Study link:
lichess.org/study/YaE3rCEKHello lichess community!
I lost an instructive game which I feel many can benefit from. I have provided my annotations for the game (without any reference to the engine so you can clearly see my thought process during the game).
It would be wonderful if you could share feedback on where either side could improve, or on anything else you feel like mentioning! The study is called Tricky Knights of Berlin (a ChessNetwork reference :D). I hope you enjoy the game!
The game is really interesting. We see so many grandmaster draws in the Berlin defense that sometimes we forget it's a legitimate attempt at winning for either side.
I became interested in the Berlin defense in 1999 when I started to play 1...e5 as a reply to 1.e4. I was playing 3...Nf6 against anything, so it made sense to play it against the Spanish as well. I couldn't figure out a single theme and I gave up, opting for 3...f5 instead. Then came the Kasparov-Kramnik match, and all the work I had done previously made sense with Kramnik answering 9.Nc3 with 9...Bd7 (instead of 9...Be6 that was the usual move with 9...Ke8 at that time).
I mention this because I think that 9...Be7 in the game is already a mistake. Not in an "objective" sense (it has been played at the top level by Ponomariov and others), but because it makes Black's game more difficult without an extensive preparation. After 9...Be7, White will easily trade the dark square bishops, depriving Black of the bishop pair (not very useful at this stage but best kept as a reserve). That's why h6 is played earlier than Be7 in most Berlin endgames (also, it guards the f7 pawn against Ng5).
Actually 9...Bd7 solves most of Black's problems as he can play h6, b6, Kc8 and take the initiative on the flanks (a5-a4, h5-h4) ; the key game here is Shirov-Almasi, Rubinstein Memorial 2000. The only way for White to fight for an advantage is to play b3/Bb2, whereby the e5-e6 push becomes a threat (Shirov-Krasenkow, Rubinstein Memorial 2000). Kramnik shows how to neutralize this with Ne7-g6 (and then h5) in the first game of his match against Kasparov (hence Black's dark square bishop must stay on f8 for quite some time). If White gains a tempo by omitting h3 (a useful move that is frequently played, but is not urgently obligatory), then Ne7-g6 comes a tempo too late and Kramnik demonstrates "plan B" in the third game of his match against Kasparov : c5, Bc6, Ne7 and Rg8! to protect the g7-pawn before developing the dark square bishop, making the e5-e6 push harmless.
I'm not sure that this is clear without a board and the full text of the quoted games (I can add this as a separate chapter of your study if you like), but my point is that you're making your life difficult with 9...Be7 instead of 9...Bd7.
Analyzing your game with the help of Komodo doesn't reveal anything surprising. The trend is favorable to White at the end of the opening, even if there is nothing critical in the short run. Black is taking a big risk with 15...c5 since this move equalizes except if White finds the pseuso-sacrifice 16.Ndb5! axb5 17.Nxb5 where Black must give the piece back in order to preserve his rook : 17...Nd5 18.Rxd5 Ke7! 19.Nxc7 Ra7 20.Nb5 Rxa2 21.Rb1 and Black is only a pawn down.
It's risky to play 15...c5, but in all cases Black was worse after Bd7 or any alternative. I think that it is a reasonable attempt at breaking the trend which was in White's favor.
Then the game proceeds logically, but Black miscalculates the "knight tour" as White's passed pawns are too strong. Instead of 24...Nxc2??, it was much better to play 24...Ke7! 25.exf7 Kxf7 26.Kxe1 Ke6 and e5 falls. Black has a rook and a pawn against two knights, with no weakness or obvious outpost which would favor the knights. This endgame is a clear draw according to Komodo, and I see no reason to disagree (i.e. no direction in which to take the analysis to discover a hidden resource).
As I understand from your notes, the risk involved in 15...c5 was not conscious. In that case, it was a "blissful unawareness" :) . Unfortunately, the next oversight (Nxc2??) was not so harmless. Your opponent has played a very impressive game, squeezing as much as White can in the Berlin endgame (just missing the tactical solution 16.Ndb5!, which only looks easy once you have considered it at all).
Life is difficult for black in the Berlin, as Kramnik himself said.
9...Be7 makes life more difficult. 9...Bd7 or 9...Ke8 have a better reputation. 10...b6 seems illogical: the point of 9...Be7 would be 10...Nh4.
Very useful insights
@A-Cielbleu, I think I did approach the game a bit too casually with my piece and pawn placement (namely Be7 and c5). 24...Ke7 I should have taken time to consider, because in the game I just saw an incorrect long-winded variation that I thought was winning and stuck to it, which is another blind spot from my side. I am able to follow your comments thanks to blindfold practice, so you don’t need to go the extra mile unless you wanted to :)
@tpr thanks for your insights as well. Yes I agree the piece placement was incongruent to the plans I executed partly due to not taking into considerations my opponents moves.
Hey! happy to provide some input as well, I was honestly a little surprised to see Be7 because it is not a main move at all, I have been watching Shirov's Chessbase DVD on beating the Berlin defence. In it, he definitely expresses his fears over the Bd7 setup, with a different idea from Kramnik, with first starting off with 9...h6 10. h3 Bd7 11. b3 c5!? with a pretty clever idea, next he is going to play Nd4, unless white plays Rfd1, when he would rather, after say 11...Kc8 Bb2, put the a rook on d1, so that the f rook can support the pawns on the kingside, as in the game alekseev Eljanov 2009
http://www.chessgames.com/perl/chessgame?gid=1554882. play then continues 11...c5!? 12. Rfd1 and the plan for black to follow is seen in Shirov-Naiditsch 2011, which I can add to the study if you like.
Yeah that’d be great, thanks for your comments as well. It’s looks like you’ve already hosted a study, but I’ll add you anyway if you’re still interested :)
of course I am interested! cheers :)
I made a study about the 9...Bd7 variation :
The 9...Be7 is apparently a Eljanov specialty, but it requires more preparation compared with the 9...Bd7 (mostly worked out by Kramnik in 2000 and well explored after that, with satisfactory results for Black).
Ty, it was definitely worth reading!
There is no question that in the starting position of the Berlin endgame white's best move if it were legal would be rook a1 jumps over knight and bishop to d1+. White can't do that because of the rules of the game. This problem means that white must choose between playing Rd1+ right away with his rook from f1, meaning that his rook on a1 has a harder time getting into the game, or not playing Rd1+ at all.
And this is important, because black has two plans in the Berlin. One is putting the king on e8, the other putting the king on c8 and possibly b7. If white plays 9. Rd1+, he does force Ke8 (9...Bd7?? 10. g4 and 11. Ne5 1-0), so every sort of Berlin player must at least know a little about the Ke8 setups, but with the rooks on d1 and a1 that is a worse version for white.
Returning to the game at hand, I don't agree with A-CielBleu's argument that 9...Be7 is particularly inaccurate. Anything played frequently by Ponomariov is not terrible. In fact, it's very likely to transpose into much more played lines. After 10. Nc3 Ke8 we have a main line where black is very healthy.
Another area in which I disagree with A-CielBleu is his belief that h6 must be played when the king is on e8. When the king is on e8 the pawn often goes to h5. This is because white's king on e8 makes the idea of Ng5 (which is more dangerous than Bg5- black isn't that afraid of the dark bishops being exchanged if it loses white a lot of tempi and his light bishop survives) not a problem, but the idea of h3-g4 remains very important. As computer analysis shows, in the game, after 11...Bxg5 and 12...Ke7 black fully equalizes despite no longer having the bishop pair. Black's bishop pair isn't the only thing keeping him alive.
In your analysis you said that instead of b6 the move 10...Nh4 was better. I don't agree. I think 10...Nh4 is playable, and makes some sense because the knight on f5 is a problem piece for black. But 10...Ke8 transposes into the most important line in arguably the whole opening. If you're trying to equalize you look there. If I were facing 10...Nh4 as white I would take it to gain a tempo, play Be3 and Rad1+ followed by the usual advance of white's kingside pawns. Playable, but a sideline.
Bg5-xe7 is not some big problem, but 11...h6 is. Getting an understanding for why 11...h6 is so much worse than 11...Bxg5 should be one of your main goals in learning from this game. A lot of the answer simply has to do with tempi and development. White achieves Rad1+ and vastly better development with initiative. And after 11...Bxg5 black can get the knight to h4 (if white really loses track of what he's doing possibly even d4) and answer the idea of g4-f4 with Ng6, put the bishop on a6 or b7, quickly put a rook on the d-file, etc.