Shakmaty Bereolos - The Official Chess Site of Peter Bereolos

7/19/20 - Bereolos-Kudrin, 1997 Chicago Open

I remember being quite frustrated by my loss against Sergey Kudrin in the 1997 Chicago Open. I knew I should be losing such an equal position, even against a strong player. My major mistake was making exchanges that improved his position. As the tide turned, I didn't put up maximum resistance either. However, the story ends on a good note as a scored 2 wins and a draw in the last 3 rounds to tie for 10th and collect what was my largest prize ever.

7/15/20 - BCE-94, Brinckmann-Rubinstein, Budapest 1929

This week's BCE correction is one of Rubinstein's classic pawn endings. I'm not sure why Fine tried to improve on Rubinstein's clear and logical play in this one.

The 1929 tournament in Budapest was fairly strong, although it was missing Alekhine and Bogoljubow who were playing their world championship match. Capablanca won convincingly with an undefeated 10.5/13 over second place Rubinstein a full point behind. Rubinstein's opponent in this game, Alfred Brinckmann finished near the bottom with 4 points.

In the game of interest, Brinckmann had an extra pawn after 59...Kxe6, but he is dead lost.

Black has all the advantages: there are a bunch of White pawns stuck on the same colored square as the bishops, the Black king is more active, Black has more space with a potential breakthrough on the queenside with b4 and d4, and the extra White pawn is a worthless doubled pawn that Black soon recovers. Rubinstein brought the point home smoothly. 60.Bd4 g6 61.Be3 Kf5 62.Kc1 Bf4 63.Kd2 Kxg5 64.Ke2 Kg4 65.Bxf4 Kxf4! 66.Kf2 The starting position of BCE-94 66...Ke4 66...Kg4 is the subject of the BCE correction. 67.Ke2 g5 68.Kf2 d4 69.cxd4 Kxd4! 70.Ke2 b4! 71.Kd2 b3! 72.c3+ Ke4! 73.Ke2 Kf4! 74.Kf2 Kg4 75.Kf1 Kg3 76.Kg1 g4 77.Kf1 Kh2! 78.Kf2 Kh1! 79.Kg3 Kg1 80.Kxg4 Kxg2! 0-1 White gets to the queenside first, but Black's advanced b-pawn carries the day 81.Kf4 Kf2 82.Ke4 Ke2 83.Kd4 Kd2 84.Kxc4 Kc2 85.Kb4 Kxb2 86.c4 Kc2 87.c5 b2 88.c6 b1Q+

7/8/20 - BCE-229a

This week's BCE position is a continuation of last week's. Fine moves the pawn back a couple of squares and claims the position is still a draw. Hey's analysis in Deutsche Schachzeitung was followed up by some discussion that I assume came from Schlecter, as he was the editor. It was noted that in the ending of K+N+h vs. K, pushing the pawn to h7 too soon is a mistake and that was likely the cause of White not winning. The following positions and analysis are then given, which lay out the winning method for BCE-229a. First, Hey's position with the pawn on h6 instead of h7.

1.Ne6 Bh4 2.Nc5 Bd8 3.Nb7 Bc7 4.h7 when it is mate in 2. Next, it is shown that having the king on g8 makes no difference.

1.Nf5 Bf4 2.Ne7+ Kh8 3.Nc6 Bc7 4.h7 again with mate in two.

A more modern take on this position, which adds more preliminary play is the following study by Andrei Zhuraviev in 1995, which took a special first prize in the Jan van Reek 50th jubilee composing competition.

I'll just give the main line. You can refer to Issue 121 of EG for analysis of alternatives. 1.Nf3! Ke6 2.Kg5! Kf7 3.h6! Bd6 4.a4! a5 5.Kf5! Kg8 6.Kg6 Bf4 7.Nd4! Bd6 8.Nc6 Bc7 9.Ne7+ Kh8 10.Nf5 Be5 11.Ng7 Bd6 12.Ne6 Be7 13.Nd4 Bd6 14.Nf3 Bf4 15.h7 1-0

7/4/20 - TCEC Season 17 Superfinal, Games 15-16, Modern Tiger

Games 15 and 16 of the Season 17 Superfinal featured a variation of Modern Defense that is a favorite of Swedish GM Tiger Hillarp-Persson, who has written two books on it: Tiger's Modern and The Modern Tiger. Game 16 was the more important game theoretically and from the match perspective. Stockfish scored the first of only two Black wins in the match, although that result had nothing to do with the opening. 1.e4 g6 2.d4 d6 3.Nc3 a6!? The starting position for this set of games. Tiger usually completes the fianchetto before playing this move. It ends up transposing after the next move.

4.Be3 Stockfish played the more aggressive 4.f4 in game 15. 4...Bg7 5.h4 A typical scenario in many fianchetto openings. Black has to decide if he wants to ignore the advance, play ...h6 with the intention of meeting h5 with g5, or Stockfish's choice to immediately stop further advance of the h-pawn 5...h5 This is the most popular move, but White has scored the best against it. It was Tiger's first choice in his first encounter against h4 back in 2003, but he now considers it dubious. His preference is to ignore the pawn with 5...Nf6, but he also played 5...h6 in one game 6.Nh3 Tiger gives White the edge after 6.Nf3 Bg4 7.Qd2, but Black could also play 6...Nf6 or 6...b5 which are typical in the variations without the advance of the h-pawns 6...Bxh3!? An interesting concept. Black surrenders the bishop pair, but now the White king can only castle queenside where Black will have more space. It also takes away an attacking piece from White's kingside attack. I thought it might be a unique concept by Stockfish, but Svidler had played this move in a French League game against Meier in 2009 7.Rxh3 Nd7 8.Qe2

A novelty from Lc0, but I don't really understand the point. 8.Qd2 as played by Meier seems much more natural, not blocking Bf1 although Svidler had few problems after 8...c5 9.0-0-0 cxd4 10.Bxd4 Bxd4 11.Qxd4 Ngf6 and a draw was agreed.

The remainder of the game was a weirdly mirror to game 12. This time it was Lc0 that self destructed. The position was fairly stable after 42...a4

Black's weak e-pawn is easy to defend and offset by White's f-pawn. Stockfish was already sitting on a 0.00 evaluation, but Lc0 gave itself a very small winning percentage. The machines began one of their maneuvering sequences with lots of Kg7-h7 and Ka2-Ka1 as well as back and forth moves by the rooks and queens. The only progress Lc0 made was to advance its f-pawn by one square and after that there were points where Stockfish offered the e-pawn for the f-pawn including after 81...Re8

As at the other times where White could take the e-pawn, Lc0 declined to do so here.instead opting for 82.f5?! 82.Rxe6 Qxe6 83.Qxe6 Rxe6 84.Rxe6 Rf7 is an equal ending. I think this situation is a bit different than what happened in Game 12. There, Stockfish had a large negative evaluation of its position, and giving up a pawn didn't change the evaluation much. Here, Lc0 gives itself a very small advantage, which is reduced by the pawn sacrifice, but still positive, whereas allowing the 50 move rule would reduce the win probability to zero. 82...gxf5 83.Rd1 Black should still have a small edge after 83.Rxe6 Rxe6 84.Qxe6 Qxe6 85.Rxe6 Kg7, but this should also be drawn. The text move kicked off another long sequence of shuffling. Lc0 steadfastly refused to regain the e-pawn, but Stockfish did not show any progress and it looked like the game would be drawn by the 50 move rule after 129...Rfa8

Three more random moves and the game would be drawn, but Lc0 decided to give another pawn with 130.g6? which Stockfish happily took 130...Qxg6, although even here, Stockfish did not come off of its evaluation of 0.00 until a few moves later. 131.R5e2 Kh8 132.Qd7 R8a7 133.Qb5 Qg8 134.Rh2? Even here White could take the e-pawn 134.Rxe6 Rxe6 135.Rxe6 Qxe6 136.Qb8+ Kg7 137.Qxa7+ and the Black king is too exposed for Black to win 134...Ra8 Now, Stockfish finally registered an advantage to Black. Strangely, Lc0 had predicted this move when playing 134.Rh2, but still gave itself the better chances, but now it also swung the advantage in Black's favor. After urther adventures Stockfish began announcing mate on move 172 and brought home the point after 196 moves.

7/2/20 - McEntee-Bereolos, 2008 Billy Colias Memorial

Another example of trying to use a dark square grip in the Benoni as compensation for a pawn is my game against Tim McEntee in the 2008 Colias Memorial. After 46.Kxe2

46...Ke5 The attempt to activate the rook doesn't work 46...Rc5 47.Rb3 Rxa5 48.Rxb7 Ra2+ 49.Ke3 Rxg2 50.Rd7 47.Ke3 g5 48.Rb3 f5

This move weakens the sixth rank,but it is already weak enough for White's purposes if Black waits passively, although he has to find a few good moves to break down the defense 48...Rd7 49.c4 Rc7 50.Kd3 Rd7 51.c5 dxc5 52.Kc4 Rc7 53.Rb6 Now, Black has no play as the king raid is too slow Kf4 54.Rh6 Kg3 55.Rh7 Kxg2 56.e5 Kxh3 57.e6 Kg3 58.d6 49.exf5? After this Black has no problems 49.Kd3! seems to give White excellent winning chances. Even working with the engines, I've been unable to find a good defense. The Black king is surprisingly vulnerable on e5. It is a quite logical move, White intends simply c4-c5, but I think sometimes the opposition is so deeply ingrained that players are hesitant to give it up. Some attempts

a) 49...f4 does nothing to stop White's plan and even cuts off an escape route for Black's king. White wins 50.c4 Rg7 51.c5 dxc5 52.Rb6 c4+ 53.Kxc4 Rc7+ 54.Kd3 Re7 55.Rg6;

b) 49...fxe4+ doesn't fare much better 50.fxe4 and here in addition to c4-c5 White also has the idea Rb2-f2-f5# 50...g4 51.Rb2 gxh3 (51...Rg7 52.Rf2; 51...g3 52.c4; 51...Rf7) 52.Rf2;

That seems to leave c) 49...g4 as probably best, but then 50.exf5 is a much better version than the game since after 50...gxf3 51.gxf3 Kxd5 (on 51...Rf7 the engine comes up with the hard to find 52.f4+!? when taking either f-pawn allows the White king to d4 to support c4-c5 and 52...Kxd5 transposes to the main line) White has 52.f4 stopping the king's return to e5 52...Rf7 53.c4+ Kc5 54.Rb6 Rxf5 55.Ke4 Rf7 56.f5 Kxc4 57.Rxd6 and White is winning;

49...Kxd5 50.f4 Rf7 51.fxg5 Rxf5 52.Rxb7 52.g6 Rg5 52...Rxg5 53.Kf2 Rf5+ 54.Ke2 Rg5 55.Kf2 Rf5+

56.Kg1?! Attempting to keep the game going because he is up a pawn. Tim's fighting spirit is commendable, but there is no real justification since the Black king is so active. White should just acquiese to a repetition with 56.Ke3 Rg5 56...Kc4 57.Rb4+ Kxc3 58.Rxh4 d5 58...Rxa5 lets the White king participate in the defense 59.Rh6 d5 60.Rc6+ Kb2 61.Kf2 d4 62.Ke2 59.Rh6 d4 60.Rxa6 d3 61.Rd6?! It's simpler if White keeps the a-pawn 61.Rc6+ Kb3 62.a6 d2 (62...Rd5? 63.a7!) 63.Rd6 Kc2 61...Rxa5 another try is 61...d2 62.a6 Ra5 63.Kf2 Rxa6 64.Rxa6 d1Q but White sets up a fortress with 65.Ra3+ Kc2 (65...Kd2 66.Ra2+!) 66.Rf3 although Black may have some practical chances here as his king has crossed over the third rank 62.Kf2 d2

63.g4? Stopping a check on f5, but White should force the Black king in front of the d-pawn before advancing his own pawns 63.Rc6+ Kd3 64.Rd6+! Kc2 65.Rc6+ Kd1 66.g4 and the game should end in a draw 63...Ra1! 64.Rc6+ Kb4 65.Rd6 d1Q 66.Rxd1 Rxd1 67.h4 Kc5 68.Kf3 Rd4 69.h5 Kd6 70.h6 Ke6 71.h7 Rd8 0-1

7/1/20 - BCE-229, Hey, 1913

BCE-225 is another position presented as a study that was based on an actual game. Oskar Hey describes it in correspondence to Deutsche Schachzeitung as coming from a consultation game at the Neuberger Schachklub between Kahl and Bpahler against Fick and Hey. In this position, the bishop needs to be able to defend both the a-pawn and prevent the knight from delivering mate on f7. Hey treats it as a corresponding squares problem giving the squares the bishop must occupy for each positioning of the knight. The additional subtlty, that was missed by both Hey and Fine is that White has a waiting move, Kh6. However, unlike last week's rook vs. bishop duel, here the bishop generally has more than one square it can be on for the defense, which allows Black to hold. They key is that either the squares must be connected or Black has to be able to give check when the White king goes to h6. That latter point is why the BCE line fails. Hey gives 3 squares the bishop can be on when the White knight is on c5: d6, g3, and d8. From the first two squares, Black can meet Kh6 with Bf4+, but from d8 the only check is on g5, which drops the bishop. Since Black cannot reach the other two defensive squares from d8, he is in zugzwang.

The only other position Hey got wrong was with the knight on d2, where he says the Black bishop should be on e3 or e7, both of these get mated fairly quickly 1.Ne4 is #3 vs. Be7 and 1.Nc4 is #6 vs. Be3. From d2, the forward knight moves are to b3, c4, e4, and f3. The common defensive square against Nb3 or Nc4 is Bc7 and the common defensive square against Ne4 or Nf3 is Bf4. That indicates against Nd2, the bishop can be on any square along the b8-h2 diagonal except c7 or f4.