine, perhaps even slightly better. If 20.Qxd4 then 20...Rac8.>

NB: The engine agrees that Polgar’s suggestion is favorable for Black.

Rodriguez: <! Again a strong move by Kramnik and most people watching the game at the press room were really impressed. It's clear that Black has the upper hand.>

GAME MOVE 20:

<20. Kh1> ( 0.59)

<Main engine preference>: <20.Nd6 Bxd6> ( 0.52)

Polgar: <Another move I did not anticipate. 20...gxf4 21.Bf3 Bxf3 22.gxf3 Qxe5 23.Nxd4 Rac8 24.Nc6 Qf5 25.Rg1+ Kh7 with an unclear position.>

Rodriguez: <There were other moves like 20.Bf3 or 20.Nd6, but Kh1 seems best.>

Pein: <20.Nxd4 gxf4 Threatening mate on g2 21.Nf3 Rac8 22.Qxf4 Rc2 23.Rf2 Rxb2 is very good for Black.>

GAME MOVE 20:

<20…Rac8> ( 0.59)

Polgar: <An interesting move. Black wants to get his heavy artillery in play. 21.Qxd4 Bc5. This is an extremely complicated position and time may become a big factor soon.>

Rodriguez: <Better than [ 20...gxf4 21.Bf3 Bxf3 22.gxf3 ( 22.Rxf3 Qxe5 ) 22...Qxe5 23.Nxd4 when Black is still better but White has play ]>

GAME MOVE 21:

<21. Qxd4> ( 0.59) <21…gxf4> (=0.09) threatening mate.

Polgar: <The idea is after 22.Bf3, Black has 22...Ba6 if 23.a4 Rc5 24.Qxf4 Rxe5 >

Rodriguez: <A move that was instantly critized at the press room, probably without reason.

[ 21...Rc2 22.Bf3 Bxf3 23.Rxf3 gxf4 24.Rg1 and White holds ]

[ 21...Bc5!? This is the move everybody was expecting. At the press conference, Kramnik said he had tried to make it work but could not see how to continue after the best defense 22.Qa4 gxf4 23.Bf3 Bxf3 24.Rxf3

A) 24...Be3 25.Nd6! (after 25.g3 Qg4 the tactics were working nicely for Black ) 25...Qxe5 ( 25...Rb8 26.Nc4 Kh8 27.Qc2÷ ) 26.Nxc8 Rxc8 with compensation but probably not more ;

B) 24...Qxe5 25.Qxf4 Qxf4 26.Rxf4 Rfd8 the ending looks very comfortable to Black ]>

NB: Some analysis of line B indicates the Black has good practical chances because of the active pawn majority on the King side, including the passed e-pawn. Variation B rates at about 0.65 in favor of Black.

Pein: <21...Bc5 22.Qd2 gxf4 23.Bf3 Be3 Looks even better.>

<Main engine preference>: <21...Bc5>: ( 0.59): <22.Qd7 Bc6 23.Qd3 gxf4 24.Bf3 Bxf3 25.Qxf3 Qxe5 26.Qg4+>

<The evaluation jump of <<0.64>> renders <21…gxf4> a <<dubious move>> as defined in the Project methodology and adds <<0.5>> to the error weighting for the game.>

GAME MOVES 22-23:

<22. Bf3> (=0.09) <22…Ba6> (=0.09)

<23. a4> (=0.00)

Pein: < 23.Qb6 Bxb5 24.Qxb5 Rc5 25.Qa4 Rxe5 26.a3 Be7 27.Qxf4=>

<Engine preference>: <23.Qb6> (=0.09): <23…Bxb5 24.Qxb5 Rc5 25.Qa4 Rxe5 26.a3 Be7 27.Qxf4 Rb5 28.Qc7>

GAME MOVE 23:

<23…Rc5> (=0.00)

Polgar: < Black chose the plan I noted above. Black has a small advantage with the Bishop pair. Black is certainly not going away without a fight. All chess fans should be excited about the fighting spirit by both players in spite of the current score.>

GAME MOVE 24:

<24. Qxf4> ( 0.33)

<Engine preference>

<24.Rfe1> (=0.00): <24…Bxb5 25.axb5 Rxb5 26.Re2 Qg5 27.g3 fxg3 28.Rg2 Qxe5 29.Rxg3+ Kh8 30.Qh4 Bd2 31.Qe7 Bb4 32.Qh4 Bd2 33.Qe7 Bb4 34.Qh4>

GAME MOVE 24:

<24…Rxe5> (=0.00)

<Engine preference>: <24…Qg5>: ( 0.33): <25.Qd4 Bxb5 26.axb5 Rxb5 27.Rd3>

GAME MOVE 25:

<25. b3> ( 0.40)

Pein: < Anand has decided to give up a pawn to head for a position with opposite coloured bishops which have a drawing tendency. 25.Be4 Bxb5 26.axb5 Rxb5 27.Rd7 Also looks well playable.> NB: As per main engine preference below, Black can play <25…f5> instead of taking the Nb5.

GAME MOVE 25:

<25…Bxb5> ( 0.40)

Polgar: < Kramnik wants to squeeze out a win with a pawn advantage in a Bishop opposite color Bishop endgame.>

GAME MOVE 26:

<26. axb5> ( 0.40) <26…Rxb5> ( 0.40)

Polgar: < This has to be the biggest advantage Kramnik has in the match so far. The real question is can he convert it? Is the advantage big enough?>

GAME MOVE 27:

<27. Be4> ( 0.43)

<Main engine preference>: (-0.40): <27.Rd7> <Rf5 28.Qc4 Qc3 29.Qh4 Qe3 30.Qc4>

GAME MOVES 27 AND 28:

<27…Bc3> ( 0.43)

<28. Bc2> ( 0.52)

Pein: <19 minutes left each. Kramnik is a clear pawn up but it's not easy.>

<Engine preference>: (-0.43): <28.Rd3> ( 0.43) <28…Be5 29.Qf2 Rb4 30.Bc6 Rfb8 31.Rf3 f5 32.Rd1 Bd4 33.Qe1 Qe5 34.Bd7 Kf7 35.Qh4 Be3 36.Qh5+ Ke7 37.Qg6 Bd4 38.Ba4 Qg7>

GAME MOVES 28-30:

<28…Be5> ( 0.52)

<29. Qf2> ( 0.52) <29…Bb8> ( 0.51)

<30. Qf3> ( 0.58)

Polgar: < Black can get the other Rook into play with 30...Rc5 then doubling up the Rooks.>

<Engine preference> <30.Qh4> ( 0.51): <30…Rb4 31.Qh3 Rc8 32.Bb1 Rf4 33.Qd3 Rh4 34.h3 Rf4 35.Rxf4 Bxf4 36.Rf1 Bg5 37.Qe4 Bf6 38.b4>

GAME MOVE 30:

<30…Rc5> ( 0.58)

Rodriguez: < Black has regrouped his pieces that are all working nicely. The big question is now how to make progress against best defense by White.>

GAME MOVE 31:

<31. Bd3> ( 0.76)

<Main engine preference>: <31.Bb1> ( 0.58) <31…Kh8 32.Rd7 Rc7 33.Rd2>

GAME MOVES 31-32:

<31…Rc3> ( 0.76)

<32.g3> ( 0.79)

<Engine preference>: <32.Rc1> ( 0.76): <32…Rxc1 33.Rxc1 Qe5 34.g3 Ba7 35.Rd1 Qe3 36.Kg2 Qxf3+ 37.Kxf3>

GAME MOVE 32:

<32…Kh8> ( 0.79)

Polgar: < The players are moving quite fast here due to time pressure. Therefore, don't expect the most precise moves. Black still has a small advantage.>

Rodriguez: <Creating the threat Rxb3>

GAME MOVE 33:

<33. Qb7> ( 0.79)

Rodriguez: < Looks active but at the same time this move creates new problems for White.>

Polgar: <One idea for Anand is to play Rf3 then doubling his Rooks on the f file. However, White is still worse here.>

GAME MOVE 33:

<33…f5> ( 0.59)

Polgar: < Trading Queen is fine for Black as he can squeeze the endgame without risks. I don't think Anand would want to trade Queen as he has to fight to earn 1/2 point. He wants to keep things in play.>

Rodriguez: < [ 33...Bc7! is much stronger since Black may be threatening 36...Rb8. Should White try to protect the b3 pawn with the natural 34.Bc4 then he runs into nasty tactics by Bxg3!!

A) 35.Rg1? Qe5! 36.hxg3 ( 36.Rxg3 Rxg3 37.hxg3 Qh5+ ) 36...Qh5+ 37.Kg2 Rc2+ ;

B) 35.hxg3? Rxg3 ;

C) 35.Qe7 Be5 36.Rg1 Bf6 37.Qc5 Bg5 ]>

<Main engine preference> <33...Bc7> ( 0.79): <34.Qb5>

GAME MOVE 34:

<34. Qb6> ( 0.70)

Rodriguez: < [ 34.Qxg7+ Kxg7 35.Bc4 Kf6 led to a better ending for Black ]>

<Engine preference>: <34. Qa6> ( 0.59): <34…Re8>

GAME MOVE 34:

<34…Qe5> (=-0.01)

Polgar: < Ouch, Kramnik blundered to allow Anand to force a draw: 35.Bxf5 exf5 36.Qxh6+ Kg8 37.Qg6+ Qg7 38.Qe6+ Kh8 39.Rxf5 Rxf5 40.Qxf5 Qb7+ 41.Rd5=>

Rodriguez: < ?! Allows White to save the game [ 34...Ba7? 35.Qxe6 Qb7+ 36.Be4! ] [ 34...Be5!? ]>

<Engine preference>: <34...Be5> ( 0.70): <35.Rfe1 Bd4 36.Qb5 Rc5 37.Qa6 e5 38.Rc1 Bc3 39.Qb6 Rcc8 40.Rf1 Bd4 41.Qa6 Rxc1 42.Rxc1>

GAME MOVE 35:

<35. Qb7> ( 1.08)

Polgar: <Anand did not see that line. Hard to calculate the whole thing through with little time.> NB: Anand took a minute or so to make this move and had 8 minutes 44 seconds left after this move.

Rodriguez: < [? In time pressure Anand misses 35.Bxf5! exf5 36.Qxh6+ ( 36.Rfe1!? Qg7 37.Re6 may even be stronger ) 36...Kg8 37.Qg5+ ( 37.Qg6+ Qg738.Qe6+ Qf7 39.Qxf7+ Rxf740.Rd8+ Rf8 41.Rxf8+ Kxf8 42.Rxf5+ ) 37...Qg7 38.Rxf5 ]>

Pein: < Anand missed 35.Bxf5! after which 35...Rxf5 36.Rxf5 exf5 37.Qxh6+ Kg8 38.Qg5+ Qg7 39.Rd8+ wins; but 35.Bxf5! exf5 36.Rfe1 Qg7 37.Re6 Bc7 38.Rxh6+ Kg8 39.Qe6+ Qf7 40.Rg6+ Kh8 41.Rh6 .>

<Engine preference>: <35.Bxf5> (=-0.01): <35… exf5 36.Qxh6+ Kg8 37.Qg6+ Kh8 38.Qh5+ Kg8 39.Qg6+ Kh8 40.Qh5+ Kg8 41.Qg6+>

GAME MOVE 35:

<35…Qc7> ( 0.37)

Polgar: < Kramnik is persistent. He wants the Queens off the board. I am not sure if he has enough to win though.>

Pein: < 35...Bc7 has a trick 36.Bc4 Rxg3! 37.hxg3 Qxg3 38.Qg2 Qh4+ 39.Kg1 Bb6+ 40.Rf2 Rg8 ; 35...Bc7 36.Qb5 may hold 36...Qxb5 37.Bxb5 Rxb3 38.Bc4 wins back a pawn because 38...Rb6 39.Rd7 Bb8 40.Re7 e5 41.Rd1 gives White too much play; But Black can refine this idea with 35...Rg8! Just ramping up the pressure and Anand would be suffering. I can't see a good move for White after this. Kramnik can calmly consolidate with Bc7 then there are threats to b3 and pressure against the kingside. 36.Qb5 Qe3 37.Qd7 Bxg3! ; 35...Rg8 36.Qf3 Rd8 .>

Simplifying too early. The queen exchange reduces Black’s options, leaving a very drawish position, the only tension being of the psychological variety emanating from the impending time control in the game and the implications for the match of a result, or lack of it, in this game.

Rodriguez: < A really poor decision since there was a lot of danger for the White king. At the press conference Kramnik said that he had not been good handling tactics in time trouble in previous games, so when he started here calculating the line 35...f4 and later discovered that it simply allowed mate in one, he decided to go for the ending. [ 35...Rg8 On his way to the press conference somebody informed Kramnik of this move that he later quoted 'as winning'. It's not easy to asses if the position is already winning or not, but certainly Black is clearly better with the queens over the board. ]

[ 35..Bc7 was also a move ]>

<35…Bc7> and <35…Rg8> were the two engine preferences. The former is more promising as despite the aggressive looking attack, the latter looks drawish, eg: <35…Rg8 36. Rf2 Bc7 (threatening <37…Rd8> or even <37…Rb8>) 37. Bf1> effectively abandoning the b-pawn>:

click for larger view

If <37…Rb8 38. Qa6 Rcxb3 39. Re2 Re3 40. Rxe3 Qxe3 41. Qc6>:

click for larger view

If now <41…Qe4+ 42. Qxe4 fxe4 43. Rc1 Bd8 44. Rc6>, Black could offer to exchange rooks with <44…Rb6>, which leads to a drawn ending with opposite colored bishops, or sac the rear e-pawn for practical chances with the rooks on the board, eg: <44…Kg7 45. Rxe6 Rb4 46. Re5 a4>. This however, looks drawish as well, notwithstanding the outside and central passed pawns.

If instead <35…Bc7> (threatening <36…Rb8>), there might follow <36. Qg2 Rxb3 37. Bc2>:

click for larger view

However, White’s best alternative after <35…Bc7> appears to be <36. Qb5> immediately putting the question to Black:

click for larger view

If <36…Qe3, then 37. Kg2 Rb8 38. Qd7 Rg8 (<38…Rd8 39. Qf7>) threatening <39…Rxg3+> and a mating attack, 39. Rf3 holds>.

Alternatively, if <36…Qxb5, then 37. Qxb5 Rxb3 38. Bd7 Rb6 39. Rc1 Bb8>:

click for larger view

Black has two extra pawns, including an outside passed pawn, with good practical chances, certainly greater than accompanied Black’s <35…Qc7>. There does not, however, seem to be a forced win in this variation either.

<The evaluation shift of <<0.71>> makes this a <<dubious move>> with a weighting of <<0.5>>.>

GAME MOVES 36-37:

<36. Qxc7> ( 0.37) <36…Bxc7> ( 0.40)

<37. Bc4> ( 0.46) <37…Re8> (=0.00)

Pein: < 37...a4 38.Rd7 axb3 39.Bxb3 Rxb3 40.Rxc7 holds comfortably for example 40...Kg8 41.Re1 Rf6 42.Re7 Rb6 43.Kg2 Kf8 44.Rh7.>

Rodriguez: < [37...a4 immediately was the only chance to keep the fight on]>

Leaches the remaining life from the game, which now flatlines to a dead draw. However, the <engine’s first preference>, <37…a4> ( 0.46), is also drawish, eg: <38. Rd7 axb3 39. Bxb3 Rxb3 40. Rxc7>:

click for larger view

The extra pawn is probably not worth much here, and this looks to be a drawn ending.

GAME MOVE 38:

<38. Rd7> (=0.00)

Polgar: <Black can try 38...a4. White can probably hold with 39.Ra1.>

Pein: <! Perhaps the ramifications of this were overlooked by Kramnik who appeared shattered after this.>

GAME MOVE 38:

<38…a4> (=0.00)

Pein: < Tricky but only good for what is most likely a draw 3 v 2 R+P endgame. 38...Bb8 39.Rfd1 a4? 40.Rd8 wins.>

Rodriguez: < now is much easier for White to make the draw>

GAME MOVE 39:

<39. Rxc7> (=0.00)

Rodriguez: < [ 39.Ra1= ]>

GAME MOVE 39:

<39…axb3> (=0.00)

GAME MOVE 40:

<40. Rf2> (=-0.30)

<<<40. Kg2>> and <,40. Rg1>> are both rated as =0.00 by the engine. However, the text flatlines at -0.30, hence the insertion of an = (in lieu of the normal ) in front of the figure in this and the next two ply. Allowing for inflation of the engine’s evaluation, moves 40 and 41 are functionally =0.00.>

Rodriguez: < Vishy had seen the line with Rd1 to search for perpetual on the seventh rank, but with only a minute left decided to play solid and then allowed sudden chances to Vladimir [ 40.Rd1 b2 41.Kg2 ]>

GAME MOE 40:

<40…Rb8> (=-0.30)

Rodriguez: < [ 40...Ra8!? was probably more dangerous ]>

GAME MOVE 41:

<41. Rb2> (=-0.30)

Polgar: <Now 41...Rc2 42. Rxb3 Rxb3 43. Bxb3 Rxc7 44. Bxe6 . 41...Rc2 42. Rxc2 bxc2 43. Bf1 Rb2 44. Kg1 e5 . However, White does have this: 42.Rxc2 bxc2 43.Bxe6 Rb1+ 44.Kg2 c1=Q 45.Rxc1 Rxc1 46.Bxf5 and White holds. Well done by Anand under severe pressure.>

GAME MOVE 41:

<41…h5> (=0.00)

Polgar: < Kramnik spent about 30 minutes for this move. He realized that 41...Rc2 would lead to a draw. This is very interesting. He is counting on the fact that White cannot make much progress.>

Pein: < 41...Rc2 42.Rxc2 bxc2 43.Bxe6! (Not 43.Bf1 Rb2! 43...Rb1+ 44.Kg2 c1Q 45.Rxc1 Rxc1 46.Bxf5 should be drawn.>

Rodriguez: < [ 41...Rb4? 42.Rc8+ Kg7 43.Bxe6 ] [ 41...e5 42.Kg2 ] [ 41...f4 42.Kg2 ]
[ 41...Rc2 unfortunately for Kramnik, this natural continuation was not enough after the forced line 42.Rxc2 bxc2 43.Bxe6 R¦b1+ 44.Kg2 c1=Q 45.Rxc1 Rxc1 46.Bxf5 leading to a theoretical ending where White puts his bishop in the long diagonal and his position is invulnerable. ]>

GAME MOVES 42-43:

<42. Kg2> (=0.00) <42…h4> (=0.00)

<43. Rc6> (=0.00)

Polgar: < Nice try by Kramnik but it looks almost certain that the game will end in a draw.>

Pein: <Threatening to unpin with Bd5, this is good enough to draw.>

GAME MOVE 43:

<43…hxg3> (=0.00)

Polgar: <White can just recapture without any danger.>

GAME MOVE 44:

<44. hxg3> (=0.00)

Polgar: <Black can play Rg8 but it will again lead to a draw. All lines seem drawish to me.>

GAME MOVES 44-45:

<44…Rg8> (=0.00)

<45. Rxe6> (=0.00)

Polgar: < Black can play 45...Rxc4 46.Rxb3 f4 47.Rh6+ Kg7 48.Rh4 =>

GAME MOVE 45:

<45…Rxc4> (=0.00)

Draw agreed. Final position:

click for larger view

<CONCLUSION>

It seems to me that Anand was forced into very difficult positions; and although he was never in an out-rightly lost position, he probably would have eventually erred his way into a losing position had Kramnik not made dubious moves.

On the other hand, Kramnik seemed to have developed a confidence crisis due to his previous lost games in the match- he distrusted his own calculations in time pressure situations, and so opted to simplify too quickly into an endgame that turned out to be clearly drawish.