Shakmaty Bereolos - The Official Chess Site of Peter Bereolos

8/30/20 - Bereolos-Sadorra, 2016 Kings Island Open

I'm not especially satisfied with my effort against Julio Sadorra at the 2016 Kings Island Open. While my attack should have yielded a draw, I think I played several second best moves in the middle game, which made the ultimate attack less strong. Compounding this, after the game turned, I did not put up much resistance. The move 34.Bc1?! is especially sour since I should have realized that avoiding a short term material loss was not worth the long term consequence of locking my rook out of the game.

8/26/20 - BCE-91c, Loyd-Winawer, Paris 1867

Today's BCE position is from the international tournament held in conjunction with the 1867 Paris World's Fair. The victory by Ignatz von Kolisch (+20 -2 =2) ahead of Symon Winawer and Wilhelm Steinitz propelled him to the top of Jeff Sonas' historical rating list. This isn't too surprising as there was not a lot of international competition back then and Sonas' entire list at that point only has 21 players, 13 of which played in Paris.

The only American player in the field was the noted puzzle creator Sam Loyd, who finished 10th. In today's game Winawer was trying to convert a pawn up rook ending with Black against Loyd. After 36.Ra2 he gave back the pawn to reach a "winning" pawn ending.

36...b5? The endgame concept of "don't rush" had not been introduced when this game was played. 37.axb5 cxb5 38.Rxa5 bxc4 The pawn ending should be drawn, Black could try to keep the game going with 38...g5+!? 39.hxg5 hxg5+ 40.Ke3 bxc4 since 41.Rxc5? dxc5 42.bxc4 Ke5 is a simple win for Black. White has to find 41.b4! when it seems that he has enough to hold after 41...Rc7 42.Ra2 (not 42.Rxg5? c3) 39.Rxc5! dxc5 40.bxc4! the starting position of BCE-91c 40...g5+ The tournament book agreed with Fine's assessment that Black was winning 41.hxg5! hxg5+! 42.Kxg5! 42.Kf3? Ke5 43.Ke3 f6 puts White in zugzwang 42...Ke5! 43.Kh6 Kxe4 44.Kg7? 44.Kg5 is the subject of the BCE correction, 44.Kh5 also holds the draw, but not 44.g5? Kf4 45.Kg7 (45.Kh5 Kg3 46.g6 fxg6+ 47.Kxg6 Kf4) 45...Kxg5 46.Kxf7 Kf5-+ as given by Fine44...f5! 0-1 The tournament book states that each side spent an hour which is somewhat shocking given that the time control was a leiesurely 10 moves per hour.

From the database, a little bit of French trivia. The players in the Paris tournament did not give a nod to their hosts, playing the French Defense (1.e4 e6) in only 4 games versus 100 games with 1.e4 e5, including today's subject game. The first time Winawer tried the Winawer variation (1.e4 e6 2.d4 d5 3.Nc3 Bb4) was a few rounds later, unsucessfully, against Steinitz. The winner of tournament, Kolisch, appears to have been the first to play the Winawer variation with Black in his match against Paulsen in 1861.

8/24/20 - TCEC Season 17 Superfinal, Games 19-20, Qb6 Sicilian

Games 19 and 20 of the Superfinal debated what is considered a sideline of the open Sicilian. 1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Qb6 This and the related system with 2...e6 aren't the most popular variation of the Sicilian, but there is still a considerable amount of theory with close to 10000 games in the database. There is an older book by Ilic, and a more recent book by Grivas, who is one of the main practioneers. Even the Grivas book is 15 years old, but theory does not seem to have developed rapidly in this line. 5.Nb3 Nf6 6.Nc3 e6 7.Qe2

This was the starting position for the two game set. The engines went their separate ways here. 7...Bb4 Lc0 went for the move considered more solid 7...Be7 but then went for activity with a pawn sacrifice 8.Be3 Qc7 9.0-0-0 d5!? instead of steering towards the Hedgehog with 9...a6 This pawn sac is known in similar positions, but was a novelty here. 8.Bd2 0-0 9.a3 9.0-0-0 was the very latest from elite level human chess is the recent online rapid game between Carlsen and Gelfand in the Legends of Chess tournament. In the post-match interview, Magnus said he wasn't too sure about the theory as this is a line you never face and he thought that maybe 9.a3 was the main theoretical move. Still, it didn't work out too bad for him as he won in only 29 moves. 9...d5 10.e5 Nd7 11.f4 a6 (Magnus initial impression was that White was already after this and suggested 11...f6) 12.Qh5 Rd8 13.Bd3 and to me it feels like Black got a bad version of the Classical French. 9...Bxc3 9...Be7 is more popular here, but the text is more principled. 10.Bxc3 e5 11.0-0-0

11...Re8 Black would like to play 11...Rd8 intending ...d5, but White has the strong 12.Rd6 threatening Ba5 12...Qc7 13.Rxf6 gxf6 14.Qg4+ Kh8 15.Qh4 with a powerful attack in Adams-Knezevic France 1997. Ilic's suggestion of 11...Ne8 intending ...d6 and ...Be6 hasn't found many takers. Instead, human practice has turned toward; 11...a5 which seems much more useful than ...Re8 and has scored quite well for Black (63.6%) 12.f3 The assessment of 12.g4 d5 is likely critical, but Black seems to be holding his own here. Grivas gives 12.Qe1 keeping an eye on a5 as an alternative. Perhaps 12.h3 preparing 12.g4 is worth further investigation, keeping open the path for the White queen to the kingside. I only found one example, which ended in a draw after a long struggle starting with 12...a5 12...Rd8 13.Qb5 Now 13.Rd6 Qc7 14.Rxf6 gxf6 has lost its sting since the White queen can't quickly join the attack, but it seems like a kind of position where a neural net engine like Lc0 would still be able to prove long-term compensation 13...Qc7 14.Na5 a6 15.Qc4 d6 16.Nxc6

The first new move. 16.Bb4 was the previous try in Van Haastert-Safarli: European Club Cup 2018. Safarli is another GM who has played the Qb6 variation fairly regularly, so his games in this line are also worth study. 16...bxc6 White has the two bishops, but neither of them is doing much. Black has covered the weakness on d5, and is ready to complete his development with ...Be6, so the position is likely dynamically equal, but there is still play left. There was one further position a couple of moves later that I wanted to comment. 17.Qe2 17...Re8 18.Re1 a5 19.Qe3 Be6 20.f4 Nd7!?

A move which illustrates the difficulties of working with engines. I fed this position to fairly recent versions of Stockfish and Lc0 and they had the obvious looking 21.f5 as their 6th and 7th preference respectively, with only very slight advantage to White. I think this would be a very uncomfotable position to play as Black with the rediculous bishop on a2. Apparently, White can't immediately win it after 21...Ba2 22.b3 a4 23.Kb2 axb3 24.cxb3 Bxb3 25.Kxb3 Reb8+ 26.Kc2 (26.Ka2 Rxa3+ 27.Kxa3 Qb7) 26...d5 with a very strong attack for the piece. But even knowing that the engines evaluate the position as equal, over the board you will be constantly worried about the bishop getting trapped. Lc0 instead played 21.fxe5 and the game was eventually drawn after 153 moves, about 100 of which were the type of shuffling of pieces while avoiding 3-fold repetition that seem to be a characteristic of games between the top engines.

So, from the engine games, it looks like the Qb6 variation is doing fine and might give Black players an alternative to the steady diet of Najdorfs and Sheveshnikovs. Still, it seems that Black has to do some serious work before trying it as when things go wrong, they can end up in disasters as the games of Adams and Carlsen showed.

8/19/20 - BCE-294, Berger, 1888

BCE-294 is another endgame study by Berger that appears in his endgame book with colors reversed. That was likely Fine's source, Berger cites at as being from Songtagsblatte f. J. a. d. V. 1888. I'm not clear on what all those abbreviations stand for, from what I could find Songtagsblatte was a German newspaper.

As a study, this position is not great. There are a few pitfalls that White can fall into, but the fact that either of the obvious initial moves of pushing a pawn are winning takes away any athetic value. Fine seemed to trust Berger's analysis and did not recognize postions that transposed, evaluating the same position as both a win and a draw. Walter Korn analyzed the position in the December 20, 1955 issue of Chess Life under the title It's Confusion, Not Logic That Reigns, which seems appropriate. I think Benko did well to remove this example from the revised edition.

8/15/20 - Bereolos-Baugh, 1990 Pillsbury Memorial

Last month, the Hey position illustrated the fortress that a defender can put up against a knight and a rooks pawn when the rooks pawn advances to the seventh. I had previously shown this idea from my 1993 game against Boris Men. I had also utilized that knowledge a couple of years earlierat to rescue a half point from a lost ending against Christopher Baugh. After 33...Kxf4

White is dead lost. Black has a simple plan to use the h-pawn to distract one White piece on the kingside, then win on the queenside. 34.Kc1 h5 35.Kd2 Nf3+ 36.Ke2 h4 37.Kf2 h3 38.Ng3 h2?! There is no need for this advance. I now have some hope as it will be a draw if I can elimniate all the queenside pawns. 39.Kg2 a5 40.Ne2+ Ke3 41.Nc3 Kd3 42.Na4 b5? This is the move that costs the half point. There was still a win but it was considerably narrower than it needed to be. 42...Kc2 43.Nc5 b6 44.Nd7 b5 45.cxb5 cxb5 46.Nc5 Kb2 47.Nb7 a4 48.bxa4 bxa4 49.a3 Ne1+! 50.Kxh2 Nd3!

and the White knight can't get on a track to stop the pawn. 43.cxb5 cxb5 44.Nc5+ Kc2 45.Nb7 a4 46.bxa4! bxa4 47.a3 Kb3 It's too late to win with 47...Ne1+!? but it was a better practical chance as White still has to find a few more accurate moves 48.Kxh2 Nd3 49.Nd6 Kb3 (49...Ne5 50.Ne4 Nd7 51.Nd6 Kb3 52.Ne4!) 50.Nc8 Kxa3 51.Nb6 48.Nc5+! Kxa3 49.Nxa4 I offered a draw here, as a way to tell him that I knew it was a draw in case he didn't. He only punished this inappropriate offer by playing a couple of more moves. 49...Kxa4 50.Kh1! Kb4 51.Kg2 1/2-1/2

8/12/20 - BCE-213, Nimzowitsch-Tarrasch, Bad Kissingen, 1928

This week's BCE position is from a game between the two great rivals, Aaron Nimzowitsch and Siegbert Tarrasch, at the 1928 Bad Kissingen tournament. I've already talked some about that tournament, won by Bogoljubow. Nimzowitsch finished 5th on +1, while Tarrasch was next to last with -3. Nimzowitsch reached a pawn up ending, but with opposite colored bishops after 38.Bxe7

38...Kb7 Black could even start with 38...Bb5 39.Kh2 The starting position of BCE-213 39...c4? 39...Bb5 is the subject of the BCE correction intending 4 0.Kg3 Bf1 putting the pawns in the crosshairs. IM Gyula Meszaros gives another way to draw 39...f4 40.Bg5 e3 41.fxe3 fxe3 42.Bxe3 Bg6 43.c3 Bb1 44.a3 c4 45.g4 b5 46.Kg3 Kc6 47.Kf4 Kd5 48.Kg5 Ke6 49.Kh6 Kf7 50.h4 Bd3 51.h5 Bc2 52.g5 Bd3 53.b3 cxb3 54.Bc1 Be4 55.a4 (55.c4 bxc4 56.Bb2 Ke6 57.g6 hxg6 58.hxg6 Bxg6 59.Kxg6 c3 60.Bxc3 b2 61.Bxb2 Kd5 62.Kf7 Kc6) 55...bxa4 56.c4 Ke6 57.g6 hxg6 58.hxg6! Bxg6 59.Kxg6 Ke5 60.Bb2+ Kd6 61.Ba3+ Ke5 40.Kg3 Kc8 41.Kf4 Kd7 42.Bb4 Ke6 43.Bc3 Bd7 44.g3 b5 45.Kg5 Kf7 46.h4 Bc8 47.Kh6 Kg8 48.b3 cxb3 49.cxb3 f4 50.gxf4 Bd7 51.Kg5 Kf7 52.f5 Bc6 53.Kf4 Ke7 54.Ke5 Be8 55.Kxe4 Bc6+ 56.Ke5 Be8 57.Kd5 Bf7+ 58.Kc5 Be8 59.Be5 Bd7 60.Kb6 Kf7 61.f6 Be8 62.f4 Ke6 63.Ka6 Kf7 64.b4 Ke6 65.a4 bxa4 66.b5 1-0

8/5/20 - BCE-298, Rinck 1914

I'm back with more BCE corrections after a couple of weeks off. This week's BCE position is a study by Henri Rinck that appeared in Deutsche Schachzeitung in 1914. As I was unable to find any scans of that year's issues, I don't know if the error was in the original or added by Fine. To make up for a couple of weeks with no corrections, I'll give a bonus correction for a position from one of Benko's games that he added to this section in the revised edition of BCE.

The 1958 Interzonal tournament in Portoroz, Yugoslavia (in present day Slovenia) is best know for the fact that a teenage Bobby Fischer qualified for the Candidates Tournament. However, Benko beat Fischer in their head-to-head encounter and finished ahead of him to also qualify. Benko had a nice endgame save as Black against Hector Rosetto in round 2. After 47...Kd7

48.Rd4+? Handing Black a critical tempo. White should immediately go after the g-pawn 48.Rg4 Nh4 49.Rxh4 Rh1 50.Rg4 Rxh2 51.Kxb5 and wins 48...Ke6! 49.Rg4 Nh4! 50.Rxh4 Rh1! 51.Rg4 Rxh2! 52.Kxb5 The same position as the previous note, but the Black king is one square closer, which is enough to save the day 52...Kf5! 53.Rg8 Kf4! threatening to build a bridge with Rh5-g5, so White can't get his pawns moving yet. 54.Ka4 54.Kc4 Kf3 renews the threat to build the bridge. There is nowhere to hide by moving forward; 54.Kc6 Rh6+ 55.Kc7 Rh7+ 56.Kc8 Rh8 which should be a draw after 57.Rxh8 g1Q 54...Kf3! 55.c4 Rh1 56.Kb5 g1Q! 57.Rxg1 Rxg1!

This is the starting point of Position 609 in the revised edition of BCE. 58.c5 Benko also analyzes 58.a4 Rg8 59.a5 Rb8+? (Black needs to bring the king closer with 59...Ke4)60.Kc5! Rc8+ and White wins with 61.Kd5 (Benko's 61.Kd6? should be a draw after 61...Rxc4! 62.b5 Ra4! 63.a6 Ke4 (Instead of Benko's 63...Ra5? 64.Kc6 Ke4 which he abandons as a draw, but the winning method is well known 65.Kb6! Ra1 66.Ka7! and the b-pawn queens.) ) 58...Ke4 59.a4 Kd5 60.a5 Rg8 61.a6 Rb8+ 62.Ka5! Kc6 63.a7! Rg8 64.Ka6 Rh8 65.b5+! Kxc5 1/2-1/2

8/2/20 - TCEC Season 17 Superfinal, Games 17-18, Budapest Defense

At first I was going to pass over games 17 and 18, which seemed to be a weird sideline in the Budapest. Then, I read GM Sandler's raves about Lc0's play in the opening of this game and then noticed that both engines entered this sideline of their own accord and decided it merited further investigation.

1.d4 Nf6 2.c4 e5 3.dxe5 Ng4 This was the starting position for this set of games 4.e3 I'm not aware that White is having any issues in the 4.Bf4 and 4.Nf3 lines, but both engines went for this modest pawn move 4...Nxe5 5.f4!?

I had once played this move, but in conjunction with 4.e4. In this particular position, I only knew the idea 5.Nh3 heading to f4 with more control over d5. The engines' idea seems to be that you can grab space, but not open the g1-a7 diagonal where a Black bishop can land. The setup is somewhat reminiscent of the Mujannah tabiya from the chess precursor game Shatranj.

The missing d-pawn doesn't seem to hurt White since the Black knights are far from the e4 square. There aren't a lot of games in the database with 5.f4, but White has scored a very healthy 72%. I tried it in an online blitz game and achieved a fairly easy win. White's structure gives him a nice space advantage and he still can pressure the d5 square. 5...Nec6 The earliest game I found with 5.f4 continued 5...Ng6 but it ended in a White win in Avram-Cramer 1968 US Open. The knight is still far away from influencing e4. 6.Ne2!? This was Lc0's novelty that got Sadler's praise. Stockfish went for more pedestrian development with 6.Nc3 and also got a nice position out of the opening, but Lc0 managed to hold a pawn down ending. 6...d6 One of the points of delaying the development of the queen's knight is to challenge the long diagonal after 6...g6 7.Bd2 Bg7 8.Bc3 7.b3 Bg4 8.Nbc3 Be7 9.g3 Bf6 10.Bd2 This move looks weird since earlier White was play involved trading the bishops on the long-diagonal. Now, Lc0 anticipates playing Nd5 and wants to grab the bishop pair rather than just exchanging bishops. 10...Bf3 11.Rg1 h6 12.Kf2 Bh5 13.g4 Bg6 14.Rg2 freeing g1 for the king. Despite all the moves that looked strange to the human eye, Lc0 has a very nice position, although Stockfish did manage to hold the draw. Overall, it looks like White has another variation to give Budapest players headaches.