Usually when I make my fifth move in a correspondence game, I am not that excited. I will have a plan where to go and a good idea of the evaluation of different lines. This game was different. When I made my fifth move all I knew was: We are about to enter a jungle.
On the one hand this is higher level chess than you will ever witness even during the Carlsen Caruana match (or in engine competitions for that matter), on the other hand the game illustrates the crisis of correspondence chess. When two guys, who know what they are doing, can rely on engine help, there will be a lot of draws.
To get decisive results, both players must be looking for a doubleedged fight (compare to our recent game in which the opponent refused a fight). Both did in this game, and even in this ideal environment they reached a standstill in which neither side could do much without having the opponent’s army spring to life suddenly.
I’ve put moves and lines below into a Lichess study for you to replay and analyze more conveniently.
Villain (2.231) – Hero (2.210)
Lechenicher SchachServer 2017, Caro-Kann
1. e4 c6
Will we see this in London? Both guys have the Caro Kann as a second tier weapon in their arsenal. Under normal circumstances they wouldn’t use it against their peers, but in case one of them needs to create winning chances while being reasonably solid with Black, we might see a Caro in one of the later games.
2. d4 d5 3. e5 Bf5 4. h4
White claims space and sets himself up to hunt down the Bf5. Along with the Short System 4.Nf3/5.Be2 this is probably the most substantial way for White to fight for an advantage in the Advance Caro Kann.
(4… h5 is the more popular and more solid move, but even here things can become messy quickly. 5. Bg5 This venomous sideline involves a nasty trap that Caro Kann aficionados should be aware of. Dozens of black players have been lured into 7…Qxb2?. 5… Qb6 6. Bd3 Bxd3 7. Qxd3 Qxb2? A poisoned pawn indeed. Taking it will be punished. (7… Qa6! as seen in Shirov,A (2709)-Anand,V (2817), Leon 2011, 0-1 (41)) 8. e6! Qxa1 9. Qb3 (diagram) with a big white advantage. If Black doesn’t play 9…Qxd4, White will go 10.Nge2 and the Qa1 is in deep trouble. If he does, b7 and the Ra8 fall, and Black is in deep trouble as well.)
(4… e6?? loses a piece after 5. g4)
Many strong players consider this to be a fun line more than anything, but in my opinion 5…Be4 is a serious weapon. However, there’s no way to be sure. 5…Be4 and especially 7…Sf6 is one of the few remaining unexplored jungles in the Caro Kann.
(5… Bd7 is the main move by far. Black intends to play a French rather than a Caro Kann, and he hopes to prove that the White kingside advance will turn out to be overextension.)
6. f3 Bh7 7. e6
Black’s latest try. Either this is good, or the whole 5…Be4 line doesn’t work.
(7… fxe6? Taking e6 is a poor choice. It softens up the kingside. There will be problems on the white squares, and the Bf8 is won’t be developed anytime soon. 8. Bd3 and White is in control.)
(7… Qd6 Directed against Bf4, eyeing towards g3. Until very recently this was considered to be an automatic choice for Black. Guys like Korchnoi and Sutovsky even awarded 7…Qd6 with a “!” in the annotated games I saw in the Megabase. However, it’s probably not that great after all. 8. exf7+ Kxf7 9. f4 Nf6 10. Nf3! Nxg4 11. h5 (diagram) A creative pawn sacrifice, invented by correspondence world champion (LSS) Joop Simmelink in 2012. White has a beautiful grip on the black position, and it’s extremely hard for Black to develop into a meaningful setup while keeping his king safe. White has more than sufficient compensation for the pawn, still, it took some time until Simmelink’s move got traction. It was first repeated in 2013 by an unknown 2.300 player from India. After that a few grandmasters have played it successfully. In the world of correspondence chess since 2012 there was only one guy stupid enough to run into 10.Nf3/11.h5, and that was me. I managed to save that game, but had to fight hard for >60 moves.)
There had been some trashtalk going on with the moves sent so far. I had told my opponent to better follow MVL instead of being on his own against me, but he said he thinks that 8.Nc3 is better than 8.Bf4. I assume he wanted to lure me into a predecessor game that turned out well for White.
(8. Bf4 Prevents Black from going …Qd6, but opens himself up for …Qb6. In my opinion the logical looking 8.Bf4 must be the critical test for 7…Nf6. 8… Qb6 9. Nc3 Qxb2 10. Kd2 and a huge mess evolved in Vachier Lagrave,M (2722) – Ding,Liren (2707), Paris/St Petersburg 2013, (1-0,39). If you ever feel like replaying achighest level fun game, I suggest this one.)
9. exf7+ Kxf7 10. Be3
Novelty, not a bad one probably. Instead of striking in the centre Black should aim for the queenside first of all, that is where the white king will try to find safety and where Black will have to organize his counterplay. The Bh7 needs a job, and since now …Nb4 is in the air it has found one. With …Re8 immediately possible …e5 may be an idea as well.
(10… Nbd7 11. Qd2 Bg6 I don’t understand this move, it looks like a waste of time, nothing else. 12. O-O-O e5 Pommrich,R (2197)-Morelli,A (2263), ICCF email 2013 (1-0, 65), is what White would have liked to repeat. But this game shows that Black didn’t understand what the position requires first of all. The centre strike …e5 looks tempting, but opening things up in front of the king after donating a tempo for nothing can’t be great. White is already seriously better. The game also shows that most ICCF players below 2.400 basically suck at chess.)
(11. Qd2!? Nb4 12. O-O-O Bxc2 13. Re1 Looks weird, probably isn’t top notch, but 11.Qd2 is not an absurd sacrifice. Black’s queenside pieces are loose, it will take some time to solidify things, and White can use that time to organize play against the black king.)
With his structure ruined it is now clear that Black has to look out for dynamics and piece play, he won’t win this one in positional style. White on the other hand has paid quite a price for foisting an isolated double pawn upon Black. He has given up the bishop’s pair, opened the b file, and the beautiful Bh7 doesn’t have a counterpart anymore. The engines are quite happy for White, though.
(12. Qd2 Qb4 13. O-O-O Rb8 A line to illustrate what Black is looking for: White is forced to weaken himself with 14.b3, and Black is happy.)
Feels a bit obvious and not too sophisticated, but it is an important move that forces White to make a structural concession himself, one that Black can build upon. With b2 and c2 under fire, White can’t just go Qd2 (like he would if the Bh7 wasn’t there). He either has to move his king or loosen up his queenside.
At first sight this gives Black a mark to attack and dissolve his poor a6/a7 pawn combo. But most of all it is now becoming apparent that the Kf7 isn’t the only king on the board that doesn’t feel comfortable. In fact this is the broad theme of the game: Both sides look for a way to penetrate the opponent’s defenses in order to get to his king while trying to keep their own lines intact.
(13. Kd2!? like in MVL-Ding, see above)
White must not play g5 with tempo, and the knight is needed on the queenside, either to support …c5 or to go to b6 and support …a6-a5-a4. Note that …e5 will rarely be a part of Black’s plans.
(14… e5?! Again, this isn’t a good idea. 15. f4! Now either the black king gets in danger if Black opens things up with …exd4 or …exf4, or he has to close in his h7 bishop with …e4. None of these options looks tempting.)
White starts rolling up the kingside, and Black must play precisely to not get rolled over. Thanks to 13…Nd7 Black has time to decide how to set himself up now. A tough decision, plenty of options, all of them playable. I had expected 15.g5, but hadn’t made up my mind how to counter it, and when the moved arrived it still took me more than a week to decide.
The sharpest reply. Black ignores the White advance towards his king and prepares countermeasures on the c file immediately. Some sample lines to show how Black could have set himself up instead:
(15… hxg5!? 16. hxg5 Kg8 17. O-O-O Bf5 18. Rxh8+ Kxh8 19. Rh1+ Kg8)
(15…Bf5 16. gxh6 gxh6 17. O-O-O)
(15… h5 16. Nf4 Bf5 17. Nd3 Qa5 18. O-O!)
16. gxh6 gxh6
(17. Bxh6 Bxh6 18. Qxh6 Bxc2 19. Qf4+ Nf6 20. Kf2 c5 Things are sharp and doubleedged, but objectively Black should be fine. His activity compensates for White having Harry the h passer.)
Stabilizes, improves a piece and provokes the following sequence. It’s hard to come up with a plan for White now. Everything is improved and positioned nicely, only 0-0-0 runs into …c5. But what else?
(17… c5?! Not yet. 18. a3 is good for White. Only after White has gone 0-0-0, …c5 will be the immediate reply.)
18. Ng3 Bg6
Marks h6 as a permanent target, but also blocks the h5 square for the Ng3.
(19. f4 The idea f3-f4-f5 would be a consistent attempt to crack the black setup. Only it doesn’t lead to anything as long as Black defends precisely. 19… Qa3! Looks slow and remote, but this is the only move that keeps the balance. Preventing White from castling combined with the options …Bb4 and …Qb2 ensures Black enough counterplay. 20. f5 exf5 21. Nxf5 Bxf5 22. Qf2 Ke6 Bold! Instead of hiding the king marches forward and works as a defender. 23. Rf1 Bb4 24. Qxf5+ Ke7 (diagram) with an unclear position. Note that after 25. Qf7+? Kd8 26. Bd2 Qb2 Black is almost winning. White doesn’t have any threats against the black king, and it turns out to be crucial that not only c3 is hanging, but without the queen on f5 c2 isn’t covered anymore.)
Now the Ng3 looks lost while its colleague on c3 lacks protection.
20. Nge2 Bf5
Back to square one (compare to the notes after 17…Bf5).
(21. Nf4 c5 22. Nd3 Qa5 Would be a way to close the h7-b1 diagonal and gain some central control. But now the weakness of the Nc3 forces White to dissolve then tension, and whatever he does, Black is ok. 23. dxc5 (23. Ne2 Qxd2+ 24. Kxd2) 23… Bxd3 24. cxd3 Bxc5 25. Bxc5 Nxc5)
Now is the time!
22. dxc5 Nxc5
The white king needs the d1 square. Black was threatening stuff like …Qa3+ Kb1 …Nxb3+
axb3 …Qxb3+ and so on. For the moment White has all black ideas under control and all his guys in play. It is now on Black to find a meaningful way to continue.
(23. Nd4? Nxb3+)
Prevents Nd4, keeps up the pressure on the c file. Now, finally, Black is aiming for …e5 (and …d4), and White needs to prevent it. From here on White has several occasions to grab the a7 pawn, but that’s something he can’t do without Black taking over immediately. It’s crucial for White to keep the d4/e5 combo under control.
(23… Bd6 24. Bd4 e5 25. Bxc5 Rxc5 26. Ne4! Qxd2+ 27. Kxd2 Doesn’t look like it at first glance, but this is a pleasant endgame for White at least.)
(23… a5?! …a6-a5-a4 looks natural, but it’s way too slow. After 24. Nd4 White is clearly better.)
(24. Bxa7 Bd6 25. Bd4 e5 and Black takes over.)
24… Rg8 25. Rxg8 Kxg8 26. Qe3 Kf7
Both players have moved their guys into an ideal configuration, and now both can’t make progress without having the opposite army spring to life. Over the board this could go either way, but this is correspondence chess, and so the game has come to a weird standstill. The position is equal with both forces neutralizing each other. I’ll leave the following maneuvring without comments. Both sides tried for a bit to get something going, but soon both run out of ideas and shake hands virtually.
27. Re1 Qd6 28. f4 a5 29. Ng3 Bh7 30. Re2 a6 31. Rg2 Qc6 32. a4 Ba3+ 33. Kb1 Rb8 34. Ka1 Rg8 35. Re2