Reports | August 27, 2013 20:50

World Cup R6: two more draws, Kramnik misses a win

World Cup R6: two more draws, Kramnik misses a win

Both games on the second day of the World Cup's semi-final ended in draws, and so the playing hall in Tromsø will see two tiebreak matches on Wednesday that will determine the finalists. Kramnik will not be happy as he missed a clear win; Vachier-Lagrave was putting up excellent resistance in a worse ending, but after close to five hours of play he eventually made a mistake that could have meant his elimination: 62.Nd7 Rf5 63.Rf8+ Kg6 64.Rg8+ Kf7 65.Ke4 +-. Evgeny Tomashevsky and Dmitry Andreikin drew their game after 29 moves.

Photos by Paul Truong

After the somewhat disappointing first day of the semis, the chess fans were well treated on Tuesday with a long and at some point dramatic game between Vladimir Kramnik and Maxime Vachier-Lagrave. Before we get to that, let's first look at the Tomashevsky-Andreikin game, a Queen's Gambit Declined, Exchange Variation. Andreikin's 8...h6 and 9...Nh5 is known as one of the most solid lines of dealing with this old system, and together with castling queenside, Black's position does look very healthy. White's minor success of creating an IQP in Black's camp wasn't worth much, and after 29 moves Tomashevsky stopped trying.

PGN string

Kramnik-Andreikin was a Slav where the Frenchman went for a slightly worse, but quite solid ending. At some point he sacrificed a pawn, assuming that he would be able to trade all pawns on the queenside. This is what happened and it was clear that Vachier-Lagrave would have to suffer for long, but at the same time the ending should be holdable.

I gave to admit it was quite close to a draw all the time,

said Kramnik, who got very far nonetheless and was even winning at one point. Vachier-Lagrave gave a wrong check on move 58 and then had to give his knight. For White, however, it was suddenly very difficult to keep his last pawn. Both players thought that the ending had to be winning, but neither of them saw how exactly.

PGN string

During the commentary, Vachier-Lagrave said:

I don't know but I was panicking for no reason all the time...

Kramnik started the tiebreak early by interrupting him:

There is a big reason; you're playing me!

Tomorrow at 15:00 CET / 09:00 EDT the tiebreaks of the semi-finals start in Tromsø. Andreikin begins as Black in the Andreikin - Tomashevsky match while Kramnik starts as Black in the Kramnik - Vachier-Lagrave match.

Results round 6



Peter Doggers's picture
Author: Peter Doggers

Founder and editor-in-chief of, Peter is responsible for most of the chess news and tournament reports. Often visiting top events, he also provides photos and videos for the site. He's a 1.e4 player himself, likes Thai food and the Stones.


hansie's picture

So the names of the 4th & 5th Candidates would be known only Tomorrow ...

Montcalm's picture

I'm positive that at the end MVL was hurt by Kramnik's attitude, if it was an equal position where you can create some chances "à la Carlsen", no! it was like a 150% drawn position, even my dog can draw this against Kramnik!
MVL looks up to Kramnik and they are sparring partners living in Paris, and playing on like that is what you'd do against a beginner who doesnt know it's a draw, it's like saying "i'm so better than you than i'll beat you in a theoritical drawn endgame, prove me you can draw this!" MVL looked at Kramnik many times like "hey that's the high esteem you have of me?" .. "thank you dad.. :(
Now MVL can tell himself Kramnik was upset like a 5 yo cause he didn't win a won endgame and needed some moves to digest it. But still there is a personnal element to it. I recall Kramnik saying the precedent round he feared "Caruana is like serious opposition" implying MVL is not?
Then after throwing the tantrum he goes into a smiley post mortem that MVL clearly was not so enjoying. Weird stuff.

Ding's picture

Your post clearly suggests that you would have no chance drawing this against Kramnik. If you seriously think that any patzer can draw this endgame, have a look at this game:

Montcalm 's picture

I'm 600 points below Kramnik but I take up the challenge. You can even do it at home against Houdini...
Taking the one in a million exception is irrelevant.
Also, yesterday was too players playing 2800+ perf in the cup, rewatch the vid you'll see MVL face in the last moves...

Montcalm's picture

Even though i'm 600 points below Kramnik, i take up the challenge! Actually last night, played this position against Houdini (Elo ~3300) on my computer and i easily reached the 50 moves and drew... I invite people to try as well.
Finding the one in a million exception is irrelevant.

I'm sorry you didn't understand the point of my post which was the psychological relation between the these two players...

Thomas Oliver's picture

Engines are relatively weak or clueless in (such) endgames. They don't think in terms of practical chances - always expecting the best move from the opponent, they will not pose little traps or opportunities to go wrong. How long did it take you to reach 50 moves and draw? In other words: did Houdini even try to make progress, which might mean exchanging pieces or moving pawns?

Carlsen probably would have played on about any time against anyone. Anyone would continue against anyone in the given tournament situation. Kramnik turned out to be right: Vachier-Lagrave did make a mistake and could have lost.
So even if there is a psychological element between the two players, Kramnik did nothing wrong, unprofessional or unethical.

Soviet School's picture

It turns out kramnik was correct as it may have tirred the Frenchman out hence his disaster in the first playoff game

Benbow's picture

Kramnik was clearly upset because he knew he was drawing a won position. Who wouldn't? So he played on more to calm his nerves than hoping to win. You're friends until you're at the board, but everyone's human. There's no shame trying to win until the end a drawn position, especially after 5 hours where MVL was certainly tired to defend and could have blundered (1% chance but why not take it).

Niima's picture

Indeed. And many such "drawn" endgames have been lost even by top GM's. Another example is Grischuk's last-minute loss in a theoretically drawn endgame at the last olympiad.

Anonymous's picture

I don't recall one single instance where a 2790 player lost this endgame though, apart from the one time Polgar did it but then the endgame started in the worst possible position to defend and she was lower on time. MVL always had more than ten minutes plus increments and it was quite optimistic of Kramnik to keep hoping for the blunder of the century for 50 moves.

Anonymous's picture

2700 player (or even 2600 player)

Thomas Oliver's picture

2600 players lost this endgame twice:
Carlsen - l'Ami 1-0, Wijk aan Zee 2011 [this had also been rook, knight and pawn against rook. Carlsen lost his pawn on move 74, l'Ami blundered on move 93]
Onischuk - Dominguez 0-1, Biel 2008 [mentioned in the comments at The last pawn disappeared on move 53, losing mistake on move 88, resignation after move 91.].

Maybe Kramnik even remembered the first game, thinking "if Carlsen can be lucky, why not me?".

euro-sceptic's picture

Apart from the comical 'I don't remember one single time ..except this one time' comment above, I think your argument is fundamentally flawed by the fact that... Karmnik acheived a winning position!
I don't understand why so many people complain about boring Super-GM draws in chess and then complain when they actually try to play in an objectively drawn position - if you follow your argument to it's logicval extreme the players would show the 'due respect' to one-another and agree a draw before the game starts.
Pipe down!

Anonymous's picture

"Apart from the comical 'I don't remember one single time ..except this one time' comment above, I think your argument is fundamentally flawed by the fact that... Karmnik acheived a winning position!"

What winning position? The one before Nd7 was obviously not in the rook + knight vs rook endgame that was discussed.

Anonymous's picture

Polgar wasn't 2700 when she lost the endgame, by the way, but she was a future 2700 player. No player rated 2700 or higher at the time of the game has ever lost the endgame in question in classical chess, so had MVL blundered it would have been unprecedented.

Anonymous's picture

Aronian's name is known already.

Thomas Oliver's picture

So are two more names, Kramnik and Topalov. Anand and Carlsen haven't qualified yet, one of them still needs to lose their WCh match.

hank's picture

Maxime should have a good chance in the rapids. Then in the Candidates on the expense of Karjakin.

RG13's picture

Kramnik shouldn't feel to bad about missing that win after close to five hours of play. The sequence beginning with 62.Nd7 does not seem to be painfully obvious, but then I'm not Vladimir Kramnik.

Caissa's picture

I agree. He was running short of time. With Rybka and a Quadcore at home it is really easy, but we must congratulate Kramnik for trying till the end. He is a superb player, but sometimes he's been overconservative. That's what upset fans. MVL is playing a great tournament, grinding Gelfand and Caruana down, but Kramnik is the only one outplaying his opponents on the regular games. Go Vlad!!

Thomas Oliver's picture

I don't know how much time Kramnik had at move 62, but apparently he spent a lot searching for a win - which existed, but either you see it or you don't see it. Vachier-Lagrave realized that he must be lost but didn't see how (not his problem ...), and it seems that both players "almost" found it in their postmortem but didn't spot 65.Ke4 at the end of the variation.

Kudos to both for joining the live commentary after their game. As far as Short is concerned: Maybe it was OK to show the win (his job!?), but it was quite unnecessary to start a lengthy analysis on whether something else than Kramnik's 40.h5 might have offered better winning chances - the verdict was the same as for the game continuation: better for white, practical chances, objectively drawn. No need to insist that Short was possibly smarter than Kramnik ... .

It might be stretchy, but to me the game is also a bit reminiscent of Eljanov-Jakovenko in round 2.2 of this World Cup. White had queen, knight and three pawns vs. queen and four pawns, but the knight was stuck on a6 and black threatened perpetual check. It took Eljanov 31 moves (Jakovenko may have started hoping for the 50-move rule), and it might have been objectively drawn with best play from black.

Anonymous's picture

Most people kept saying that Nd7 was difficult to find with only two minutes on the clock, but two minutes was what he had left after spending 15 minutes trying to find it.

Anonymous's picture

Still, saying that Kramnik's move was a huge blunder, and giving it double question marks at Chessbase (GM Ramirez) looks harsh.

Anonymous's picture

Short is being smarter :)

Vde's picture

62.Nd7 was not obvious at all, it was a study-like win. But still some people probably rated below 1600 called Kramnik's move a blunder.

o anão gigante's picture

Fully correct: it is far from obvious. And Kramnik did not see Ke4 until Short mentioned it. There is a long way to go after the best defence: Rf5. In spite of what is being said in here, VK never saw a wining line.

choufleur's picture

The Kramnik Vachier Lagrave ending is incredibly reminiscent of the Kramnik Kasparov fourth game of the 2000 wc match. Same material R+N+P vs R, same incredible result, draw ! It is very likely that Kramnik's play was affected by this precedent failure. The difference is that today the pawn was a g-pawn, whereas it was a a-pawn versus Kasparov.

Anonymous's picture

Many +1 pawn endgames are much less "won" than most people (and computers) might think.
A lot of endgames with an extra pawn are, with best play, actually drawn!

Kronsteen's picture

This was a +1 pawn +1 PIECE endgame.

Anonymous's picture

Not really. Because a knight+rook against rook is drawn also.

Frits Fritschy's picture

Following this logic, a draw with an advantage of two knights and a pawn isn't anything special either, as just two knights can't win against a lone king.

Remco G's picture

K+2N v K is drawn too, so two knights don't matter either. But K + 2N v K + Q is drawn too, so queens don't matter either.

Soviet School's picture

Very good Remco, and Snooker (or pool) is chess with balls so snooker without balls is chess.

Vde's picture

Hope that Kramnik wins the tournament because he is clearly the best in classical chess. In rapids a lot of luck is involved.

-:)'s picture

Fair play VPL on squeaky bum defence but K will be fuming.

europatzer's picture

Kramnik tried hard but the win was obscure
MLv played in a very creative style QB8 then Putting the Q on E8 and e6 very original

maybe his time has come !!??

Alexander's picture

Kramnik has played well in rapids as well. He didn't go to blitz so far but it can happen tomorrow.

alfa's picture

Nice game and a thrilling final part.I think that Kramnik had the same endgame against Piket(rook-knight+3 pawns vs rook-knight+2 pawns) long time ago in Monaco,though I am not sure if my memory is still what it used to be.He won that game with similar mating ideas,I think.

Frits Fritschy's picture

Can't find it, maybe you meant this? (which is a bit different)

Anonymous's picture

Did you see on video how Kramnik was distracting - maybe unconciusly- Vachier in the first time control before the last move 40? These things could be forbidden.

Sarunas's picture

If Ramirez puts a double question mark for 62.Ke4 in Chessbase, then I'm glad he got eliminated in Round1st at Tromso. When the young lad will learn that engine backed assesments may hurt our best artists' dignity, he will better understand the nature of chess and even may hope for progress. In fact, nobody spotted 62.Nd7 idea with relevant ramifications -nor Kramnik nor MVL neither Short. As Nigel put it in afterwards, it got tossed by most powerful of all Dirk Geuzendam. In fact Ke4 is an excellent move, but only at step 65. CB should reckon that anf reframe criticism of Ke4 accordingly.

UltimatePatzer's picture

Of course Ramirez put a ?? ("Returning the favor!") and the article calls the move "a huge blunder".

Funny, couple of years ago (without engines and tablebases) commentators would not dare to evaluate such ending, yet to make a resolute judgment. Nowadays every scribbler with an engine calls a missed study-like win a huge blunder.

Frits Fritschy's picture

Even when you try 65 Ke4 in your mind, you still have to get to the idea of sacrificing your pawn with 67 Ne5+. Not easy at all, even for a super GM.

Soviet School's picture

Amazing coincidence that kramnik drew the same material configuration with Kasparov in 2000 after an amazing game. I happened to follow both these games live over internet. It was interesting that Short did not remember this earlier Kasparov game till prompted by a tweet, it shows that memory of even World Champion challengers is not as good as we may think.

He also fell for a miracle draw with Carlsen a piece up in London classic a couple of years ago.

I thought theNd7 winning possibility was easier to see then many of his earlier moves but there you go,

choufleur's picture

see my above comment

Anonymous's picture

Amazing blunder by Maxime QxL on c2. It's amazing....even I don't blunder like that!!

Anonymous's picture

Hahaha what a patzer move, turns out Big Vlad did the right thing in forcing MVL to play out that endgame.

Anonymous's picture

AMAZING game by Vladdy, crushing win with black in 22 moves, compare to Carlsen being crushed by MVL in Biel!

Soviet School's picture

Did he play qc2 as a sort of resignation since he was losing exchange ny way? Otherwise it is bizarre

Thomas Oliver's picture

Either that, or - seems more likely to me: he hallucinated that black has to play (immediately) 22.-Qxc2 and that 23.Be4 regains the queen with equal material or leads to perpetual check. Never mind that this variation also has a hole: 23.-Qxd2 24.Bxh7+ Kh8 25.Bf5+ Qh6

Latest articles