Shakmaty Bereolos - The Official Chess Site of Peter Bereolos

12/26/18 - BCE-362, Lasker-Steinitz 1896

The World Championship rematch between Lasker and Steinitz was even more of a rout than the first match. Again, the first four games were decisive, but this time Lasker won all four of them. It took him only 17 games this time to reach the required 10 wins versus 2 wins for Steinitz and 5 draws. Today's BCE correction is from the 14th game. At this point Lasker had 7 wins to Steinitz' 2, so even if Steinitz had held the draw, it is doubtful that he would have come back. The lopsided nature of this match may have also caused this endgame not to get a lot of scrutiny. The game transitioned from a double rook ending to a single rook ending after 54. Kxc2

54...Re2+ 55.Rd2 Re4 56.Rf2 Rg4 57.Kb2 Re4 This was long condemned as the losing move. In addition to Fine it was also given a question mark by Euwe and Hopper as well as by Browne in the 1985 edition of the rook endings volume of the Encylopedia of Chess Endings. 58.g3 I didn't find a book on this match, but there was a contemporary analysis by Tarrasch in Deutsche Schachzeitung where he stated that now Black is slowly pushed back and that he would draw if the pawn stood on g5 or h5. 58...Re5! I'm assuming that Tarrasch's comment was based on the fact that with the Black pawn on g5 or h5 he could counterattack g3 with 58...Rg4 which fails here to 59.Rf4+! 59.Rf4+ Kb5 60.Ka3 Rd5 61.Rf3 Ka5 62.b4+ Kb5 63.Kb3 Kb6 This doesn't lose yet, but Black had some options instead of immediately giving ground 63...Re5 64.Kc3 (This is more testing than 64.Rd3 which Tarrasch gives with no further comment. Black doesn't have any trouble there with 64...Re4 65.Rd5+ Kb6 and the threat of Re3+ forces White back) 64...Kc6! it is important to cover d5 so the White king can't penetrate after 65.Kd4 Re4+. This point is further illustrated by the variation 64...Rd5? 65.Rd3! (not 65.Re3? as given by Euwe and Hopper when Black can use the defesnive technique he missed in the game with 65...Kc6! 66.Kc4 Rd8! heading to g8) 65...Re5 66.Kd4! Re4+ 67.Kd5! Re8 68.Rf3 and White will trade b for f and win with the g-pawn thanks to the offside position of the Black king. There is also a tactical path to the draw with 63...f4 64.gxf4 (64.Rxf4 Rd3+) 64...Rd4 65.f5 Rxb4+! 66.Kc3 Ra4 when the threat of Ra3+ prevents the White pawn from immediately advancing 67.Kd3 Kc5 68.f6 Ra3+ 69.Ke4 Rxf3! 70.Kxf3! Kd6! and the pawn is caught. It's important to study all the alternative drawing methods because in your games there may be only one that works. Conversely, from the attacking side, you should try to narrow the path your opponent has to make the draw. 64.Kc4 Kc6 65.Rb3 Re5? The real losing move as indentified by Petronijevic in the 2014 edition of ECE (65...Rd8!=) 66.b5+ Kb6 67.Kd4 Re4+ 68.Kd5 Re8 69.Kd6 Re1 70.Rf3 Kxb5 71.Rxf5+ Kc4 72.g4 Kd4 73.g5 Rg1 74.Ke6 Ke4 75.Kf6 Ra1 76.g6 Ra7 77.Re5+ Kf4 78.Re7 1-0

12/21/18 - Harris-Bereolos, 1982 HF Closed Championship

The next rook ending from the 1982 HF Closed Championship occurred in the 9th round against William Harris. This game was actually delayed and turned out to be the last game played, but I decided to present it in the order that the rounds were scheduled. I'm going to start with a position much before the ending in hopes that in hopes that it might throw some light on the incredible double blunder that occurs at the transition into the endgame. We were still well within established theory after 15...Bxf6

16.Ng5 He had previously played 16.Bg5 against me, with the game eventually ending in a draw. The line I wanted to draw attention to is 16.Qxe6+ Qxe6 17.Bxd5 Qxd5 18.Rxd5 Nb4 19.Rc5 Nc2 20.Rb1 Rab8 21.Rxb8 Rxb8 22.Nd2 Re8 23.Nf3 Rb8 with equality per an old analysis by Larsen. 16...Bxg5 17.Bxg5 h6 18.Be3 Ne5 19.Bb3 Qd6 20.h3 c5 20...Rae8 is the move preferred by theory. 21.Bxc5 Qxc5 22.Qxe5 Qxf2+ 23.Kh1 Qf5 24.Qxe6+??

Here, this shouldn't work at all. Instead 24.Rxd5 Qxe5 25.Rxe5 Bxb3 26.axb3 is a much better double rook endgame for White than what appears in the game 24...Qxe6 25.Bxd5 Qxd5?? 25...Rf1+ winning is a classic example of the hook and ladder tactic. Maybe I would have looked a little harder at this if I hadn't known the variation with 16. Qxe6 26.Rxd5 Rac8 27.Rc1 Rf2 28.Ra5 Rd8 28...Rc6 with the idea of ...Rg6 is possibly an easier way to draw, but doubling the rooks on the seventh is also fine. 29.Rxa6 Rdd2 30.Rg6 30.Rg1 Rd3 (30...Rxa2 31.Rxa2 Rxa2 32.Rc1) 30...Rxa2 31.Rg3 Kf7 32.Kg1 g5 33.Rf1 Rxf1+ 34.Kxf1

34...Ke6 35.Re3+ Kd5?! There is no need to head to the queenside. White doesn't have a good way to make progress against 35...Kf5 36.Re2 Ra3 36...Ra1+ 37.Kf2 Rc1 is a more accurate defense 37.Re3 37.Rc2 Kc4 38.Kf2 37...Ra2 38.Kg1 Kc4? 38...Rc2 39.Kh2 h5 39.Kh2? White needs to immediately go after the kingside pawns before Black realizes his mistake 39.Re6! 39...Rc2? Black could correct his previous errors with 39...Kd5 but that iwould be hard to play after just marching the king to c4. 40.Kg3? 40.Re6! 40...h5 41.Kf3

41...Ra2? Missing the final chance to return with 41...Kd5 42.g4 hxg4+ 42...h4 is a little more resistant 43.Ke4 Rg2 44.Kf5 Rg3 45.Re7 Rxh3 White must play 46.Kxg5! and White wins. Instead, the very natural 46. Rc7+? protecting the c-pawn is a mistake as after 46...Kd3! 47. Kxg5 Rh1! 48. Rh7 h3! 49. Kg6 Black can use his King to support the h-pawn 49...Ke4 50.g5 Kf3! 51.Kf7 Kg2+ 53.g6 Ra1 and Black holds. 43.Kxg4 Ra5 44.Rf3 Rb5 45.Rf5 Rb1 46.Kxg5 Kxc3 the Black king is cut off both horizontally and vertically, so there is no chance to hold. 47.h4 Kd4 48.h5 Rg1+ 49.Kf6 Rh1 50.Rg5 Ke4 51.Kg6 Kf4 52.h6 Ra1 53.h7 Ra6+ 54.Kh5! Ra8 55.Rg8 Ra5+ 56.Kh4 Ra1 57.Rf8+ 1-0

Lessons from this game: 1. Even if an idea is familiar, double check to make sure it applies in the current position. 2. While an active king is desirable in the endgame, it needs to be active in the correct place. A typical defensive method in a rook ending against an outside passed pawn is to have the rook try to control the pawn, while the king is ready to counterattack when his opposite number goes to the other side of the board to help the pawn advance.

12/19/18 - BCE-359, Steinitz-Lasker 1894

This week's BCE installment comes from the World Championship match between Steinitz and Lasker. This match was played under the rules preferred by Fischer where victory went to the first player to reach 10 wins. This seems pretty mind boggling in light of the Carlsen-Caruana match, but it only took Lasker 19 games to reach the goal. The match started off in back and forth fashion with the players alternating victories in the first four games. Today's entry is from the 4th game. There are some interesting moments before the BCE position. I'm going to start from the last diagram for this game in the book on the match, which has the rather elaborate title The Games in the Steinitz-Lasker Championship Match with Copious Notes and Critical Remarks by Messrs. Gunsburg, Hoffer, Lasker, Mason, Pollock, Steinitz, and Others, and Illustrated by Numerous Diagrams of Interesting Positions together with Biographical Sketches of the Two Players. That's quite a mouthful! In the rook ending after 41...Kd7

Both Mason and Pollock point out that 42.Rg1 makes White's job much easier 42...Rd2+ Gunsburg's suggestion 42...Rc2 doesn't really change the evaluation after 43.Rg1 43.Kc5 Rc2+ 44.Kxb5 Re2 45.e6+? Now Black blockades the pawns and should hold a draw. It was not too late to finally go 45.Rg1 45...Kd6 46.Rd1 maintaining the connected passed pawns. Another try for White is 46.Kxa4 Kxd5 47.b4 Kxe6 48.h3 (48.Rd1 Rxh2) 48...Kd7 when Black should be able to hold the draw. It might take a lot of effort to convince your favorite chess engine of that this is the case. This is the kind of position engines still struggle with. They evaluate the position as very favorable for White because of the extra passed pawn and more active king, but they will flail around for a winning plan. Even if they have the 50 move rule programmed in, there is still a horizon effect if it can even get that deep because White can keep resetting the count with the moves b5, h5, and h4 in various orders. 46...Rxb2+? Black takes the pawns in the wrong order, he should start with the h-pawn 46...Rxh2 47.Kxa4 Rxb2 and Black should have no difficulty holding the draw with the White king cut off. 47.Kxa4 Rxh2 48.Re1 Ra2+ 49.Kb5 Ra8 50.Kc4 g5

51.hxg5? This move is passed over in the tournament book, but it should have cost White the win. In general the attacking side should avoid pawn exchanges. Here, White wins with 51.e7 Re8 52.Re6+! Kd7 53.Rxh6 gxh4 54.d6! Even 51.h5 is better than the text, as White can later target the h6 pawn. 51...hxg5! The starting position for BCE-359 52.Kd4 Ra4+ unlike Fine, both Mason and Pollock correctly evaluate 52...Rf8 as equal 53.Kd3 Ra3+? The losing move as identified by Mason who points out that Black could correct things with 53...Ra8! The game finished 54.Ke4 g4 55.Kf5 Ra8 56.e7 Re8 57.Kf6 g3 58.Kf7 Kd7 59.d6 g2 60.Rg1 1-0 This win brought Steinitz level at 2-2. After draws in games 5 and 6, Lasker won an epic game 7 then rattled off 4 wins in a row after that in games 8-11 to take full control. The match ended in Lasker's favor 10-5 with 4 draws and the chess world had its second World Champion.

12/12/18 - BCE-237f

The second B+egh vs. N+gh is a tougher task because Black can set his pawns on the opposite color of the White bishop. Still, as Shipov showed, this defense can also be broken down. I found a good practical example of this ending in the game Kupper-Binet Tapaszto 1964 Tel Aviv Olympiad after 47. e5

47...Nd7 48.Bg8 Kf8 49.Be6 Nb8 50.Bc8 Nc6 51.Ke6 Ne7 52.Bb7 Ng6 53.Be4 Ne7 54.h4 Ng8 55.h5 Ne7 56.Bb1 Ke8 57.Bg6+ Kf8 58.Kd6 Ng8 59.Kd7 Ne7 60.g4 Nd5 61.Be4 Nb6+ 62.Kd8 Kf7 63.Kc7 Na4 64.Kd6 Nb6 65.Bd5+ Kf8 66.Ke6 Ke8 67.Bc6+ Kf8 68.Bb5 Nc8 69.Ba4 Nb6 70.Bb3 Nc8 71.Kd7 Ne7 72.Ba2 g6 73.Be6 1-0

The knight has been corralled as Shipov described. White may not have executed the win in the fastest way, but you don't get any bonus points for speed.

I also found a very interesting example between two US Chess Hall of Famers, Bill Goichberg and Pal Benko. While Goichberg is rightly lauded for his career in organizing chess events, he was a strong player back in his playing days. In this ending from the 1966 US Open, the players reached a position similar to our position of interest after 37. Nxb2

The obvious difference here is that the Black e-pawn is much further back than in the BCE position. If White can manage to blockade on e4, the game will be a draw since the knight will have 8 squares it can move to instead of the 6 it would have blockading on e2. So the big question is who can get to e4 first. 37...Kf7 37...Bc5+ 38.Kf1 e5 39.Ke2 Kf7 40.Kd3 38.Kf2 38.h3 Bc5+ 39.Kf1 e5 and the pawn will reach e4 38...Kf6 38...Bxh2 39.g3! would be similar to the game 39.Nc4 to stop the Black king from advancing to e5, but this allows Black one final chance. Instead, White can reach a blockade on e4 with 39.h3 Ke5 40.Kf3 Kd4 41.Nd1 e5 42.Nf2 39...Bxh2!? After other moves, White has time to blockade at e4, so Benko takes a risk free chance in the king and pawns versus knight ending. 40.g3! Kf5 41.Ne3+ Ke4 42.Ng4 Bxg3+ 43.Kxg3! h5 44.Nf2+ Kf5 45.Kf3 g5 46.Nd1 g4+ 47.Kg3 Kg5 48.Nc3? Knight versus pawn endings are notoriously difficult for the defender. ere, the only move was 48.Ne3! this move prepares to blockade with Ng2-h4 and after 48...h4+ 49.Kf2 f5 is covered 49...e5 (49...Kf4 50.Ng2+ forces Black back) 50.Nd5 covering f4 50...Kf5 51.Ne3+ and the attack on g4 again thwarts Black 48...h4+! 49.Kg2 Kf4 50.Ne2+ Ke3 51.Nc3 g3? a somewhat strange slip. Black should play the moves in reverse order. 51...Kd3 52.Nd1 g3 52.Nd1+? Again, it is a difficult defense. The knight suffers from lack of space on the back rank. He could have gotten the half point back by giving the knight more options with 52.Nb5 e5 (52...Kf4 53.Nc7 e5 54.Nd5+; the White king can help in the fight against the e-pawn after 52...Kd3 53.Kf3 which was the problem with pushing the g-pawn.) 53.Nc3 52...Kf4 53.Nc3 Kg4! 54.Ne4 h3+! 55.Kg1 e5 0-1

12/11/18 - Matsuo-Win, 2018 Batumi Olympiad

The Nimzowitsch variation has become a popular try for White against the Petroff, even making an appearance in the recent World Championship match. In Yearbook 128, SP Sethuramen explores some new ideas for White in one of the main lines. No one in the Olympiad chose the moves he looked at, so I selected a game that follows a more well trodden path. This game was played in the 10th round of the Olympiad between Tomohiko Atsuo of Japan and Tun Win of Myanmar. 1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Nc3 Nxc3 6.dxc3 Be7 7.Be3 Nc6 in the 11th game of the match, the Challenger went for kingside castling and the Champion didn't make much headway against 7...0-0 8.Qd2 Nd7 9.0-0-0 Nf6 10.Bd3 c5!? and a fairly uneventful game was eventually drawn between Carlsen and Caruana. 8.Qd2 Be6 9.0-0-0 Qd7 10.Kb1

By far the most popular move. Sethuramen's survey looks at 10.a3 and 10.b3 10...h6 This move is fairly uncommon although Mamedyarov has played it several times. Caruana has played the most popular 10...Bf6 a couple of times and Li Chao has championed 10...a6 11.Be2 Bf6 12.Nd4 Nxd4 13.cxd4 d5 a fairly level position has arisen. White has a slight lead in development, which he uses to gain some space on the kingside. 14.h3 0-0-0 15.g4 g5 16.f4 Qe7 17.Rde1 Rde8 18.Bf3 Qd6 19.Rhf1 Rhg8 20.Bg2 gxf4 21.Bxf4 Bg5 22.Re5 Bxf4 23.Qxf4 Rg5 24.Rfe1 Reg8 25.Qf6 R5g6 26.Qf2 c6 27.a3 Qd8 28.Rh5 f5 29.Bf3 fxg4 30.Bxg4 Bxg4 31.hxg4 Rxg4 32.Rxh6 White's nagging pressure has finally seemed to evaporate, but the game only lasts one more move 32...Qg5? Black needed to cover the seventh rank with something like 32...R4g7 33.Rh7 1-0

An abrupt conclusion. White is suddenly threatening Qf7 or Ree7 with devestation along the seventh rank. 33...Rg7 34.Qf8+ Kd7 35.Rxg7+ Qxg7 36.Qf7+ drops the queen; 33...Rf4 34.Qh2 is no help as the rook is pinned against c7 and R1e7 is still on tap. The ever resourceful engine comes up with a way to keep the game going 33...Re4 when White needs to find 34.Ka2 as neither 34.Qf7 Rxe1+ 35.Ka2 Re7 nor 34.Rxe4 Qg1+ 35.Qxg1 Rxg1+ 36.Ka2 dxe4 leads anywhere Now it likes like White will worm his way in. For example, 34...Qg6 (34...Rxe1 35.Qxe1 and White penetrates via e6 or a5) 35.Reh1 Qe8 36.Qf6 Re6 37.Qf7 Qxf7 38.Rxf7 followed by R1h7, but at least here Black can still try to fight.

12/7/18 - Bereolos-Ivanov, 2007 Land of the Sky

Like myself, Alexander Ivanov is a perennial participant at the Land of the Sky tournament. This year's meeting was our 5th time playing each other in the event. We have faced off a couple of times on first board, including in the last round of the 2007 event. In this game, he got a slightly easier middle game to play with a better pawn structure, and eventually ground out the point in a rook ending. I've mostly kept the notes from my original tournament report. I had a quickly look at the ending again using the engine. I should have been able to hold the draw. I want to look at it in more detail to see if my original analysis that 50. Kf3 was the losing move, the engine didn't immediately show a win after the later 53. Ke4.

12/5/18 - BCE-237e

My original post on the BCE revision also mentioned a couple of related positions that had been examined by GM Sergey Shipov on the old Club Kasparov website. These positions look at B+egh vs. N+gh. These should be winning for White, but can be more difficult depending on the configuration of the defender's pawns. In the first case, Position #237e, the pawns are on the same color as the bishop. Here, White should win without much trouble, although Fine gets the move order crossed in one variation.

12/3/18 - Sirosh-Mosadeghpour, 2018 Batumi Olympiad

Back to the closed games for my next Olympiad/Yearbook 128 entry. Davorin Kuljasevic looks at another variation where useful waiting moves are the trend, the main line of the Fianchetto Grunfeld. This game between Ilja Sirosh and Masoud Mosadeghpour is from the 5th round match between Estonia and Iran. 1.d4 Nf6 2.c4 g6 3.g3 Bg7 4.Bg2 d5 5.cxd5 Nxd5 6.Nf3 0-0 7.0-0 Nb6 8.Nc3 Nc6 9.e3 Re8 This is the big tabiya of the variation. Kuljasevic notes that 18 different moves have been tried from this position. I've personally had this position as Black on 3 occassions and faced a different move each time. The five games that reached this position in the Olympiad saw White try 4 different moves, but Black more than held his own with 3 wins and 2 draws. 10.h3

This move is the focus of the survey. In the very next round Sirosh varied with 10.Nh4 against Chirila 10...Be6 Kuljasevic considers the main moves to be 10...e5 or continuing the waiting game with 10...a5. He did not consider the text good enough for Black to achieve equality. 11.b3 a5 Kuljasevic only looked at 11...h6 12.Bb2+/= The text makes sense since White has created a target with b3, a move he certainly would not have played in response to 10...a5 12.Bb2 The only other game to reach this position continued 12.Ng5 Bc8 13.Nge4 e5 and Black was OK and went on to win in Olander-Heinola 2016 Finnish Team Championships. 12.Ba3 is the choice of Komodo 12 and Stockfish each liking White a little over 0.4. 12...a4 13.Nxa4 Nxa4 14.bxa4 Qd7 15.Kh2 Bc4 16.Rg1 odd looking, he was probably worried about 16.Re1 running into an eventual ...Nd3 16...Nb4 17.Nd2 Ba6 18.Ba3 Nd3 19.Ne4 Bc4 20.Nc5 Nxc5 21.Bxc5 Bd5 22.Rb1?! 22.Rc1 keeps the bishop protected, which avoids the tactics that follow. 22...Bxg2 23.Kxg2 e5

This variation revolves around Black trying to play ...e5 under favorable circumstances. Here, the move wins material. 24.d5 24.dxe5 Qc6+ picks up the bishop. Meanwhile, Black was threatening 24...b6 winning the pawn on d4. However, the push does not save the pawn. 24...Ra5 25.Bb4 Rxd5 Black is slightly better because of the weak a-pawns. 26.Qc2 e4 27.Rbc1 This looks like a clear case of the wrong rook. What is Rg1 doing? White should have preferred 27.Rgc1 27...c5 28.Be1 28.Bxc5 Rc8 pinning and winning 28...h5 29.a5 Qf5 30.Qe2 Red8 31.Rf1 Rd3 32.Rc2 Rd1 33.h4 Ra1 34.Rd2 this allows Black to finish things quickly, but White is very tied up and it is hard to suggest moves here. 34...Qf3+ 35.Qxf3 exf3+ 36.Kg1 Rxd2 37.Bxd2 Rxa2 38.Rd1 c4 0-1

12/1/18 - Tennant-Bereolos, 1982 HF Closed Championship

In the summer of 1982 I played in my only double round robin to date, the championship of the Homewood-Flossmoor Chess Club in Illinois. I was still just an expert at the time, this tournament was a great learning experience for me as I got to play stronger players on a consistent basis. I started out pretty well with 2.5 out of 6 despite being the lowest rated player. However, I didn't add to that score, finishing with 4 losses. I couldn't find a crosstable from the event, I would guess that 25% was probably last. I would have done better if I had held a draw in any of the rook endings I had in the event. The first one was in round 2 against the highest rated player, Steve Tennant. We reached a balanced queen and rook ending after 25.f4

25...Qc7 This unecessarily gives White some chances. Black should start with 25...Rc4 26.Qd8+ Qxd8 27.Rxd8+ Kg7 28.Rd7

threatening both e6 and Rxa7. 28...Rc1+?! Black might be able to grovel to a draw with 28...Kf8 29.Rxa7 h5 but there is no need for such measures. Instead, Black should remain calm and play 28...Rc4 29.g3 (29.e6 Rxf4 30.Rxa7 Kf8) 29...Ra4 30.e6 Kf6 31.exf7 (31.e7 Re4 this is the point of putting the rook on the 4th rank 32.Rxa7 Rxe7=) 31...Kg7 32.f8Q+ Kxf8 33.Rxh7 Rxa2

Compare this position to the one after White's 33rd move in the game and you can see the huge differences in Black's favor. The Black pieces are basically on the same squares, but the White king is tied to the back rank and he no longer has an e-pawn. 29.Kf2 Rc2+ 29...a6 30.e6 Kf6 31.e7 Rc8 32.Rc7 Re8 33.Rc6+ Kxe7 34.Rxb6 is a better defensive try for Black 30.Kf3 Rxa2? the final mistake, there were still some chances to hold with 30...a6 or 30...Kf8 31.Rxa7 h5 although both lines are inferior versions of those looked at earlier. 31.e6! Kf8 Now, 31...Kf6 32.e7 and the misplaced Black rook can no longer come to the rescue. 31...Ra3+!? was worth a try because of the variation 32.Ke4?! (instead the careful 32.Ke2 Ra2+ 33.Ke3 Ra3+ (33...Kf6 34.e7 Ra1 35.Ke2) 34.Kd4 secures the point) 32...Kf6 33.e7 Ra1 34.e8N+! when White might still have technical difficulties to solve. 32.Rxf7+! Ke8 33.Rxh7

The comparison position. The active White king and extra pawn make all the difference and he easily finishes things off. 33...b5 34.g4 b4 35.f5 gxf5 36.gxf5 Ra6 37.Rb7 Ra1 38.Ke4 Re1+ 39.Kd5 Rd1+ 40.Ke5 Rd8 41.f6 1-0

Lessons from this ending: 1. King activity in rook and pawn endings is key. Even giving up a pawn can be well worth it. 2. Try to maintain flexiblity with the rook. 28...Rc4 would have allowed black both defensive options with ...Re4 as well as attacking options with ...Ra4. 3. Don't panic. Obviously, I had not correctly evaluated the queen exchange. That is a time to reassess the position and perform concrete calculations. 4. Always try to present your opponent with opportunities to go wrong. Black is still losing after 31...Ra3+ but it would give White a chance to make things more difficult whereas the game continuation was just an automatic win for White.