Black to move
My moves so far have been fine, although lots of deep computer analysis and dozens of games played against Stockfish on my iPad have convinced me that it would have been more accurate to place my rook on the first rank before moving my king towards the queenside.
33...Rd1
This move does not yet squander the advantage, but would have been an appropriate moment for the correct sacrifice.
34.Rf4 Rd3+ 35.Rf3
Black to move
This move does not yet squander the advantage, but would have been an appropriate moment for the correct sacrifice.
34.Rf4 Rd3+ 35.Rf3
Black to move
35...Rdxd4??
This sacrifice, which I was proud of during the game, was a blunder that should have led to a draw.
36.Bxd4 Rxd4 37.Rb3
This sacrifice, which I was proud of during the game, was a blunder that should have led to a draw.
36.Bxd4 Rxd4 37.Rb3
Black to move
37...Rg4+ 38.Kh3??
This unfortunate error restored my advantage. Had White's king stayed near my central pawns, soon to advance, they would have been stopped.
39...Re4?
39...d4 retains the advantage
39.Rxb5 Rxe5 40.Rb8+ Ke7 41.a4 d4 42.Rb3 Rd5
Placing my rook behind a passed pawn.
Placing my rook behind a passed pawn.
White to move
43.Rd3 e5 44.Kg3 e4 45.Rd1 Ke6
White to move
The final error.
46.b4 Ke5 47.Kf2 e3+ 48.Kf3=
46...e3-+ 47.Ke4 e2 48.Re1 d3 49.Kf3 Re5 White resigned 0-1
Computer Training
After running quick analysis using the website's analysis feature, I realized my play was not as brilliant as it had felt while playing. But, it was not clear to me what I should have played to convert the clear advantage Stockfish found in the position.Running the engine and playing its favored moves seemed to involve a great deal of shuffling and often unforced weakening moves by White.
I was intrigued.
The key idea that I missed was a pawn sacrifice to open the a-file with the idea of getting both Black rooks on the first rank. Black then aims to force the exchange of rooks.
Having learning the correct idea from computer analysis, I began playing and replaying the position against the machine. I spent one evening playing more than a dozen games against Stockfish on my iPad. White's resources were revealed in games ending in a draw and even one loss. I won a few, too.
Over the course of several weeks, I analyzed these games on my computer. Then, while writing this article, I played Stockfish 17 on my computer from the position the top of this post.
30...Rc1 31.Kg3 Kf8 32.Kf2
The beginning of a maneuver that the iPad version of Stockfish never found.
32...Ke8 33.Ke3 b4
The correct idea, but here, perhaps, timing is off. Possibly better is 33...Rg1 34.Rf2 Rh1 35.g3 Rg1 36.Rf3 b4
34.axb4
Black to move
My first effort continued 34...Rc7 35.Kd2 Rg1 36.Kc2 Rxg2+ 37.Kb3 and I did not find the only move that holds an advantage: 37...Rg1
35.g3 Rc7
Better than 35...Rc8 due to White's attack on the f-pawn.
36.Kd2 Ra7
36...Rh2+ is better 37.Kc1 Ra7
37.Kc2 Raa1 38.b5 Rhf1
The f-pawn must be protected so that the king can get in front of White's passed pawn.
White to move
Perhaps 39.Rf6 was a better choice.
39...Kd7 40.Bb4 Rf2+ 41.Kb3 Kc7 42.Bc5 Rb1 43.Ba3 Kb6 44.Rc3 Kxb5 45.Rc5+ Kb6 46.Kb4
Another slight inaccuracy, according to Stockfish when it has more time to analyze this position than it did to play it.
46.Ka4 is more stubborn.
Black to move
In this position, my original exchange sacrifice is now correct.
47.Bxb2 Rxb2 48.Ka4
48.Kc3 drops the g3 pawn, but it seems to me that the d-pawn is more important and it soon falls.
49.Rb5+ Kc6 50.Rc5+ Kd7 51.Ra5 Rxd4+ 52.Kb5 Rc4
White to move
Black to move
I stopped playing against Stockfish here because I had places that I needed to go, but I was confident that after 59...Re2 I would have the sort of rook ending that I can win against the engine. Maybe I will continue at a later datew.
No comments:
Post a Comment