Shakmaty Bereolos - The Official Chess Site of Peter Bereolos

11/22/02 - Kings Island Open

Last weekend was the 11th edition of the Continental Chess Association's Kings Island Open. This year's event seemed much stronger than in the past featuring 9 GMs, 5 IMs, and 1 WGM in the open section.

I haven't had a lot of time to analyze my games in detail, so I'll just show some positions of interest. In round 1, I had White against IM Ben Finegold's wife Kelly. She wrote Kelly Finegold on her scoresheet, but was still Kelly Cottrell on the wall chart. I got a very nice position out of the opening as can be seen in the position after 16...Bd6

This position is very strange in that it is symmetrical, except that White has managed to obtain 7 extra moves! Bf4, b4, Na4-c5, Qb3-c3, and Rc1. I don't recall running into anything like that before.

I won that game and, oddly enough, in Round 2, I played Ben Finegold with Black. We reached a theoretical position in the Modern Benoni after 17. Qd2

This position is roughly level, but Black needs to make sure Ng7 does not become passive. Kapengut gives 17...Be5 18. Bf4 Bxf4 19. Qxf4 Re8 with the idea of f5, exploiting the position of the knight, as equal. I played 17...a6 planning 18. a4 Rb8 19. a5 b5 20. axb6 Rxb6 with the idea Ng7-e8-c7-b5 but it was probably better to implement this plan by first maneuvering the knight starting with 17... Ne8 18. a3 b5 This may be a mistake. I saw his plan, but didn't evaluate the position correctly. Again, 18... Ne8 or 18... Be5 should be considered, when b4 can be met by b6. 19. b4 c4 20. a4 Bd7 21. Bg4 h5 I had been planning 21... a5 but when we got here I didn't like the looks of the exchange sacrifice 22. Bxd7 axb4 23. Nxb5 Bxa1 24. Qxb4 Qxd7 25. Rxa1 when White has two nice bishops and should be able to pick up the c4 pawn to have two pawns for the exchange. 22. Bxd7 Qxd7 23. Bd4 Bxd4 24. Qxd4 Now the placement of the two knights puts the position in White's favor. He went on to win.

In Round 3 I had White against William Harris. I managed to get a huge space advantage out of the opening that had his pieces stepping on each others toes. I got to finish things off with a nice tripling of major pieces on the h-file 26. Qh3 [1:0] as his knight drops off.

I had Black in Round 4 against David Friedman. Around the time control I picked up a couple of pawns and looked to be cruising to victory. I got a bit careless and ended up giving back one of the pawns and had to trade quite a few others. Still, I think the position after 70. Bd7 is winning for Black.

Here, Black should transfer the knight to d4 so that in lines after c2 and Kc3 he will be threatening mate in 1. Instead, shades of my game with Chris Chambers, I think I pushed prematurely. 70...c2? 71. Be6+? I still need to look at this one more carefully, but it seems like this may be an exchange of blunders. I thought I could win by triangulation after 71. Kb2 Kd3 72. Kc1 Kc3 73. Ba4, but now looking at it, I don't see a way to do it since White has plenty of squares for his bishop. 71... Kc3 Now Black is going to be able to reposition the knight since White isn't in position to attack c2. 72. Bf5 Nxf5 73. gxf5 73. g5 Nd4 -+ 73...Kd3 [0:1]

This put me a half a point up on the rest of the U2400 field heading into the last round. I was the odd man so actually ended up getting paired top seed GM Alexander Goldin! This did enable me to get my due White. He played the same variation that GM Serper played against me a few years ago, so I faced a difficult decision: Should I follow the line played in that game, which leads to a =/+ ending that isn't too difficult to hold, or should I deviate and try to take advantage of a rare White against a GM? I ended up deciding based on the tournament situation. A draw guaranteed me a share of the U2400 prize. It may have been a somewhat cowardly decision, but I don't think there is much shame in drawing a player rated more than 300 points above you. So, we chopped the pieces down into the ending, which I held without much trouble. One interesting point came just before the end when he played 45...Nd6

Earlier, when the h-pawns were on their original squares, the pawn ending would have been a dead draw after an exchange on d6 followed by g4. Likewise, if the f-pawns were exchanged, the position holds no ideas. Here, however, I thought he might be able to engineer a breakthrough with ...g5 and ...f4. It turns out that this doesn't work. His king has to be close enough to stop my g-pawn, but then d5 gives White counterplay. Intuitively, during the game, I didn't like lines in the pawn ending with d5 since normally, Black would then just create an outside passed pawn and win. Since there wasn't any danger in the minor piece ending I just played 46. Bd8 and after 46...Nc4 I claimed that 47. Bg5 would repeat the position for the third time as the position had already occurred on moves 38 and 40 (although when I first told him my claim, I erroneously indicated moves 40 and 43. After he said that the knight was on b7 on move 43 I looked at the scoresheet again and got it right). Incidentally, 46...fxe4+ was the only way to keep the game going as 46...Nb7 was another 3-time repetition by 47. Be7 [1/2:1/2] Things worked out very nicely for me in the other games too as all the 2400+ players beat the U2400 players and the only game between two U2400 players with 2.5 ended in a draw. The only exception was IM Angelo Young who beat GM Blatny with an inspired piece sacrifice. So it was only a two-way tie and we each got a nice pay day.

11/10/02 - Taubenhaus-Pollock, Bradford 1888

After the digression on my games from the Southeast Open, I'm ready to continue the discussion of pawn endings. The date in the title is not a typo, this is a 19th century example. This particular ending happens to be BCE #85. I started looking at it after Richard Toll sent me some analysis to correct that in BCE. While Richard's analysis was correct, I was bothered by an earlier move in Fine's line. I looked at it with Russell Linnemann and we found some very interesting lines.

I'm going to start with Fine's analysis as the main line. 1...d5 For history buffs, the actual game continued 1...c5 2. Ke3? (2. c4 is a blockade) 2... d5? (Black wins with 2... c4 3. Kf2 d5) 3. exd5+? (again 3. c4 was a draw) 3... Kxd5 and Black won after 4. Kd3 c4+ 5. Ke3 a4 6. a3 c6 7. Kf2 Ke4 8. Ke2 c5 2. Ke3 Euwe and Hopper quote analysis by Castadli saying that 2. a4 is the only way to draw (I'd be interested in hearing if anyone has this analysis. I didn't have it in any of my endgame literature. Hopefully I'm adding something to it and not just reproducing it) for example 2...c5 (2... d4 3. Ke2 c5 4. c4=) 3. c4= So the big question is if 2. Ke3 is an alternative way to hold. 2...a4 3. Kd3 Kd6 4. Ke3 Kc5 Euwe and Hopper stop here in their note with the implication that Black wins, but Fine continues on 5. b3 axb3 6. axb3 Kd6 now Fine continues with 7. Ke2 c5 and in the notes to 8. Kd3 gives 8. exd5 Kxd5 9. Kd3 c4+ as decisive based on the variation 10. bxc4+ Kc5 11. Ke4 Kxc4 12. Kxe5 Kxc3 13. Kf5 c5 and queens first. As Richard Toll pointed out the position after 14. Kxg4 c4 15. h4 Kd2 16. h5 c3 17. h6 c2 18. h7 c1Q 19. h8Q Qc4+ 20. Kf5 Qd3+ is drawn. This is all very well, but when I looked at it, the play around move 7 seemed quite odd. After Fine's 7. Ke2 then making a passed pawn with 7...d4 is going to be decisive For example, 8. cxd4 exd4 9. b4 Ke5 10. Kd3 c5 11. bxc5 c6 But can White defend better? 7. Kd3 c5 8. exd5 Kxd5 9. Ke3 c4 10. b4 Kd6 11. Ke4 Ke6 12. b5 Black invades the queenside after 12. Ke3 Kd5 13. Ke2 Kc6 12... Kf6

13. Kd5 The other main line is 13. Ke3 Kf5 14. Kf2 Ke4 15. Ke2 Kd5 16. Ke3 Kd6 (the queen ending looks drawn after 16... Kc5 17. Ke4 Kxb5 18. Kxe5 Ka4 19. Kf5 Kb3 20. Kxg4 Kxc3 21. h4 Kd3 (21...Kd4 22. h5= ) 22. h5 c3 23. h6 c2 24. h7 c1=Q 25. h8=Q) 17. Ke4 Ke6 18. Ke3 Kd5 19. Kf2 Kc5 20. Kg2 (20. Ke3 Kxb5 21. Ke4 Ka4 22. Kf5 e4 23. Kxg4 e3 24. Kf3 Kb3 25. h4 Kxc3 26. h5 Kd3 27. h6 e2 28. h7 e1=Q 29. h8=Q Qe4+ 30. Kf2 Qd4+ -+) 20... Kxb5 21. h4 gxh3+ 22. Kxh3 Ka4 23. g4 Kb3 24. g5 e4 25. g6 e3 26. Kg2 Kxc3 27. g7 (27. Kf3 Kd2 28. g7 e2 29. g8=Q e1=Q 30. Qxc4 Qe3+ 31. Kg4 Qe2+ 32. Qxe2+ Kxe2 33. Kf4 Kd3 -+) 27... e2 28. g8=Q e1=Q -+ This position is won even without the pawn on c7 13... Kf5 14. Kc5 This move, that Linnemann found, looks like the only practical chance. Both 14. Kxc4 and 14. Kc6 fall short, so White waits for Black to block his pawn 14...Ke4 15. Kc6 Kd3 16. Kxc7 e4 17. b6 e3 18. b7 e2 19. b8=Q e1=Q

and Black stands much better, despite being a pawn down. However, there are always technical difficulties in converting queen endings like this.

We haven't answered the question fully if 2. Ke3 can hold or not, but it looks like White could be in some trouble. If any readers think they can shed further light, I'd be interested in hearing from you.

11/7/02 - All Rook Endings Are Drawn

It looks like my analysis of Moreira-Bereolos may have been as sloppy as my play in that game. In my analysis after 60. Rg7

I gave 60... Rc6 61. Rg6 Kb7 62. Kf4 (in the original analysis, this was 62. Kf5, but there is no reason to allow 62...Rc5+) 62...b4 63. Ke4 b3 64. Rxc6? b2! winning. However, White makes an easy draw with 64. Rg1 and White will round up the b-pawn. I thought I found an improvement with 63...Rc5 since 64. Rxh6? allows 64...b3 and he can't even sacrifice his rook to stop the pawn 65. Rg6 b2 66. Rg1 Rc1 so Black is going to pick up the h-pawn and win right? Well, after 64. Kd4 Rxh5 65. Kc4 Rh4+ 66. Kb3 I think White is still holding the draw despite being two pawns down

I don't see any way for Black to make progress here. He can bring his king to the kingside, but then White will shift his rook to the b-file and check from the side. Black can then sacrifice his b-pawn to advance his king, but then White will make a Vancura draw without much problem. I guess my mistake (besides the terrible 40...a4?) was to allow the pawn configuration h5/h6 on the kingside. I still feel the ending after 44...Rf6 is winning for Black, but it will require more analysis to find just where I went wrong.

11/3/02 - Southeastern Open

I visited South Carolina for the first time last weekend to play in the Southeastern Open in Rock Hill. This brings my total up to 21 states that I have played tournaments in. Despite the large prize fund and lack of competing events, the turnout was pretty poor. Alex Wojtkiewicz was the only GM. I thought the turnout from North Carolina was disappointing. The tournament was held about 10 miles from Charlotte, but only 3 of North Carolina's 20+ active masters showed up. Here in Tennessee, we only have about 5 active masters, and 3 of them were in attendance.

The open section was still pretty strong. In the first round I had White against Klaus Pohl on board 2. On board 3, Adam Caveney played Todd Andrews. The last time these pairings took place was in the Tennessee Open. But there, it was round 3, not round 1; and we were on Boards 1 and 2, not 2 and 3. In my game, Pohl got in trouble in the 4 pawns attack after 15. Nd5

He went pawn grabbing with 15... Qxb2? Eliminating the knight with 15... Bxd5 was essential. Petrosian gives this as leading to a Black advantage 16. Rb1 Nxf3+ 17. Bxf3 Qd4+ 18. Qxd4 This is much better than letting Black swap on d1 since the White rooks are ideally placed on the b and f files. 18...Bxd4+ 19. Kh1 Nd7 It's too late to get rid of the knight now since after 19...Bxd5 20. Bxd5 both b7 and f7 are attacked. 20. Nc7 Bxa2 Maybe 20...Nb6 offered faint hopes of resistance keeping the rook out of the attack. 21. Rxb7 I haven't been able to find a defense from this point. White just has too much pressure. The d-pawn is ready to become a monster 21...Nf6 22. Nxe8 Rxe8 23. Bc6 Bc4 23... Re6 24. Bh6 exposes another problem, the back rank. 24. Bxe8 Bxf1 25. Bxf7+ Kf8 25... Kh8 26. Bh6 would also soon be mate 26. Bh6# [1:0]

In round 2, I had Black against Jairo Moreira. This game changed the face of the tournament greatly. I outplayed him in the middlegame to win a pawn, then made some "safe" moves in time pressure that gave him some drawing chances. The rook ending then turned into a comedy of errors in which I made the last one to give him a draw. He went on to win his next two games and then was given a draw in the final round to give him clear second. There were better opportunities for me earlier, but I'm going to pick up the analysis in the rook ending after 32. RfxNc1

32... Rxa2 33. Rc7 Rd8 34. R1c3 Rd5 This was the only chance I saw to keep the game going. 34... Rdd2 35. Rf3 f5 36. g4 and it will be White who gets his rooks on the 7th with at least a draw. 35. Rf3?! Black doesn't care about the doubled pawns. Much better is 35. g4 keeping the Black rook off of f5. 35... Rf5 36. Rxf5 gxf5 37. Kf3? Now White should lose as Black can force connected passed pawns. Better was 37. Rb7 When White would still have drawing chances. 37... Ra3 38. Kf4 38. Rc3 a5 39. Re3 b5 38...a5 39. Rc3 Kg7 40. Kxf5 a4? A totally careless move 40... b5 was correct. This is pretty unexcusable since I had seen the early variation with b5. 41. Rc4 Rxb3 Now 41...b5 is met by 42. Rg4+ 42. Rxa4 Rf3+ 43. Kg5?! Better is 43. Ke5 Rxf2 44. Rg4+ Kf8 45. Rh4 Kg7 46. Rg4+ Kh8 47. Rb4 with a probable draw. 43... Rxf2 44. Rb4 Rf6 45. h4 Re6 46. g4 Kf8 47. h5 Kg7 48. Kh4 Rc6 49. g5 f6 50. gxf6+ Kxf6 51. Kg4 Ke5 52. Rb1 Rd6 53. Kg5 h6+ 54. Kg4 Rf6 55. Re1+ Kd5 56. Rd1+ Kc6 57. Rc1+ Kd7 58. Ra1? I was just going to make sure that he avoided 58. Rd1+ Rd6 59. Rxd6+? Kxd6 60. Kf5 b5 61. Kg6 b4 62. Kxh6 b3 63. Kg7 b2 64. h6 b1=Q 65. h7 and Black wins despite the rook pawn on the seventh since his king is close enough. I was going to repeat moves with 59. Rb1 Rf6 60. Rd1+ and then continue with my plan of 60...Kc7 61. Rc1+ Kb8 62. Rb1 Kb7 which I thought was zugzwang, but now I'm not so sure. He may still be holding after 63. Rb2 Kc6 64. Rc2+ Kd5 65. Rd2+ Kc4 66. Rc2+ Kd3 67. Rb2 Kc3 68. Rb1. After the game continuation Black should win. 58... b5 59. Ra7+ Kc8 60. Rg7 Rb6? Rooks belong behind passed pawns, but here it is a mistake. 60... Rd6? was also wrong since the White king can then gain a tempo 61. Rg6 Kc7 62. Kf4 b4 63. Ke5 Rd7 64. Rxh6 b3 65. Rg6 b2 66. Rg1= The right way is 60... Rc6! 61. Rg6 Kb7 62. Kf5 b4 63. Ke4 b3 64. Rxc6 b2! (64... Kxc6 65. Kd3=) 65. Rxh6 b1=Q+ and wins 61. Rg6 Kb7 62. Kf4 Rc6 62..b4 63. Ke4 Rb5 64. Rxh6 b3 65. Rg6 b2 66. Rg1 b1Q 67. Rxb1 Rxb1 68. Kf5 is also a draw since the Black King is too far away. 63. Ke3 b4 64. Rxc6 Kxc6 65. Kd3 Kd5 66. Kc2 Ke4 67. Kb3 Kf4 68. Kxb4 Kg4 69. Kc3 Kxh5 70. Kd3 Kg4 71. Ke2 Kg3 72. Kf1 Kh2 73. Kf2 h5 74. Kf1 h4 75. Kf2 h3 76. Kf1 Kh1 77. Kf2 h2 78. Kf1 [½:½]

In the third round I had White against Mark Hathaway. My previous two games probably influenced my play in this one. I played an e5 sacrifice in the 4 pawns attack that was probably not totally sound. I knew there were important differences between the position from Round 1, but I had won so easily that round and was still a bit frustrated from blowing Round 2. I may have been close to having compensation at several points, but he seemed content with a draw. I took his offer after 27...b6

since regaining the pawn with 28. Qc4+ Be6 29. Qxa6 looked dangerous for me after 29...Qh4. I should perhaps have continued with 29. Qb3+, which is still about equal after 28...Be6 20. Qxb6 Qh4 21. Qc6, but sets a trap of 28...Qe6?? 29. Re1 and White wins.

I had Black against Angel Mera in round 5. I thought I got a nice position out of the opening, but made some inaccuracies before the time control and had to bail out into an inferior ending. I found a nice drawing combination at the end after 50. Kxa3

50...Nxb4 The most accurate move order 50... h5? 51. gxh5 Nxb4 52. Bg6+ and White wins 51. Kxb4 h5! Again, accuracy is required. The immediate 51... Be7? fails to 52. Bxe7! (not 52. Bg6+? Kxg6 53. Bxe7 h5 54. g5 h4=) 52... Kxe7 53. Bg6 +- 52. gxh5 Be7 53. Bg6+ Kf6 54. Bxe7+ Kxe7 55. h6 Kf8! The final precise move. Black could still go wrong here with 55...Kf6? 56. Bh5 +- 56. Bh7 Kf7 57. Kc5 Kf6 58. Bf5 Kf7 [½:½]It's humorous to give the diagrammed position to your chess computer. It won't even consider 50...Nxb4 and even with 5 piece tablebases, it will say White is winning after you give it the the first two moves. The tablebases will let it see that 52. gxh5 is a draw, but it thinks it is just up a piece if it doesn't take. Yet, somehow Kramnik could only draw with the beast...

In the final round, I had White against Armin Burghart and got to add another chapter to the saga of the Botvinnik Variation. We followed the main line to 15...b4

16. Na4 This is the older move. These days the sharper 16. Rb1 is more often played 16...Qa6 My opponent is playing on the cutting edge, though. 16...Qb5 is the older continuation here, but I think these days 16...Qa6 is considered more reliable 17. a3 b3 Shirov says this is not a serious idea, although it has scored better in my database than the principle move 17...Bd5 which carries the idea 18. Bxd5 Ne5 with unclear play. 18. Nc3 Bh6 This looks like the first new move. I found earlier games with 18...Nb6 and 18...Ne5. 19. Bxh6 Rxh6 20. Qd2 20. dxe6 as on almost every move in this variation deserves attention. 20...Rdh8 Perhaps, 20...Rxf6 should be considered. 21. h4 Ne5 22. Rad1 a case of which rook to play to d1. In the game continuation White has to be careful that if he takes everything on d3 that Ba6 doesn't skewer his rooks. However, keeping the rook on f1 allows White to cover up with f3 if necessary. 22...Nd3 23. Be4 exd5 24. Nxd5 24. Bxd3 cxd3 25. Qxd3 Qxd3 26. Rxd3 Ba6 27. Rxd5 Bxf1 28. Rxc5+ Kd7 29. Kxf1 also promises an advantage to White 24... Kb8 Black is probably lost after this move since White is able to exchange off all of the minor pieces. Black's king is too exposed in the resulting major piece ending. 24... Qe6? was also bad because of 25. Bf5 so 24... Bxd5 probably had to be tried although it seems that White still has a fairly sizeable plus after 25. Bxd5 25. Bxd3 Bxd5 26. Qf4+ Kb7 27. Be4 Bxe4 28. Qxe4+ Kb8 on 28... Qc6 it is probably simplest to leave the queens on with 29. Qxc4 when the Black king is still bare. 29. Qe5+ Ka8 30. Rd6 Qc8 31. Rfd1 Rg8 32. Qe4+ Kb8 33. Qf4 [1:0]


Yet another counter