Last night, mrkajz44 and I finished a long and draining game, with some huge swings in the advantage over 50 moves. A real grind:
https://www.chess.com/analysis/game/...169?tab=review
I had the white pieces. The game opened with a Caro-Kann from black, and we played the advance variation with 1. e4, c6, 2. d4, d5, 3. e5, Bf5, where black develops the light square bishop to sit outside of what usually becomes a semi-Slav pawn structure with a pawn on e6.
Now I have a weapon against Bf5 that I’m sort of proud of. It’s to play 4. h4, which has the sneaky idea of trapping the bishop if black plays a number of very natural looking moves. So in this game we had 4.h4, e6, 5. g4, and then the bishop is on the run: 5. … Be4, 6. f3, and the bishop is trapped.
So I’ll admit, at that point I thought I was looking at an easy win. But… BOY was I wrong.
The problem with this position for white is although you win the bishop for a pawn, your pawns on the king side are very committed (you basically cannot castle king side anymore), and you’ve spent a bunch of time pushing pawns in the opening, neglecting development on the king side.
So my opponent let me capture the bishop on e4, but found the sneaky resource 6.Be7, which is a move I underestimated. After 7. fxe4, Black played 7. … Bxh4+, and suddenly it dawned on me that I might have a problem. 8. Rxh4 loses the exchange to Qxh4+, and the king still has to move somewhere, and now I’ve lost castling rights and my king is ridiculously exposed due to the overextended pawns on the king side. So I was forced to play Ke2, (basically a delayed bongcloud) which is a very ugly square for the king. At that point I was convinced I had blundered somehow and that I was probably losing. We ended up with a position where Black had a very annoying bishop on h4, a passed protected pawn on e4, and my king is preventing my queen from developing.
Now, it turns out stockfish isn’t scared by that at all. It assesses the position at +4 for white, I guess on the basis that white is up a piece for two pawns and can somehow move the king out of danger with Kd2, Kc1 or something like that, I guess.
Suffice it to say I didn’t play the position as accurately as stockfish might have liked, but nevertheless ended up with a position on move 24 in which stockfish says white is better despite the fact that the king is on e3 and way out in the open. I tried to mount an attack on the king side with some queen and rook batteries, and I played 24. Nf4, with what I thought was a very clever idea: Nxg6, which wins a pawn because the pawn on h7 is tactically pinned (black cannot take back because 24. … hxg6 loses a rook after Rxh8, Rxh8, Qxh8). It was all looking good,
as long as my opponent didn’t see that tactic.
Reader, he saw it. Black calmly played Rg8, adding a defender to g6, and the plan no longer works. I persisted with this idea adding an attacker with 25. Rh6, and what followed was some weird maneuvering where both sides tried to point absolutely all of their pieces at black’s pawn on g6.
On move 28 I had another clever idea. Except it turns out I should stop having those, probably—this one got me into a lot of trouble.
My idea was to push my e pawn with 28. e6, with the sneaky idea of opening the h2-b8 diagonal, and then try to sneak my queen into the other side of the board while all of black’s major pieces are stuck defending g6.
Stockfish immediately sees the problem. This move really weakens the pawn chain in the Center, and it doesn’t only open a path for white’s queen—it opens a path for black’s queen too, which is now pointed straight at the white pawn on d4. Now, that pawn is defended, so it’s all good right? Not…. really. Its only defender is the king, and any number of pawn pushes will chase the king away.
I then played b3, trying to create some pawn advances on the queen side. That’s apparently a blunder, because stockfish says it loses the queen, after a difficult-to-find sequence that neither player saw.
But then I blundered again: I played 30. Kd2 to try to get my king to a safer position and away from annoying pawn checks, forgetting that the king was the only defender of the d4 pawn, which in turn defended the knight on c5. So that move doesn’t just lose a pawn: it loses a pawn AND a knight, after 30. … Qxd4+, 31. Kc1 (Nd3 was better, apparently), and Qxc5.
It dawned on me then that I was probably losing. And now stockfish agrees: the position is almost -4 for black.
I spent the rest of the game trying to find some counterplay, but knowing I would be lucky to find a draw. It seemed like I had found that on move 39, winning the a pawn and then trading queens down to what I hoped was a decent endgame. Trading queens was another blunder: stockfish says white is better if it avoids the queen trade, and chases the queen away with Re2 instead.
Both players had some minor inaccuracies after that but found the right moves to secure a drawn endgame, where both sides had a rook and 3 pawns. Mrkajz44 offered a draw on move 50, and I accepted—somewhat relieved not to have lost if I’m being honest.
Crazy game.
