Shakmaty Bereolos - The Official Chess Site of Peter Bereolos

6/26/19 - BCE-215, Kling and Horwitz 1851

The final BCE position for this set that Fine labelled Horwitz 1880 is an opposite colored bishop ending with 2 pawns versus 1, none of which are passed. Given that description, it is not surprising that the position turns out to be a draw instead of a win.

However, BCE-215 was one of three similar position Kling and Horwitz published in their periodical The Chess Player in 1851 and 1852. I guess it was Fine's bad luck to use the one that wasn't a win. In 1851, there was also this version.

1.Bc4+! b5 1...Kb6 2.a5+ Kc6 3.b5+ Kc5 4.Kxb7 Kxc4 5.Kxa7 Kxb5 6.a6! 2.Bxb5+! Kb6 3.Be2 Kc6 4.Bf3+ Kb6 5.Be4 Ka6 6.Bd3+ Kb6 7.Bb5 Bb8 8.a5+ 8.Kxb8? is stalemate 8...Ka7 9.Bd3 Be5 10.b5! Bd4 11.Kc7 Be5+ 12.Kc6 Bd4 13.b6+ Kb8 14.a6 -- White is ready to mate with 15.a7+ Ka8 16.Be4 followed by a king move 1-0

The following year, there was another version of the theme

1.Bf3+! Kg5 2.Be4 Kh5 3.Bxg6+ Kg5 4.Bd3 Kh5 5.Be2+ Kg5 6.h4+ The idea from the previous study does not work here because the White bishop is of the opposite color of the one the h-pawn promotes on. 6.Bg4? Bg7! 7.h4+ Kh6! 8.Bf3 Bd4 9.g4 Be3 10.Kf6 Bd4+ 11.Kf5 Bf2 12.g5+ Kg7 13.h5 Be3 14.h6+ Kh8 and 15...BxP 6...Kf5 7.g4+ Kf4 8.Bd1 1-0

6/25/19 - Palatnik-Bereolos, 1996 Fairfield Glade

My first outing against GM Sam Palatnik was a fairly miserable effort on my part. An off-the-cuff novelty with a bad followup left me with a totally passive position and I was slowly squeezed off the board.

6/19/19 - BCE-166, Horwitz 1880

This week's BCE entry is another Horwitz study with a bishop battling 3 connected passed pawns. This time White has some pawns of his own, so no complaints about needing to switch the colors. It is a bit strange that Fine missed the stalemate trick in the first main line as he uses it in the second main line.

This particular position seems to have been first published in Volume 1 of The Chess-Monthly. The trapped position of the White king in the starting position is a bit strange. However, an earlier study in 1860 in The Chess Player's Chronicle shows how it might have gotten there.

This problem is Black to play and draw and is attributed to Mr. Horwitz, so I guess no credit to Kling on this one. The solution here is 1...Ke6! 2. Kxa1 Kf7! and draws, which is BCE-166 except here the Black pawn is on a5 instead of a6. The starting position of the Black a-pawn doesn't seem to matter, but perhaps Horwitz thought he found something which led to the 1880 version. I don't understand why he didn't leave the two introductory moves, although the position is still fairly artificial. I tried to add to it, but only came up with one additional move.

I've followed modern color conventions here, so the task is White to play and draw with the solution 1.Ra1! Kb2 2.Kd3! Kxa1 3.Kc2! etc. This still isn't really satisfactory since I can't figure out what a logical previous move for Black might have been to reach the last diagram.

6/12/19 - BCE-143, Kling and Horwitz 1851

Back in February and March, I had a series of BCE corrections from the work of Kling and Horwitz. This week begins a new trilogy of positions from the Bishop and Pawn Endings section of BCE that Fine attributes to Horwitz 1880. Researching the history of these positions, it looks like they all predate 1880 and should also give Kling some credit. Fine seems to be following the citations in Berger, which Fine gives as his primary source. Berger cites Volume 1 of The Chess-Monthly. The Chess-Monthly was a periodical that began in 1879 edited by Hoffer and Zukertort. Each issue had a section titled End-Games by B. Horwitz, but there is no accompanying text to give the reader a clue if these were original or from previous work.

BCE-143 is a battle of a lone White bishop against 3 connected passed pawns. In another one of BCE's inconsistencies Black is clearly the only one playing for a win, despite Fine's saying in the introduction that he uses White to denote the superior side and even has examples where he has switched the colors of actual game positions to that end. I bring it up here because unlike BCE-143 and Berger, the position in The Chess-Monthly has the colors reversed.

I think Berger switched the colors because all of his examples of piece versus pawns have White with the piece. Fine seems to have just copied this position and forgotten about what he said in the introduction.

The history of this position goes back further. Kling and Horwitz edited their own periodical The Chess Player from 1851 to 1853 and this position appears in the September 27, 1851 issue attributed to The authors, so I think the correction citation is Kling and Horwitz 1851 as I have put in this post's title. This postion is seems to have come from an actual game as the authors state This beautiful strategem happened in a game played at the Philidorian Chess Rooms, Strand.

That seems to mostly solve the origin story, but the analysis of the position also has some history behind it. The original solution given in The Chess Player as well as in The Chess-Monthly and Berger (with colors reversed) begins 1. f6?, which leads to a draw as shown by Fine in BCE. The van der Heijden study database credits the discovery of the winning cook 1.Kf4! to Max Karstedt in Deutsches Wochenschach 1906. Unfortunately, I was unable to find a copy of that volume on the web, so I don't know if he just gave the line shown in the database, which corresponds with Fine's main line, or if he included other variations. Since the BCE correction is in a side variation, it isn't clear if Fine originated this error or if he just passed it down from Karstedt.

As a contemporary example of this ending, I'd like to look at the game Nakamura-Hillarp Persson from the 2005 Sigeman tournament. Before Norway Chess came along, the Sigeman was probably the top tournament in Scandanavia. It is usually held in Malmo, Sweden, but the 13th edition in 2005 was split between Copenhagen, Denmark and Malmo. Timman and Sasikiran topped the table in 2005, with Nakamura a half-point back. It was not a good tournament for Hillarp Persson who was dead last with 1.5/9. Against Nakamura, he abandoned his favorite Modern Defense for the French. The endgame was one that all French Defense players dread with a knight against the light-squared bishop in a fixed pawn structure after 40...Kxa5

The position is so closed that Black should objectively hold, but White can play for the win with no risk, so Nakamura makes the effort. Step 1 is to open in roads for the king on the queenside. 41.Kb3 Kb6 42.Nc3 a5 43.Ka4 Bd7+ 44.Ka3 Bc6 45.Na4+ Kb5 46.Kb3 Bb7 47.Nc3+ Kb6 48.Ka3 Bc6 49.b4 Ka6 49...axb4+? 50.Kxb4 Bb7 51.Na4+ and Black has to give ground 50.Kb3 Kb6 51.Na4+ Kb5 52.bxa5 White could win a pawn with 52.Nc3+ Ka6 53.b5+ Bxb5 54.Nxd5 but the passed a-pawn should allow Black to hold. It is somewhat similar to the famous Karpov-Kasparov ending where White should not rush to capture the d-pawn since it gives the bishop more scope. 52...Kxa5 53.Nc5 Be8 54.Ne6 Bf7 55.Nd8 Be8 56.Nb7+ Kb6 57.Nd6 Bd7 58.Kb4 Kc6 59.Nb5 Kb6 60.Nc3 Bc6 61.Nb1 61.Na4+ Bxa4 62.Kxa4 Ka6 Black has the opposition, so it is a draw. The sacrifice 61.g4 to clear g3 for the knight doesn't work either 61...fxg4 62.Ne2 Be8 and the Black bishop can't be driven off the e8-h5 diagonal. 61...Bd7 62.Nd2 Bb5 63.Nb3 Bd7 64.Nc5 After a bit of maneuvering, Nakamura targets the one pawn that the bishop can't protect, on f6. 64...Be8 64...Bc8? 65.Na4+ forces the Black king to give ground, but now the knight can get at f6. 65.Ne6 Bf7 66.Nf8 Kc6 67.Nh7 Kd6 68.Nxf6

White has won a pawn, but his knight is trapped. However, to collect it Black must allow the White king to penetrate. 68...Bg6 69.Ng8 Bf7 70.Nf6 Bg6 71.Ng8 Bf7 72.Nh6 Be6 73.Kb5 Bd7+ 74.Ka5 Be6! 75.Kb6! Kd7 76.Kc5 Kc7! Now, to make further in roads, Nakamura sacrifices the knight. 77.Nxf5 Bxf5! 78.Kxd5! Kd7 79.Ke5! Bg4 80.Kf6

80...Kd6? after this, the king is out of position for the ensuing ending against the 3 connected passed pawns. Instead, Black holds the draw with 80...Bf3 81.f5 Ke8 82.Kg6 Kf8 83.d5 Bxd5! 84.Kxh5 81.f5! Kd7 82.Kg6 Bf3 83.d5 Bxd5 84.Kxh5! Ke7 85.Kg5? White needs to use his king to force the Black king into a weaker defensive position on the back rank. 85.Kg6! Be4 (85...Kf8 86.g4 Bf7+ 87.Kf6) 86.h5 Kf8 87.h6 Kg8 88.g4 and wins 85...Kf7! 86.g4 Bf3 87.h5 Kg7!

This is the model defensive setup shown by Fine in the previous BCE position (#142). The bishop is in an active position where it can target the White pawns from behind ("pawns in the crosshairs" is Dvoretsky's term for this technique.) to stop them from advancing. 88.Kh4 8On 8.h6+ Kh7! 89.Kh5 Black can make an immediate draw with 89...Bxg4+ 88..Bd1? 88...Be4! was necessary to prevent g5 89.g5! Bc2 90.h6+ Kh7 91.Kg4! The pawns are now too far advanced for Black to hold. Nakamura first improves his king then advances the pawns. 91...Bb1 92.Kf4! Bc2 93.Ke5 Bd3 94.Ke6 Bc4+ 94...Bc2 95.f6 Kg6 96.Ke7 Bb3 reaches the same end 95.Ke7 Bb3 96.f6 Kg6 97.f7 White can also reverse the order with 97.h7 Kxh7 98.f7! Bxf7 99.Kxf7! 97...Bxf7 98.h7! Kxh7 99.Kxf7! Kh8 100.Kg6 Of course not 100.g6? stalemate 1-0

6/11/19 - Martinez-Smith, 2018 Batumi Olympiad

Like the other Yearbook 128 survey on the Sicilian Taimanov, the main line of one by Pavel Skatchkov and Dimitry Frolyanov was not reached in any Olympiad games. The ones that did reach it were not too interesting, so I picked one that deviated a couple of moves earlier. The line chosen in the game between Jose Martinez Alcantara and Shreyas Smith from the Peru-Jamaica match looks like one that needs to be put to rest.

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Qc7 6.Be3 a6 7.Qf3 d6 7...Nf6 8.0-0-0 Ne5 9.Qg3 b5 was the subject of the survey 8.0-0-0 Bd7 9.Qg3 b5?! Black probably has to stick with the main move here 9...Nf6. Since it will take some time to castle kingside and once the king gets there it is staring into the face of Qg3, I'll toss out the untested novelty 9.0-0-0 for those looking for an original path. 10.Bxb5!

10...Nf6 Black probably has to accept the sacrifice, to at least have some material for his trouble. 10...axb5 11.Ndxb5 Qb8 (11...Qb7 was brutally dealt with in Ragger-Maiwald, 2017 Bundesliga 12.Qxd6 Rc8 13.Bc5 Nce7 14.Qd3 1-0) 12.Nxd6+ Bxd6 13.Rxd6 and Black suffers from the same problems as in the game trying to figure out how to get his pieces out of the kingside. White scored another fast win from here in Lentjes-Jancke, Lueneburg 2015. 13...Nf6 14.e5 Nh5 15.Qg5 g6 16.Rhd1 h6 17.Qh4 0-0 18.g4 Rc8 19.gxh5 Nxe5 20.hxg6 Rxc3 21.gxf7+ Nxf7 22.Rxd7 Rxa2 23.Rg1+ Ng5 24.Rxg5+ hxg5 25.Qh7+ Kf8 26.Qh8# 1-0 11.Bxc6 Bxc6 12.f3 One high level game got this far. In Alekseev-Potkin 2017 Russian championship White removed the bishop pair with 12.Nxc6 and Black's compensation is again questionable. 12...Bb7 Maybe 12...Bd7 keeping the b-file open, and supporting a future advance of the a-pawn is better, but Black's big problem is how to develop the kingside. 13.Kb1 Rc8 14.Nb3 Be7 Now the king is a permanent resident of the center, but a fianchetto didn't look feasible because of the weakness of the d-pawn. 15.Qxg7 Rg8 16.Qh6 Rxg2 17.Rhg1 Rxg1 18.Rxg1 d5 19.Bg5 Qb6 20.Rd1 Qd8 21.Qg7 h5 22.h4 Ba8 23.Nd4 Bb7 24.e5 Rxc3 25.exf6 Bf8 26.Nxe6

A small bit of tactics to finish things off. 26...fxe6 27.f7+ Kd7 28.Qxc3 Be7 29.Bxe7 Qxe7 30.Qg7 1-0

6/5/19 - BCE-594, Chigorin-Janowski, Karlsbad 1907

This week's BCE correction comes full circle to complete the trio of positions from the games of Mikhail Chigorin as we return to his final tournament at Karlsbad in 1907. In the second round Chigorin had White versus David Janowski and they reached a two rook versus queen position after 49.cxd4

At first glance it would seem that Black should have a relatively easy win with soon to be connected passed pawns against White's split pawns. The tournament book takes this point of view basically saying that it is a matter of technique. However, the Black rooks are somewhat uncoordinated and after analyzing this ending extensively with assistance from the engine and tablebases, I think the position is objectively drawn. 49...Rxg4 50.Kc4? It is better to play 50.Qa7 to protect both pawns and to stop the rooks from getting organized 50...h5? It looks like Janowski missed an opportunity here. In the Queen vs. Two Rooks section of his Endgame Manual, Dvoretsky emphasizes A standard method is doubling the rooks to gain, or at least stop an enemy's pawn. Janowski had a chance to that here with 50...Rc1+ 51.Kd5 Rc7 threatening to win the d-pawn with Rd7+ and a sample variation is 52.Qe8 Rg5+ 53.Kd6 Ra7 54.d5 Rg6+ 55.Ke5 Ra5 56.Qd7 Rg5+ 57.Ke6 h5 and White is in zugzwang and must lose one of his pawns. 51.Qe8 Rg6 52.Qb5 White could even advance his pawn here as 52.d5 Rxa4+ 53.Qxa4 Rg4+ 54.Kb5 Rxa4 55.Kxa4 is a drawn ending 52...Kh6 53.a5 Rc1+ 54.Kd3 Rcc6 Black's rooks stop White's passed pawns, but they are otherwise passively placed 55.Qb8 The starting position of BCE-594 55...Rcd6 56.Kc4 h4 57.Kc5 Kg5 58.d5 h3 59.Qe8 Kf4 60.Qe1 Rh6 61.Qf2+ Kg4 62.Qg1+ Kf5 63.Qf2+? The BCE correction is that after 63.Qf1+ Black can't escape the checks. The point is that h1 is covered by the White queen, so trying to follow the game continuation would lose 63...Kg6? 64.Kxd6! h2 65.Qh1! 63...Kg6! 64.Qc2+ now 64.Kxd6 h2 and Black will queen 64...Kf7 65.Qh2 Rdf6 66.Qc7+ Kg8 67.a6 Rxa6 Rook and two versus queen and one pawn is almost always winning for the rooks, and Janowski has no problems converting. 68.Qb8+ Kh7 69.Qb1+ Rag6 70.d6 Rh5+ 71.Kc4 h2 72.d7 h1Q 73.Qxh1 Rxh1 74.d8Q Rg4+ 75.Kb5 Rh5+ 76.Kb6 Rg6+ 77.Ka7 Rf5 78.Qd3 Rgf6 79.Kb7 g6 80.Kc7 Kg7 81.Qd4 Rf4 82.Qc3 Re4 83.Kd7 Ra4 84.Ke7 Ra7+ 85.Ke8 Ra8+ 86.Ke7 Raa6 87.Qb2 g5 88.Qc3 g4 89.Ke8 Rac6 90.Qg3 Rce6+ 91.Kd7 Re4 92.Qh4 Kg6 93.Qh8 Kf5 94.Qh5+ Kf4 95.Qh2+ Kg5 96.Qd2+ Rff4 97.Qg2 Rd4+ 98.Kc6 Rf3 99.Qe2 Rdf4 0-1

6/4/19 - Newsom-Bereolos, 2010 South Carolina Open

I was checking some of my analysis with an engine when I found a big error in my game against Gary Newsom at the 2010 South Carolina Open. After 44.d7

I went for the pawn ending. 44...Nxd7? The path to the draw was 44...Ke7! 45.Ke3 Kd8 and neither side can make progress (Black also has the tactical solution 45...g4 46.hxg4 Nxg4+ 47.Kxe4 h3 48.gxh3 Nf2+ and the Black knight will eventually return to take the d7 pawn after taking h3.) 45.Nxd7! Kxd7 46.Ke3 Kd6 47.Kxe4 Ke6 48.b3 The only alternative I looked at in my original notes was 48.b4 Kf6! 49.Kd5 Kf5! 50.Kc5 Kf4! 51.Kb6 Kg3! 52.Kxb7 Kxg2! 53.b5 g4 54.bxa6 gxh3! 55.a7 h2! 56.a8Q h1Q! with a draw, but White does have a win 48.g4 hxg3 49.Kf3! Kf7 50.Kxg3 Kg7 51.Kg4 Kg6 52.b3 Kf6 53.Kh5 Kf5 54.Kh6 g4 (54...Kf6 55.b4 Kf5 56.Kg7 Kf4 57.Kf6) 55.hxg4+! Kxg4 56.Kg6!

and although Black will capture a b-pawn first, White will win both of Black's queenside pawns. 48...Kf6! 49.b4 It's too late for 49.g4 since in the above variation, White needed both of his tempo moves with the b-pawn 49...hxg3! 50.Kf3! Kf5 51.Kxg3 Kf6! 52.Kg4 Kg6! 53.b4 Kf6 holds. White would actually lose here if he tried to win 54.Kh5? Kf5! 55.Kh6 g4! and it will be Black who wins both pawns on the queenside 49...Ke6! 50.Kf3 Kf5 51.Ke3 Ke5 52.Kf3 1/2-1/2

Lessons from this ending: 1. Don't automatically discard backwards moves. The key move to win was a king retreat Ke4-f3. This can be particularly hard to find when the king has just moved forward on the previous move. 2. As always, king endings can be boiled down to pure calculation.

6/1/19 - Diermair-Matsuo, 2018 Batumi Olympiad

I'm reaching the home stretch of my Olympiad/Yearbook 128 series. The only remaining surveys on the closed openings are on the English Opening. The first of these was a gambit line reviewed by Carsten Hansen, which didn't find any takers in the Olympiad. I picked the game that came closest to that line, a somewhat one-sided win by Andreas Diermair over Tomohiko Matsuo in the Austria-Japan match.

1.Nf3 Nf6 2.c4 c5 3.Nc3 d5 4.cxd5 Nxd5 5.d4 Hansen's survey was focused on the gambit 5.e3 e6 6.Nxd5 exd5 7.b4!? 5...e6 6.e3 Nc6 7.Bd3 cxd4 8.exd4 Be7 9.0-0 0-0 10.Re1

A very common IQP tabiya that can arise from a wide variety of openings. 10...Ncb4 Putting a piece on f6 with 10...Bf6 or 10...Nf6 is much more common with thousands of games with each 11.Bb1 Bd7 This move is pretty rare. The most played move is 11...Nf6 so that Black can maintain the blockade of the d-pawn using the b4 knight after 12.a3 Nbd5. However, the engines suggest some other ideas. Fire suggests the similar idea 11...Nc7. This reserves f6 for the bishop, and the knight can support the ...b5 push. The drawback is that there is one less piece to defend the kingside and the c-file is blocked. Houdini doesn't fool around with preparation and pushes 11...b5!? immediately looking for fast development for a pawn after 12.Nxb5 Ba6 13.Nc3 Rc8. Stockfish wants to bring the queen out with 11...Qb6, but White has to be happy after 12.a3 Nc6 (12...Nxc3 13.bxc3 Nd5 is a worse version of the game after 14.Qd3) 12.a3 The database shows White with a nice +7 -2 from here. 12...Nxc3 Relieving White of the isolated pawn, but 12...Nc6 just looks like two tempi down on normal positions. 13.bxc3 Nd5 14.Qd3 g6 The only top level game that got here was Portisch-Bagirov Beverwijk 1965 which continued 14...Nf6 but the great Hungarian champion scored a quick victory in only 24 moves. Portisch tied for first in that tournament, the first of four times he finished on top of the tournament that is today known as Tata Steel Chess and now played in Wijk aan Zee 15.Bh6 Re8 16.c4 Nf6 17.Ne5 Rc8 18.Ba2

18...Nh5? Black is under a lot of pressure here and commits a tactical mistake. Perhaps the point of the text was to play ...Bg5 to trade bishops or maybe ...Bf8-g7. The immediate 18...Bf8 19.Bg5 Bg7 doesn't bring much relief after 20.Ng4 when Black will likely have to make further weaknesses around his king with ...h5 in order to deal with the pin. Maybe Black's best is something like 18...Bc6 trying to hold up d5, but I don't really see a plan for Black if White continues the buildup with 19.Rad1 19.Qf3 attacking both f7 and b7 19...Bf6 19...f6 20.Nxd7 Qxd7 21.Rxe6 Qxd4 22.Rae1 is no better 20.Qxb7 Bxe5 21.Rxe5 Rb8 22.Qf3 Ng7 23.g4 dominating the knight on g7 23...Qb6?? Black doesn't have any compensation for the pawn, but this move ends the game immediately. 24.Qf6

White will soon mate on g7 1-0