See You in Baku

I just wanted to let anyone who happens to read this know why I am not posting these days. There are two reasons. One is that I have been finishing a book that I have been writing. The second reason is that I am moving from Iceland to Baku, Azerbaijan. We will arrive in Baku in August. I won't be able to play chess for some time, since I will be settling in to the new position, but I hope to eventually play in Baku. They sure have a strong chess tradition, including being the birthplace of Garry Kasparov. If any Azeris see this blog, perhaps you can help me out with finding good FIDE-rated events in Baku. I've tried emailing the clubs and federations that I could find, but no one responded.

Reykjavik Open 2009 Rd. 9

I have played in many tournaments in my life, and 455 tournament games, yet this event has been the worst performance I have had. Given how the rest of the tournament went, this last round was fitting. I played a young boy with no official rating - though FIDE shows that he has just begun playing and so far has a 1670 rating - and he played like each of my previous opponents has - mistake free. I have had games in the past where the computer showed no real mistakes by my opponent, but this is the first time I have had an entire tourney with no such mistakes!

I noticed some interesting patterns between this year's tournament and last year's. Last year I never once had to play down against a player lower rated than myself, and I gained around 30 rating points. This year I had to play down five times, and I lost around 18 rating points. Last year I didn't lose a single game with white; this year I nearly lost every white and didn't lose any with black!

[Event "Reykjavik Open 2009"]
[Site "Reykjavik, Iceland"]
[Date "2009.04.01"]
[Round "9"]
[White "Cross, Ted"]
[Black "Karlsson, Mikael"]
[Result "1/2-1/2"]
[ECO "B40"]
[WhiteElo "2076"]

1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nf6 5. Bd3 Qb6 6. Nb5?!
The proper move here is 6. Nb3, but I was feeling particularly aggresive and I didn't care if he took the pawn.
6...a6
He would have been fine taking the pawn: 6...Nxe4 7. Bxe4 Qxb5 8. Nc3 Qb6 9. O-O d5 and black has a small edge.
7. Be3
The computer likes 7. N5c3 better.
7...Bc5 8. Bxc5 Qxc5 9. N5c3 Nc6 10. Nd2 b5 11. O-O Bb7 12. Kh1 O-O 13. f4 d6 14. a3 Rfd8 15. Nb3
Looking back on it, I don't like this move. The computer recommends it, yet in the game this knight was just out of play for a long while.
15...Qb6 16. Qe2 Qc7 17. Rad1
The computer also recommends this, but if I was really going to follow the plan that you see over the next few moves, then this is just a wasted tempo.
17...Rac8 18. Rf3
I told myself not to do this. I knew I wouldn't play like this against a higher rated opponent. Yet, I felt that lower rated players often panic when someone directly attacks their king, and I didn't quite like the idea of playing g4 and g5 instead, since my king is potentially exposed to the black bishop.
18...Ne7 19. Rh3?
I couldn't back down, even though I knew I should. Since I could see that his knight was swinging over to g6, this move doesn't do any good anymore. Perhaps it would have made better use of itself on g3. Best of all would have been to abandon that line of play and just play it back to f2 or f1.
19...Ng6 20. Rf1
You see how moving this rook to d1 was a waste of time?
20...e5!
An excellent move that highlights the drawbacks of my overly aggressive plan. I either have to allow the opening of the d file for black's rook or further weaken my kingside with g3. Neither is appealing. So, black gets a perfect outpost for his knights on f4.
21. fxe5 dxe5 22. Rg3 Nf4 23. Qf2
I really thought I had him here. So many 1600 level players would have gone right in for the knight trade on d3, but given the way this tournament has gone, I am not surprised he didn't go for it.
23...N6h5
I had hoped for 23...Nxd3?? 24. Qxf6 and white wins.
24. Rf3 Nxd3 25. cxd3 Nf4 26. Nc1
Here I missed the chance for equality with 26. d4!
26...Rd7 27. g3
Better was 27. N3e2 f6 (27...Nxe2 28. Nxe2 Rcd8 29. b4 f6 30. Rc1 Qd6 31. Rc3 Rc7 32. Rxc7 Qxc7 33. h3 with a slight edge to black.) 28. Nxf4 Qxc1 (28...exf4 29. Ne2 g5 30. Nc3 =) 29. Nh5 Rf7 30. b4 Qc2 31. Qb6 Qc6 32. Qf2 Qe6 with a slight edge to black.
27...Ne6 28. Nd5 Bxd5
Black overlooked a much easier win here with 28...Rxd5! 29. exd5 Bxd5.
29. exd5 Nd4 30. Re3 Rxd5 31. Re4 Rc5 32. Ne2 Nxe2 33. Rxe2 Rd8 34. d4! exd4 35. Qxd4 Qc6+ 36. Qe4 g6?!
Black should have played 36...Qxe4+ 37. Rxe4 Rc2 38. Rfe1 g5 39. R4e2 Rdd2 40. Rxd2 Rxd2 41. b4 Rd3 42. Ra1 with a good advantage for black.
37. Qxc6 Rxc6 38. Re7 Rf8?!
Much better was 38...Rc2 39. Rexf7 Rdd2 40. Rf8+ Kg7 41. R1f7+ Kh6 42. Kg1 Rxh2 43. Rf2 Rhxf2 44. Rxf2 Rxf2 45. Kxf2 Kg5 and black should win.
39. Ra7 Kg7 40. Rf2 h5 41. h4 1/2-1/2
Though my move is the last one, it was actually he who offered the draw. He is too new to chess to know that he has to make his move first before offering a draw, I guess. I was lucky to escape alive. So, how come 1600 level players are all of a sudden playing like experts?

Reykjavik Open 2009: Photos from Round 1

A friend took some photos in round 1, and I just received them. Oddly enough, after so many tournaments, this is the first time I have photos of me playing in a tournament (not a simul):

Here is top seeded Ukrainian grandmaster Alexander Areshchenko in round 1, playing against fellow Ukrainian Anastazia Karlovich.

U.S. champion GM Yuri Shulman is the second seed.

Here I meet my first round opponent, Icelandic grandmaster Throstur Thorhallsson.

Me playing against GM Thorhallsson.

Just before the game begins.

One last shot of me playing in round 1.

Reykjavik Open 2009 Rd. 8

I finally managed to play a pretty good game from start to finish. I am happy with this one. I learned my lesson from two rounds ago and decided to play a brand-new opening line so I would avoid my opponent's preparation. It is strange playing something that you don't know, but I don't think I made too many mistakes.

[Event "Reykjavik Open 2009"]
[Site "Reykjavik, Iceland"]
[Date "2009.03.31"]
[Round "8"]
[White "Thorsteinsdottir, Hallgerdur"]
[Black "Cross, Ted"]
[Result "0-1"]
[ECO "B22"]
[WhiteElo "1951"]
[BlackElo "2076"]

1. e4 c5 2. c3
I knew from the databases to expect either an Alapin Sicilian or a Rossilimo Sicilian, so I prepared for both. This is the Alapin.
2...Nf6
My surprise! I have never played this variation before. I prepared as best I could, but there is just too much to actually learn in one day.
3. e5 Nd5 4. d4 cxd4 5. Nf3 Nc6 6. cxd4 d6 7. Bc4 e6 8. O-O Be7 9. exd6
This is where she took me out of my preparatory knowledge.
9...Qxd6 10. Nc3 a6
Other moves played here are 10...O-O or 10...Nxc3.
11. a3
This is a novelty according to my database. 11. Ne4 has been played previously.
11...O-O 12.Qd3 Rd8 13. Bg5 Bxg5 14. Ne4 Qe7
The computer really likes black now and likes a line beginning with 14...Nf4. I actually considered this move, but it looked far too complicated to actually try in this situation. The mainline analysis runs 14...Nf4 15. Nxd6 (15. Qb1 Qc7 16. Nexg5 h6 17. Qh7+ Kf8 18. Qh8+ Ke7 19. Qxg7 hxg5 20. Nxg5 Rxd4 21. Rad1 (21. Qxf7+ Kd6 22. Rad1 Qxf7 23. Nxf7+ Ke7 24. Rxd4 Nxd4 25. Rd1 Nf3+ 26. gxf3 Kxf7 and black wins.)) 15...Nxd3 16. Nxg5 (16. Bxd3 Bf6 17. Ne4 Bxd4 18. Nxd4 Nxd4 with a large edge for black.) 16...Nxb2 17. Ngxf7 Nxc4 18. Nxd8 Nxd6 19. Nxc6 bxc6 and black has a large advantage.
15. Nexg5 f5!?
Much safer is 15...Nf6.
16. Rfe1 Qf6 17. Rad1 h6 18. Nh3 g5 19. Ne5 Bd7 20. Bxd5 exd5 21. Qb3 Rab8
I knew I had to give up the central pawn, but I thought I saw more than enough compensation for it.
22. Qxd5+ Be6 23. Qc5 Rd5 24. Qc3 Rc8 25. Nxc6 Rxc6 26. Qe3 f4 27. Qe4 Rcd6
Amazingly, all of these past moves are the computer's first choice!
28. f3 Bxh3 29. gxh3
White had a better play for equality here. The line is long and there are many possibilities, so I will just give the computer's main line: 29. Qe8+ Qf8 30. gxh3 Qxe8 31. Rxe8+ Kf7 32. Re4 Rb5 33. Rd2 Rb3 34. Kf2 Rdb6 35. Ree2 Re6 36. Rxe6 Kxe6 37. h4 gxh4 38. d5+ Kd6 39. Kg2 a5 40. Kh3 Rxf3+ 41. Kg4 Rf1 42. Kxh4 Rc1 43. Kh5 Rc5 44. Kxh6 Rxd5 45. Rf2 Rf5 46. Kg6 Rf8 47. h4 Ke5 48. h5 Rg8+ 49. Kf7 Rh8 50. Rh2 f3 51. Kg7 f2 52. Rxf2 Rxh5 =.
29...Rd8
This had to be played or I could get into serious trouble.
30. Qe6+ Qxe6 31. Rxe6 Rxd4 32. Rxd4 Rxd4 33. Rxh6 Rd1+ 34. Kg2 Rd2+ 35. Kg1 Rxb2
I was certain I could win this endgame.
36. Rg6+ Kh7 37. Rxg5 Rb3 38. Rf5 Rxf3
I overlooked the more convincing 38...Rxa3 39. Rxf4 b5 and wins. However, I am still winning.
39. Kg2 Rxa3 40. Rf7+ Kg6 41. Rxb7 a5 42. Rb5 a4 43. Ra5 Ra1
This wins because once the pawn comes to a2 I can move the rook with check against her king (at the right moment). If she tries to avoid this, then I get to move the 'f' pawn down.
44. h4 a3 45. Ra6+ Kh5 46. Kh3
This is a very good try for her, looking for stalemate.
46...f3!
This must be played. Pushing the pawn to a2 would be a terrible mistake: 46...a2?? 47. Rxa2 =.
47. Ra5+ Kh6 48. Ra6+ Kg7 49. Ra7+ Kf6 50. Ra6+ Ke5 51. Kg3 a2 52. h5
If she tried 52. Kf2 then 52...Rh1 wins due to the skewer of the king and rook once she takes on a2.
52...Rg1+ 0-1
I am very happy with my play. My opponent didn't make any major mistakes (the theme for this event!), yet I was still able to find enough pressure to earn the win.

Reykjavik Open 2009 Rd. 7

Well, what had so far been a dismal tournament has now officially sunk below anything I ever imagined in my worst chess nightmares. Sorry, I have nothing positive to say on this, because that is the truth. Besides shedding a boatload of rating points, I feel as if my opponents play as perfectly as computers and there is little I can do about it. Today I faced the lowest rated opponent I have faced in ages, only 1660 ELO. I am more than 400 points higher rated, so my statistical chances are supposed to be about 99%, if I recall correctly. Well, this opponent played as mistake-free as any of the much higher rated opponents I have faced. I kept finding devious little ways of trying to win in the endgame, but she never fell for a single one, though in my past experience 1600 level players would have fallen for at least one of them. I know it's just how I feel right now, but I feel like I should give up chess. If I were prone to that sort of thing, I would weep.

[Event "Reykjavik Open 2009"]
[Site "Reykjavik, Iceland"]
[Date "2009.03.30"]
[Round "7"]
[White "Cross, Ted"]
[Black "Finnbogadottir, Tinna"]
[Result "0-1"]
[ECO "B34"]
[WhiteElo "2076"]
[BlackElo "1660"]

1. e4 c5 2. Nf3 g6
Ah, of course. Every game I saw in the database had her playing consistently with 2...e6, so naturally she varies now. It's what I thought I should do yesterday, but didn't and paid the price.
3. d4 cxd4 4. Nxd4 Nc6 5. Be3 Nf6 6. f3
I am sure she didn't know this, but this was a perfect opening choice against me, because I do not allow the normal dragon, yet she got me into it. Now I have to play an opening I don't know.
6...Bg7 7. Nc3 O-O 8.Qd2 d5 9. exd5?!
This shows my lack of knowledge of this version of the dragon.
9...Nxd5 10. Nxd5 Qxd5 11. Nxc6 bxc6 12. c3 Rb8 13. b3?
A terrible move (see, I am the only one who makes mistakes these days!). I could have had a slight edge with 13. Qxd5 cxd5 14. O-O-O
13...Qa5 14. b4 Qe5
Better is 14...Qa3! 15. Rc1 Bf5 16. Bc5 Rfd8 17. Qe3 Qb2! 18. Be2 Bd3! 19. Kf2 Bxe2 20. Rb1 Qxa2 21. Ra1 Qb2 22.Rab1 Qxc3 23. Qxc3 Bxc3 24. Kxe2 and black should win this.
15. Rc1 Bf5 16. Bc4 Bh6 17. Kf2 Rfd8 18. Qe2 Bxe3+ 19. Qxe3 Qxe3+ 20. Kxe3
You see, even with a less than stellar opening, I have fought back to a slightly better endgame. If my opponent only played like a normal 1600 player, I could expect to win this!
20...Rd6 21. Rhd1 Rbd8 22. Bb3
I played this because in my past experience, this level of player, more often than not, likes to trade down, and this looks like a very plausible plan here, though in fact I would likely win if she did. So, naturally she doesn't.
22...Be6
Of course. If 22...Rxd1 23. Rxd1 Rxd1 24. Bxd1 I am pretty sure I could win this given my centralized king.
23. Rxd6 exd6 24. Rd1 Re8 25. Bxe6
The computer likes 25. Kf4 better, but it just leads to equality after 25...Bxb3 26. axb3 d5 27. Rd2 Re1 28. b5 cxb5 29. Rxd5 a6 30. h4 Rb1 31. b4 =.
25...Rxe6+ 26. Kd3 Kf8 27. c4 Ke7 28. a4 Kd7 29. Rd2 Kc7 30. Re2 Kb6
Of course she doesn't trade rooks like a normal 1600 player often would.
31. a5+ Ka6 32. Rb2
My move isn't really a mistake, but easier to play would be 32. Re4 c5 33. Kc3 Rxe4 34. fxe4 h5 35. g3 cxb4+ 36. Kxb4 f6 37. Ka4 Kb7 38. Kb5 a6+ 39. Kb4 Kc6 40. Kb3 Kc5 41. Kc3 f5 42. exf5 gxf5 43. h3 d5 44. cxd5 Kxd5 45. Kd3 Kc5 46. Ke3 Kb4 47. Kf4 Kxa5 48. Kxf5 Kb4 49. g4 =
32...f5 33. Kd4 Re5 34. h4?!
I keep making my life harder than it needs to be. Better is 34. Rd2.
34...f4 35. Rd2 c5+ 36. bxc5 Rxc5 37. Kc3 Kxa5 38. Rxd6 Rh5 39. Rd5+??
Strange. I looked at this move first and thought it didn't work. Then I looked at the correct move, 39. Rd4 and wasn't sure if it quite worked or not. When I looked at my current terrible move again, I thought I saw something better. I just miscalculated. Drawing would be 39. Rd4 Rxh4 40. g4 g5 41. Rd5+ Kb6 42. Rxg5 Rh3 43. Rb5+ Kc6 44. Kd4 Rxf3 45. Rf5 Rf1 46. Rf6+ Kc7 47. Rf7+ Kc6 =.
39...Rxd5 40. cxd5 Kb6 41. Kd4 Kc7 42. Ke5 Kd7?
The first and only mistake that she makes in the entire game! Flat out winning is 42...a5 43. h5 a4 44. hxg6 hxg6 45. Kd4 Kd6.
43. Kxf4 a5 44. Ke4 a4 45. Kd4 Kd6 46. g3??
I blow it. I could have held the draw with 46. Kc4 a3 47. Kb3 Kxd5 48. Kxa3 Ke5 49. Kb3 Kf4 50. Kc4 Kg3 51. Kd4 Kxg2 (51...Kxh4 52. Ke5 Kg3 53. Kf6 Kxg2 54. f4 Kg3 55. Kg5 h5 56. Kxg6 =) 52. f4 Kg3 53. Ke5 h5 (53...Kxh4 54. f5 gxf5 55. Kxf5 =) (53...Kg4 54. h5 =) 54. f5 gxf5 55. Kxf5 =).
46...a3 47. Kc3 Kxd5 48. h5 a2 49. Kb2 Kd4 50. Kxa2 Ke3 51. f4 Kf3 52. Kb3 Kxg3 53. hxg6 hxg6 54. f5 gxf5 55. Kc3 f4 56. Kd2 Kg2 0-1
Of course, playing the stellar way that she did, she deserves the win. I just don't understand how a 1600 player plays like this. I know she did it on her own, because she rarely left the board, so congratulations to her, and massive frustration for me.

Reykjavik Open 2009 Rd. 6

I never have photos of me playing chess, so it is nice that someone posted this one on the skak.is blog. This is me on the right playing today's opponent, while in the center of the photo is the lady to whom I lost in the next round.

I deserved what I got today, and possibly should have even lost. As black I have a pretty predictable set of lines against 1. e4, and that bit me hard today. I considered playing something I have never played before, just to throw off my opponent's preparation, and I should have done so I guess, but I decided to stick with what I knew. My opponent planned things perfectly, using a line I have never had played against me and which I have never studied. I was lucky to get a draw...

[Event "Reykjavik Open 2009"]
[Site "Reykjavik, Iceland"]
[Date "2009.03.29"]
[Round "6"]
[White "Eiriksson, Vikingur"]
[Black "Cross, Ted"]
[Result "1/2-1/2"]
[ECO "B33"]
[WhiteElo "1882"]
[BlackElo "2076"]

1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Bg5 a6 8.Na3 b5 9. Nd5 Be7 10. Bxf6 Bxf6 11. c3 O-O 12. Nc2 Bg5 13. h4
My opponent deserves full credit. He found a gaping hole in my opening knowledge here. I knew instantly that I was going to suffer. I know only the next two moves of this line, and I just knew that my opponent had spent all day studying this line on his computer. He confirmed that afterward.
13...Bh6 14. g4 Bf4 15. Qf3 Be6
A common alternative, though I didn't know it, is 15...Rb8.
16. Bd3
This move is fine, but also decent is 16. Nxf4 with a game so complex that I am not even going to try to give you any analysis.
16...Ne7?
Here is where my lack of knowledge hurt. This is bad. Instead, 16...Re8 is fine here.
17. Nxf4 exf4 18. Qxf4 Ng6 19. Qg3 Ne5 20. Be2 Bc4?!
Again not so good. Better was still 20...Re8.
21. f4 Nc6?!
Better is 21...Nd7.
22. Bxc4 bxc4 1/2-1/2
I offered a draw because I saw more danger of losing than of being able to make any sort of comeback to win. I saw some possible pressure against the 'e' pawn if he went after the pawn on c4, but he could just castle queenside and my attack would probably not be enough.

Reykjavik Open 2009 Rd. 5

I have noticed a pattern - my opponents never blunder. Oh, they make occasional second-best moves, but overall they just don't mess up. I believe I have played relatively well in this event, yet my results are miserable, and if you examine each of the games you will notice that my opponents never really make any mistakes. I have to assume that this is because my playing style is not generating enough confusion...

Today I played a new line (for me) of the French, simply because this line was the only example of a French that I had for this opponent. I played relatively well, yet I never gained any advantage, and my opponent kept playing almost all the best moves throughout the game, so that slowly but surely I was strangled. Note that my old Fritz 8 is not sufficient to be able to analyze the complexities that arose in this game, so I don't trust all of the analysis. I sure wish the new Fritz 11 or the Rybka 3 were more affordable.

[Event "Reykjavik Open 2009"]
[Site "Reykjavik, Iceland"]
[Date "2009.03.28"]
[Round "5"]
[White "Cross, Ted"]
[Black "Halldorsson, Gudmunder"]
[Result "0-1"]
[ECO "C03"]
[WhiteElo "2076"]
[BlackElo "2248"]

1. e4 e6 2. d4 d5 3. Nd2
I have never played this variation before, but it was the only one for which there was an example of my opponent's previous play.
3...Be7 4. Ngf3 Nf6 5. Bd3 c5 6. dxc5 O-O
I'm sure he purposely varied from his previous game in this line, which was played earlier in this same tournament against a GM. He played 6...Nc6 in that one.
7. O-O a5 8. a4
This is a novelty according to my database, and while my computer seems to think it is okay, I felt throughout the game that I had some trouble because of this move. ({RR} 8. Re1 Na6 9. exd5 Qxd5 10. Re5 Qd8 11. Nb3 a4 12.Nbd4 Nxc5 13. Bc4 Qc7 14. Nb5 Qb6 15. Nfd4 Bd7 16. Qf3 Rac8 17. Be3 Qa5 18. c3Bxb5 19. Bxb5 Bd6 20. Rg5 Nce4 21. h4 h6 22. Rg4 Nxg4 {Martinez Lopez,E-Bermejo Martinez,J/Mondariz 2003/CBM 95 ext/0-1 (37)}) ({RR} 8. Qe2 Na6 9. c4 Nxc5 10. cxd5 Nxd3 11. Qxd3 b6 12. Re1 Ba6 13. Qb3 Bc5 14. h3 exd5 15. e5 Ne8 16. Nf1 Nc7 17. Bg5 Qd7 18. Ng3 Ne6 19. Rad1 d4 20. Ne4 d3 21.Be3 Bb4 22. Bd2 Nc5 {Ruiz Jimenez,F-Baron Rodriguez,J/Mondariz 2002/CBM 91 ext/1/2-1/2 (51)})
8...Na6 9. e5
I didn't even look at the move the computer likes best: 9. exd5!? Qxd5 10. Bxa6 Rxa6 11. Nb3 and my old Fritz 8 says white is a little better here.
9...Nd7 10. Nb3 Naxc5 11. Nxc5 Nxc5 12. Bb5 Bd7 13. Qe2 Rc8 14. Be3 Bxb5 15. Qxb5 Qc7 16. Rfe1 Rfd8 17. c3 Qc6 18. Nd4 Qxb5 19. Nxb5 Nb3
I am sure there are ways to improve in the previous moves, though nothing really stands out. This 19...Nb3 move really hurt, because I had to choose between giving up a pawn and remaining somewhat active, or placing my rook on a terrible square.
20. Ra2 Rc4 21. Bb6
An alternative worth exploring is 21. Rd1.
21...Ra8
The computer likes 21...Rd7 much better here and gives black a slight edge.
22. g3
I actually looked at the move the computer likes best - 22. Bc7!? - and it didn't look bad to me, but I wasn't sure enough about it.
22...g5 23. h3
23. f3 or 23. Rd1 are possibly better alternatives, but too complex to give lines here, especially since my computer isn't up to snuff.
23...h5 24. Nd6
The computer also likes 24. Kg2, but again I don't trust the analysis.
24...Rc6 25. Be3 b6 26. Nb5 Nc5 27. f4
My computer can't make up its mind about the move I played versus 27. Rd1.
27...gxf4 28. gxf4
The computer prefers that I first play 28. Nd4 Rcc8 and only then 29. gxf4. I imagine this is because it keeps black's knight out of b3.
28...Kh7 29. Kh2 Rg8 30. Rg1?!
Here again I had the chance to keep the knight out of b3 with 30. Nd4.
30...Rxg1 31. Kxg1 Nb3
You see, my rook gets stuck again.
32. Na7!?
The computer prefers 32. Kf2, but it seems to just lead to a lost game anyhow: 32. Kf2 Kg6 33. Kf3 Bc5 34. h4 f6 35. Na7 Rc7 36. Nb5 Rc8 37. Bxc5 bxc5 38. Nd6 Rd8 39.Ra3 c4 and black has a winning advantage. During the game I actually thought I might be winning here, but I was soon to be proven wrong.
32...Rc4 33. Bxb6 Kg6!
I had only looked at him taking on f4 with the rook, whereupon I thought I could win with 33...Rxf4 34. Nc6 Bc5+ 35. Bxc5 Nxc5 36. b4! Ne4 37. bxa5 Nxc3 38. Rf2 Rxa4 39. Rxf7+ Kg6 40. Rf6+ Kg5 41. Rxe6 and it is white with the winning chances.
34. Kg2
The alternative 34. Be3 still just loses to 34...Kf5 35. Nb5 Rxf4! 36. Bxf4 Kxf4 37. Kf1 Kxe5 38. Na7 Bf8 39. Ke1 Ke4 40. Kd1 Kd3 41. Nc6 f6 42. Nd8 e5 43. Ne6 Bd6 44. Ng7 e4 45. Nf5 e3 and black wins.
34...Kf5 35. Be3?
Better was 35. Nb5 Kxf4 36. Nd4 Nxd4 37. Bxd4 Rc8 38. Ra1 Rb8 39. Rf1+ Ke4 40. Rf2 Bh4 41. Rd2 Bg5 42. Rc2 Kd3 43. Rf2 Bd2 though black still has a large edge.
35...Ke4 36. Kf2
It's hard to call this a mistake when everything loses now. For instance 36. Bf2 Kd3 is also losing.
36...Bh4+
Also winning was 36...Kd3!.
37. Ke2 d4! 38. Bxd4
I know this looks like a terrible blunder, but it really is no worse than anything else. I was vaguely hoping I could find a way to pick up the 'a' pawn and start pushing my pawns. Just as bad was 38. Bd2 d3+ 39. Kd1 Bg3 40. Nb5 Nxd2 41. Kxd2 Bxf4+ 42. Kd1 Rc5 43. b3 Rxe5 44. Rf2 Rg5 45. Rf1 Rg2 46. Re1+ Kd5 47. c4+ Kc5 and black wins.
38...Nc1+ 39. Kd2 Nxa2 40. Nb5 Kxf4 41. b3 Rc6 42. Kc2 Nb4+ 43. Kd2 Nd5 44. c4 Ne7 45. Bc3 Ra6 46. Nc7 Ra7 47. Nb5 Rd7+ 48. Ke2 Nc6 49. c5 Bd8 50. Nd6 Nxe5 51. Bd2+ Kg3 0-1
It gets frustrating to watch other games and see people making mistakes, while in my own games the only one who seems to ever make mistakes is me...