Annotated Game #272: The power of the pawn lever

After a disappointing first round loss, I was set to fight back as White and was able to get a classic Stonewall Attack on the board, with the result living up to the opening's name. I'm still learning about the ins and outs of the position type, with the primary lesson from this game being the huge power of the e4 pawn lever when the conditions are right. This idea kept reappearing in analysis, including in a final version with a rook sacrifice. The main point is that White's pieces can be unleashed against an insufficiently defended Black kingside, so the sacrificed material is irrelevant. This is a common theme across similar openings, such as the Colle System, so is important to keep in mind, even early on.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class B"] [Result "1-0"] [ECO "D00"] [Annotator "ChessAdmin/Dragon 3.2"] [PlyCount "59"] 1. d4 d5 2. e3 {my opponent thought for a while here, evidently not familiar with the opening setup.} c6 3. Bd3 Nf6 4. Nd2 e6 5. f4 c5 {actually the best move according to the engines, but of course it's still a waste of a tempo moving the c-pawn twice. This is one of those chess opening paradoxes that it is best not to think too hard on.} 6. c3 {we now have the Stonewall Attack formation.} Be7 7. Ngf3 c4 {gaining space, but at the expense of lessening the central pressure on d4.} 8. Bc2 Nc6 9. O-O {castling is fine, but not the only option.} (9. e4 $5 {this pawn lever is a key idea which I did not fully examine, being a temporary pawn sacrifice.} dxe4 (9... Nxe4 {is similar after} 10. Nxe4 dxe4 11. Ne5 {and now} f5 $2 {loses to} 12. Qh5+) 10. Ne5 $16 {with the c4 pawn threatened, while Black's e4 pawn is indefensible.}) (9. Qe2 {is also a standard preparatory move for the e-pawn lever, both to project force along the e-file generally and to defend the e3 square in particular.}) 9... O-O (9... Ng4 {targeting the e3 square is something White players always need to look out for, although here it's not scary.} 10. Qe2 f5 {and now White has several good options, including the thematic Ne5 or simply h2-h3.}) 10. Ne5 Nd7 {clearing the way for the f-pawn to kick the Ne5, although this increases the level of cramping in Black's position.} 11. Rf3 $6 {this is premature. In addition to the idea of the e4 pawn lever, which was examined before, there are other useful moves.} (11. e4 {still looks best and most aggressive:} f5 (11... dxe4 $6 12. Bxe4 $1 {keeping the b1-h7 diagonal open for the attack is more important than getting the Nd2 centralized immediately, especially since it continues to pressure c4.} Ncxe5 (12... Qc7 13. Ndxc4 $18) 13. fxe5 f5 14. exf6 Nxf6 15. Bf3 $18 {this is essentially a positionally won game for White, whose pieces are all better with no real weaknesses, while Black has to cope with weak e- and c-pawns.}) 12. exf5 exf5 13. Nxd7 Bxd7 14. Qf3 Be6 15. Qh3 $16 {White has attacking possibilities on the kingside and e-files, While Black will struggle to find counterplay.}) (11. b3 {looking to provide an outlet for the dark-square bishop and pressure the c-pawn.}) (11. Ndf3 {giving White's minor pieces more space.}) 11... Ndxe5 $2 {this removes an attacking piece, but White's e5 pawn is very strong as a replacement on the square.} (11... f5 $1 $11 {blocking the attacking diagonal and preventing further f-pawn advances.}) 12. fxe5 $18 {visually it's not hard to see White's advantage, with most of his pieces pointing towards a bare Black kingside.} f6 (12... f5 {now this doesn't work to block the file.} 13. exf6 {followed by e4, after the recapture on f6.}) 13. Rh3 $16 {the pressure down the h-file is good enough for an advantage, but not optimal. White can effectively enter the above variation with} (13. exf6 Bxf6 14. e4 $1 $18 {unleashing the energy of White's pieces.}) 13... h6 14. Qg4 $14 {I have an edge with pressure on the kingside at this point, but still have not unblocked my pieces for development, especially the Nd2 and Bc1; these would all be helped by an e3-e4 pawn break. The other rook also needs to get into play. I thought here for a while and thought g4 would be a flexible place to develop the queen while making threats; h5 is perhaps better.} (14. Qh5 f5 15. Rg3 Bh4 16. Rg6 Ne7 17. Rxh6 gxh6 (17... Bf2+ 18. Kxf2 gxh6 19. Nf3 $16) 18. Qxh4 $16) 14... f5 (14... fxe5 {is preferred by the engine but would appear risky-looking over the board.} 15. Rxh6 Bg5 16. Bh7+ Kh8 17. Rh5 Bh6 18. Bc2 Qf6 {holds.}) 15. Qg3 $16 {maintaining the pin on the g-file and therefore threatening to capture on h6.} Kh8 {breaking the pin.} 16. Nf3 (16. Qg6 {I thought about and rejected because of} Qe8 {which just helps Black's queen move to the defense.} 17. Qg3 $16 {preserves the advantage, however.}) (16. e4 {I did not consider because it seems like just losing the pawn, but the engine prefers it due to the follow-up sacrifice on e4, unleashing White's pieces.} dxe4 $2 17. Nxe4 fxe4 18. Bxh6 $1 $18 {is the point, as the bishop now decisively enters the game.}) 16... Kh7 {covering the g6 and h6 squares, so the e4 sacrificial lever no longer works.} 17. Bd2 {I thought for a while here and judged it best to activate the Ra1.} Rg8 {now I became worried about Black pushing the g-pawn towards a pawn fork on g4.} 18. Qf2 {this is sufficient to keep the advantage.} (18. Re1 $1 {the engine spots a strong sacrifice on e4, again to unleash White's pieces. Moving the rook into the game would also be consistent with the previous move.} g5 19. e4 {now ...g4 is not possible, due to the Bd2 being unleashed against h6.} dxe4 (19... fxe4 20. Qg4 {exploiting the absence of the f-pawn to penetrate Black's kingside, while the Nf3 is protected tactially by the pin on the b1-h7 diagonal.} Bd7 21. Rxe4 $18 {and Black can't take on e4 without exposing his king.}) 20. Rxe4 $1 {the most important idea is to open the diagonal for the bishop.} fxe4 21. Qg4 $1 {the immediate bishop capture on e4 should also be good enough, but the queen's penetration into Black's camp after capturing on e4 cannot be denied.} Kg7 22. Qxe4 $18) 18... g5 {this looks very threatening, but actually loses.} 19. g4 $1 {the only move that keeps the advantage, which I spotted when deciding to play the previous one. The g-pawn is physically blocked from advancing, while the pin on the f-pawn is decisive.} Kg7 20. Qg2 $18 {lining up against the king, which sets up the follow-on tactic.} Bd7 {this is too slow, but other moves don't help much either. If the king tries to run away, the h-pawn hangs.} 21. Rf1 {bringing another piece into the attack before moving to break through.} (21. gxf5 {immediately is also possible and considered best by the engine.} exf5 22. Nh4 $18) 21... Qb6 22. gxf5 exf5 23. Nh4 $1 $18 {this evidently suprised my opponent. Now his position crumbles.} Qxb2 {desperately trying to distract from the kingside attack underway.} 24. Nxf5+ Bxf5 25. Bxf5 Raf8 26. Rg3 {simple and sufficient to win.} (26. Rhf3 $5) (26. Qxd5 {is the engine's choice, but much more complicated to see at the board.} Qxd2 27. Qe6 Rxf5 28. Rxf5 {with mate to follow, unless Black exchanges queen for rook, for example} Qd1+ 29. Rf1 Qxf1+ 30. Kxf1 $18) 26... Kh8 27. Qh3 $1 {now White has too many pieces against Black's thin defense.} Kg7 {my opponent showed surprise at the queen sac that now came.} 28. Qxh6+ $1 Kf7 29. Bd7+ {elegantly cutting off Black's escape at the same time a discovered check is given.} Bf6 30. Qxf6# 1-0

Annotated Game #271: Play what you know - or at least what your opponent does not

The first round of the next tournament I played featured a poor strategic choice on my part. I hadn't properly prepared the opening (a Caro-Kann Two Knights) so was afraid to enter into my own repertoire line. Instead, I chose to play the main line - having no real experience in it - and handed my opponent a big lead in development, positional plus and an early attack. My deficient understanding in the line was bad, but really the strategic opening decision was even worse, choosing something that logically my opponent would have more experience in, rather than "risking" entering into my own repertoire sideline, which it was much less likely my opponent would know, even if my memory was also faulty.

Showing the value of not giving up early, analysis identified where I objectively got back in the game (and could have even achieved counterplay, with some active choices). However, my opponent did not give up her attack either, and eventually broke through. A well-deserved win on her part.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "1-0"] [ECO "B11"] [Annotator "ChessAdmin/Dragon 3.2"] [PlyCount "43"] {[%evp 0,43,28,43,66,57,57,28,63,37,29,33,46,15,44,44,64,56,49,49,45,29,16,-4,21,34,70,77,75,32,35,-20,4,13,13,13,29,23,89,-9,67,147,1096,325,1234,1171]} 1. e4 c6 2. Nc3 d5 3. Nf3 Bg4 {out of my own personal book here, but I had not prepared for the Two Knights variation and did not trust my normal choice of line without that.} 4. h3 Bxf3 5. Qxf3 e6 6. exd5 exd5 $6 {the start of my troubles.} (6... cxd5 {I was overly concerned about Bb5+, but after ...Nc6 there is nothing for White.}) 7. d4 Be7 {rather slow and passive development.} (7... Qf6 $5) 8. Bd3 $16 {now White is significantly ahead in development.} Nf6 9. Be3 Nbd7 $6 {this is too slow, as White is already ready to launch an attack if she wants.} (9... O-O) (9... Na6 {would at least threaten to go to b4.}) 10. Ne2 {this is also rather slow, although the idea of bringing the knight around to the kingside is ultimately useful.} O-O (10... Qa5+ $5 {would be a bit more active.}) 11. g4 {now my opponent decides to get active, although it would have been even more effective the previous move.} g6 $2 {this was played to block the diagonal against the Bd3, but otherwise worsens Black's position, creating a target for the coming pawn storm.} (11... Qa5+ 12. c3 Bd6 $14 {is more active and prevents O-O-O.}) 12. O-O-O $18 {now White is very well developed and has the obvious plan of a kingside pawn storm, while I have no meaningful counterplay.} Ne8 {after this White can just roll through.} (12... Nb6 $5 {at least trying to get another piece active, eyeing c4.}) 13. h4 Nd6 14. Bh6 $6 {this just slows things down for White, but was what I was expecting.} (14. h5) 14... Re8 15. g5 $6 $14 {at this point I felt that I had roughly equalized and Dragon 3.2 agrees. Locking in the Bh6 goes completely against what White needs to do on the kingside, contradicting the idea behind the pawn storm.} Rc8 $6 {this is too optimistic, I should have continued focusing on immediate defense.} (15... Bf8) 16. Rdg1 Ne4 $6 {I thought the interference on the diagonal would be more effective that it turned out to be. White avoids the trap of taking the knight and resumes her pawn storm.} (16... c5 $1 {would be consistent with the previous move.} 17. h5 c4 $11 {hitting the bishop just in time.}) 17. h5 (17. Bxe4 dxe4 18. Qxe4 $2 (18. Qh3 $14) 18... Bxg5+ 19. Bxg5 Rxe4 20. Bxd8 Rxd8 $17) 17... Nf8 $2 {poorly calculated on the defense. Protecting h7 is not enough.} (17... Bf8) 18. hxg6 fxg6 19. Nf4 $2 {now I can take on g5 with either piece, and choose the wrong one.} (19. Qg2 $18 {is best per the engine, but not so easy to find.}) (19. Bxe4 {maintains an edge, although is not necessarily decisive.} dxe4 20. Qh3 $16) 19... Nxg5 $4 (19... Bxg5 $1 $11 20. Bxg5 Nxg5 21. Qg2 Rc7 {and now if} 22. Qxg5 Re1+ $1 {a deflection tactic aimed at the Qg5, which I missed.} 23. Kd2 Qxg5 24. Rxg5 Rxh1 $19) 20. Qg4 $18 {The Ng5 is indefensible and now pinned against g6, with the threat of Bxf8 followed by taking on g6 to break through.} Nf7 21. Bxf8 Bxf8 22. Bxg6 1-0

Book completed - American Grandmaster: Four Decades of Chess Adventures

 

I recently completed American Grandmaster: Four Decades of Chess Adventures by GM Joel Benjamin. I had picked it up as a bargain used book from a dealer at chess tournament a while back, and finally decided to read it during lunchtime at work, my interest primarily being for daily annotated games review. It was not particularly well suited for this, however, as the majority of the actual chess content consisted of individual positions, game fragments, or complete scores but with only a couple comments included. There were enough games at a sufficient level of annotation to be worth going through, but this was a minority of it. If you are looking for something more comprehensive and challenging for annotated games but still with a personal perspective, My Best Games by Victor Korchnoi is world class, while GM Walter Browne's The Stress of Chess...and its Infinite Finesse is a much more detailed and improvement-oriented version of a chess career book with well-annotated games.

From my perspective, it seems the book tried to be multiple things, but was mostly a breezy and candid, but not quite a tell-all, recounting of GM Benjamin's chess career. This mostly consisted of complaints (and occasional praise) about tournament conditions, memories of poor treatment (or occasionally special experiences), and chess politics. The Deep Blue vs. Kasparov project got an interesting treatment and GM Benjamin's perspective was certainly unique, although the narrative was not exactly comprehensive. The book was published in 2007 but the focus was on Benjamin's junior and earlier career, really through the 1990s. As such, a lot of the references - especially to the politics and various failed professional associations - feel quite dated. If you weren't around the 1990s, therefore, it's probably not going to be particularly relevant.

Annotated Game #270: Learning the Slow Slav the hard way

This last-round tournament game is yet another example of the main lesson from Annotated Game #267: How openings are really learned. Here it's a Slow Slav that I had little depth on previously, but studying this game and looking at a couple database examples now have armed me much better for future clashes in the opening.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class A"] [Black "ChessAdmin"] [Result "1-0"] [ECO "D12"] [Annotator "ChessAdmin/Dragon 3.2"] [PlyCount "55"] 1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. e3 Bf5 5. Nc3 e6 6. Nh4 Bg6 7. Nxg6 hxg6 8. Bd3 Qc7 {targeting h2 and prompting White's next.} (8... Nbd7 {is the standard move here.}) 9. h3 Nbd7 10. Bd2 Be7 11. Rc1 O-O 12. O-O {up to this point, although we have not followed the full main line, play has been pretty standard as both sides develop. Now I have to start thinking for myself in the middlegame, without however having much experience with it.} a6 $5 {not a standard idea in this position, but not a bad one either. The point is to possibly prepare b5, while also serving as a waiting move.} (12... Rfd8) (12... Rac8) 13. Qf3 {increasing pressure on d5 and connecting the rooks.} Qd8 (13... dxc4 {is a common liquidation of the center at this stage.} 14. Bxc4 c5 {would be a standard follow-up.}) 14. cxd5 cxd5 $6 {this poor decision is the start of my problems. Opening the c-file is a bad idea, with White's rook already occupying it.} (14... exd5 $11) 15. e4 $6 {however, this is an overly aggressive reaction.} dxe4 {the correct response.} 16. Nxe4 b5 {I saw White's next move, but did not handle it properly.} 17. Rc6 Nb8 $6 {this contributes to the cramping of Black's position and un-develops a good piece.} (17... Nxe4 18. Bxe4 Nf6 $11) 18. Rc2 $14 Nbd7 19. Rfc1 {now the c-file is a much more obvious problem. Essentially I have done nothing in the last few moves while my opponent has managed to double his rooks on the only open file on the board.} Nd5 {best chance to keep things together, centralizing the knight in front of the isolated queen pawn.} 20. Nc5 {the best choice to keep up the pressure. The defense becomes more complex now and I falter.} Bf6 $6 {this pressures d4 but underestimates White's attacking potential.} (20... Bxc5 {I considered this but obviously did not like the creation of a strong passed pawn. However, it should be containable.} 21. dxc5 Ne5 $1 $14 {and the knight can blockade on c6 or exchange off the Bd3, both of which would be helpful.}) 21. Qg4 {now sacrifices on e6 and g6 are in the air.} Nxc5 22. Rxc5 $6 {this lets up the pressure.} (22. dxc5 $16 {the Nd5 is not in as good a position to blockade the c-pawn, plus White's two bishops look dangerous.}) 22... Be7 $11 {enough for equality.} (22... Ne7 {is better, thereby guarding c6 and c8 and g6, while jumping to f5 later would also be good.}) 23. Rc6 Bb4 {not wrong, but it betrays my lack of understanding of the need to keep pieces available to defend the kingside.} (23... Bf6) 24. Bg5 Qa5 $2 {the losing move. Now White unleashes a breakthrough sacrifice.} (24... Be7 $11) 25. Rxe6 {now I realize that I'm simply lost, so try a desperation shot at counterplay.} Qxa2 26. Rxg6 Qxb2 27. Rxg7+ Kxg7 28. Be7+ 1-0

Annotated Game #269: Symmetrical = wild??

Due to having a bye round in the tournament (and having quickly lost my first game as Black), I surprisingly ended up having two Whites in a row. This one, despite being a Symmetrical English, turned into one of the wildest games I have played in at least the past few years. 

Similar to Annotated Game #265 and others, I often get the sense from my opponent's initial hesitation to respond to 1. c4 that they have decided to simply mimic moves, rather than actually follow a prepared repertoire. The problem from White's perspective is that in this opening, that works reasonable well for Black up until around moves 8-9. The positive aspect, as I've also learned, is that once asymmetry is reached, Black may have less of an idea about what to do in the middlegame.

In this game, Black's asymmetric move 9 allows White to obtain a small positional plus and led to what could be a comfortable, even winning middlegame. One of the main lessons for me was a failure on move 14 to kick Black's strong centralized knight on d4 - this is something that I recognize I have repeatedly failed to do in other games, also with poor results. A big benefit of analyzing your own games is recognizing these types of specifc patterns that recur and then correcting them; I articulated this in more detail in Annotated Game #192: The problem of mental perspective. 

Starting on move 17 the game gets a lot crazier and mutual time pressure contributes to a number of swings in evaluation. I wish I had been able to spot the more sophisticated way to play, but felt relatively good at the end of an exhausting wide ride, escaping with a draw.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class A"] [Result "1/2-1/2"] [ECO "A38"] [Annotator "ChessAdmin/Dragon 3.2/Stockfish 16"] [PlyCount "117"] 1. c4 Nf6 2. Nc3 c5 3. Nf3 Nc6 4. g3 g6 5. Bg2 Bg7 6. O-O O-O 7. d3 d6 8. Rb1 Rb8 {this sort of mirror-imaging can be annoying, but ultimately it has to stop somewhere, and does the next move.} 9. a3 Ng4 $6 {this is a dubious maneuver, however, since it does nothing to counter White's queenside plans or help Black's development.} 10. Bd2 $14 Nge5 {this was evidently the idea behind the maneuver, but the knight has taken up time to move to a square which is not evidently better.} 11. b4 (11. Nxe5 Nxe5 12. b4 $14 {is a similar idea to get in the pawn advance while opening the long diagonal; it seems to be easier/better for White to play, since the bishop in the game loses a tempo retreating from f3.}) 11... Nxf3+ 12. Bxf3 Nd4 13. Bg2 {although it's only a small positional advantage for White, I was quite happy here with control of the long diagonal and the d5 square, and having already got in b2-b4.} Bd7 14. bxc5 {this wins a pawn, but Black has significant compensation.} (14. e3 {is the engine recommendation, getting the strong centralized knight out of d4. A common flaw in my play has been leaving such knights in place for too long.} Nc6 15. b5 Ne5 16. f4 Ng4 17. Qe1 $14 {covering e3 and preparing to push the e-pawn.}) 14... dxc5 15. Bxb7 Nc6 $6 {this passive move allows White to keep the pawn and a positional advantage as well.} (15... Bg4 {is the most aggressive response, pressuring e2 and combining well with the centralized Nd4.} 16. Be4 {simply retreating does not help.} (16. Bf4 e5 17. Bd2 {eliminates the threat to the Nc3, but after} Qc7 {Black renews his threats and can pick up the e-pawn again by capturing on b1.}) 16... Rxb1 17. Qxb1 Bxe2 $11) 16. Ne4 $16 {now my own centralized knight causes Black problems.} Na5 {this seems to just lose another pawn, but eliminating the bishop is better for Black.} 17. Nxc5 (17. Bd5 $1 {positionally alone this might be best, but there is also a tactical justification.} Rxb1 18. Qxb1 e6 19. Nxc5 $1 exd5 20. Nxd7 {and the Na5 will be hanging if the Qd8 moves off the diagonal to recapture, so} Nxc4 21. dxc4 Qxd7 22. Rd1 $18) 17... Nxb7 {I thought for a while here and picked the wrong piece to recapture with.} 18. Rxb7 $6 (18. Nxb7 Qc7 19. Na5 $16) 18... Rxb7 19. Nxb7 Qa8 $1 $11 {I had missed this idea, which exploits the weakness of the long diagonal on White's king position.} 20. Na5 {essentially forced} (20. Nc5 Bh3 21. e4 Qc6 22. Be3 Rc8 23. d4 Bxf1 $17) 20... Bh3 21. e4 {this works to block the mate threat, but leaves the e-pawn as a target for a Black pawn lever.} (21. f3 $5 {I don't recall considering this at length, not sure why.}) 21... Bxf1 22. Kxf1 $15 {Black has a small advantage in the middlegame, as his pieces will be able to coordinate better and White's two extra pawns are not yet able to be mobilized.} Rb8 (22... f5 {would immediately hit at White's e-pawn and threaten to open more lines for the rook.}) 23. Nb3 $6 {it's understandable that I wanted to get the knight off the rim, but this is too slow.} (23. Bb4 $11 {mobilizing the bishop with tempo, as it threatens e7.}) 23... Rb7 (23... Qc6 $1 $17 {and Black threatens to move to a4, causing greater problems for White.}) 24. Ke2 {it's still a little early to be moving the king. Again, mobilizing the bishop looks better.} Qb8 25. Bb4 {I finally hit on the best idea, although it is not quite as effective as before. At least the b-file is shut off, and I felt that I could hold the position now.} Bd4 $4 {both my opponent and I hallucinated that the bishop was safe on the square. Mutual time pressure likely played a part.} 26. Nc5 $4 (26. Nxd4 {and wins.}) 26... Bxc5 (26... a5 $1 {is even better.} 27. Nxb7 axb4 28. Na5 bxa3 $19 {and the advanced passed a-pawn will be decisive.}) 27. Bxc5 Qe5 28. Bb4 Qh5+ $6 {this sequence was what I had anticipated with my (erroneous) calculations on move 26, and in fact it is balanced.} (28... a5 $17 {is the idea that both my opponent and I overlooked, which would allow the rook to penetrate.}) 29. Ke1 Qe5 30. Qd2 {again failing to defend against ...a5. However, luckily my opponent continues to miss it as well.} (30. Qa4) 30... Qa1+ 31. Qd1 Qd4 32. Qd2 $6 (32. Qc2) 32... e5 33. Bc3 {some complex and rather stressful defensive calculations were required here.} (33. Ke2 $5 {looks simpler.}) 33... Rb1+ 34. Ke2 Qd6 35. Bb4 $11 {now there should not be any threat of a breakthrough.} Qf6 36. Bc3 {just playing it safe.} g5 {my opponent meanwhile continues to try to press.} 37. Qe3 Qe7 {covering a7.} 38. Bb4 Rb2+ 39. Kd1 {keeping the king attached to the defense of the f2 pawn is a much better idea.} (39. Ke1) 39... Qf6 {immediately targeting f2.} 40. Qc5 $2 {this aggressive counterattacking defense has a major flaw, but Black overlooks it in time pressure.} (40. Bc5) (40. Bc3) 40... Qd8 $6 (40... h6 $1 $19 {and Black's king will be able to eventually avoid White's checks, while the Q+R combo can effectively penetrate and attack White's position.}) 41. Qd6 $11 {forcing the queen trade, to what should be a balanced position.} Qxd6 42. Bxd6 Rxf2 $2 {this is too greedy, but I miss the refutation.} (42... f6 $11) 43. Bxe5 $6 {thinking too materialistically here.} (43. c5 $1 $18 {passed pawns must be pushed! Leaving the bishop in place covers the f8 square, so Black's king is cut off.}) 43... Kf8 $11 44. c5 {one tempo too late.} Ke8 {as often happens in endgames, sometimes good-looking moves have subtle refutations.} (44... Ke7 $11) 45. h4 $2 {I was obviously worried about protecting the h-pawn, but this is done better with} (45. g4 $11) (45. d4 $1 {and the central connected passed pawns win, as the engine points out.} Rxh2 (45... Rf1+ 46. Kd2 Rf2+ 47. Ke3 Rc2 48. Kd3 Rc1 49. Bf6 $18) 46. d5 $18) 45... f5 $1 $19 {now Black should be winning.} 46. Ke1 Ra2 $2 {now White should be winning.} (46... Rf3 $1 $19) 47. exf5 $1 gxh4 48. gxh4 Rxa3 49. Ke2 $6 {another subtle endgame misstep and we are back to equality.} (49. Kd2 {and the king can chase the rook on the queenside.}) 49... a5 $11 50. c6 Ra2+ 51. Ke3 Rc2 52. c7 a4 53. d4 $2 {looks reasonable, but should lose.} (53. Kd4 a3 54. Kd5 Kd7 55. f6 a2 56. f7 Rf2 57. c8=Q+ Kxc8 58. Ke6 Kd8 $11 {and now neither pawn can queen.}) 53... a3 $19 54. d5 a2 55. d6 Kd7 56. f6 {desperate pressure added to Black, but it works. There is only one winning move.} a1=Q (56... Rc5 $1 57. Bd4 Rc1 $19) 57. Bxa1 Kxd6 58. f7 Ke7 59. Bf6+ {this is a little flashy and sets an obvious trap, but my opponent sees through it. Draw agreed.} (59. Be5 {also draws.} Kxf7 60. Kd3 $11) 1/2-1/2