Games versus GMs 
Welcome to my chess page. This is mostly random thoughts and analysis in the form of a chess diary with other sections of the site slowly developing. A lot of the content will come from my own experience. There are two reasons for this. One, so I can use this site as a selfimprovement tool. Two, so you the readers will have content that is not found on other chess sites. Follow the link to the left to reach my annotated games against grandmasters. Send me comments and ideas If you link to this site, please point to www.bereolos.net so that your link will be correct no matter where I host the site. Pete 

Corrections to Basic Chess Endings  
Links  
Archives  
I had a pretty decent result at Kings Island back in November, only losing in the final round to my former Tennessee Tempo teammate Alex Shabalov. I'm not too satisfied with this game. I didn't really set him any difficult problems to solve and although I resisted a long time, I didn't manage to find my way to a draw. These kinds of losses are the ones I need to remove in order to get better.
This week's BCE position bears a lot of similarity to last week's. Again, it is a bishop versus pawns study by Reti (this time from 1923) that was cooked by Bondarevsky in 1955 and later repaired by Benko. However, there are a couple of new twists to the tale this time.
Unlike BCE162, Benko included his correction in the revised BCE as Position 305.
This is Reti's position with all the pieces shifted one file to the right. Now it really is a draw after 1.Kg7! For some reason Benko inverted the starting moves with 1.e5? when Black wins with 1...fxe5! 2.Kg7 e4! 3.h6 e3! 4.h7 e2! 5.h8Q e1Q! and Black will consolidate his extra piece. 1...Bd3 2.e5! fxe5 2...dxe5 3.h6! Ke6 4.Kg8 Bc2 5.Kg7 with a positional draw was Reti's idea since Black can't advance his pawns without blocking the bishop's diagonal. Benko's version also introduces a new theme into the position with 3...Bh7 4.Kxh7 Kf7! 5.Kh8 e4 6.h7 e3 7.g5 fxg5 stalemate 3.h6! d5 The point of shifting the position is now revealed as the cook with Bh8 would now be Bi8, which is not on the board 4.g5! d4 5.g6! Bxg6 6.Kxg6! d3! 7.h7! d2! 8.h8Q! d1Q! 9.Qxe5+ and a draw
Benko didn't stop there, and attempted to further improve the study. While still a draw, this version also has a few flaws.
1.g4
Benko gave 1.Kg7 a question mark because of 1...f5
but White still has a fantastic draw here with
2.exf5! d5 3.h4! d4 4.h5! Bxh5 5.f6! d3 6.g4! Be8
(6...d2 7.gxh5! d1Q 8.f7!)
7.g5! d2 8.g6! d1Q 9.f7! Qd7 10.Kg8 Qe6 11.Kg7!
1...Kd8
1...Kd7 2.Kf8! Bg6 3.Kg7! Bxe4 4.Kxf6!
2.h4! Ke7 3.h5! Bc6 4.e5! fxe5 5.h6! Be4! 6.g5 d5 7.Kg7 d4 8.g6! d3 9.h7! d2! 10.h8Q! Bc2!
Benko concluded the study with 10...d1Q? 11.Qh4+! with a draw
because Black is losing his bishop.
However, the resulting Q+P vs Q+P ending is a win for White thanks to his far advanced gpawn.
Shades of
Kasparov vs. The World!
In the 10th round, I lost yet another rook ending, this time to Walter Brown. Note, this is not the 6time US champion who spelled his last name with an e on the end. Walter mostly stopped playing tournaments in the 1990s, but stilled stayed active in chess as a TD for the CCA. We entered the rook ending after an exchange of knights on c2 42.Rxc2
42...Rc8 Black could try to activate his rook along the ffile with 42...Rf8 when 43.Rf2 can lead to some interesting pawn endings. 43...Rxf2 44.Kxf2! Kc6 45.Ke3 Kc5 46.a3! d5 47.cxd5! Kxd5!
The engine says it is a draw, which goes against all I thought I knew about pawn endings. It seems like White should be able to decoy with the queenside pawns and then win on the kingside. But because of the more active Black king, White is unable to trade off all the queenside pawns, so Black maintains the balance with his bpawn. A couple of sample lines 48.g4 (48.h4 e4 49.g4 Ke5 50.g5 Kd5 51.a4 Kc5 52.Kxe4 Kb4 53.Ke5 Kxb3 54.Kf6 Kxa4 55.Kg7 b5 56.Kxh7 b4 57.h5 gxh5 58.g6 b3 59.g7 b2 60.g8Q b1Q+) 48...g5 49.Kd3 (White actually loses with 49.h3? e4! 50.a4 Ke5! 51.b4 Kd5 52.a5 Kc4! 53.Kxe4 Kxb4) 49...e4+ 50.Ke3! Ke5 51.a4 Kd5 52.a5! Kc5 53.Kxe4! Kb4! 54.Kf5 Kxb3! 55.Kxg5 Kb4 56.Kh6 Kxa5! 57.Kxh7 b5! 58.g5 b4! 59.g6 b3! 60.g7 b2! 61.g8Q b1Q+! and the queen ending with an hpawn is drawn.
Instead of trading on f2, Black also has 43...Rf5 44.g4 Rf4 45.Rxf4 exf4
and it seems that there are a couple of ways White can make a drawn, but both just barely. a) 46.h4 Letting Black have a protected passed pawn is losing (46.Kf2? g5! 47.Kf3 Kc6! 48.Ke4 Kc5! 49.a3 b6 50.Kd3 d5!) 46...h6 47.g5! hxg5! 48.hxg5! Ke6 49.Kf2 Kf5 50.b4! Kxg5 51.a4 Kf5 52.a5 d5 (52...Ke6 53.b5! Kd7!
and now White just waits with 54.Kf3 g5! 55.Kg4 as the threat of a6 ties the Black king to the queenside) 53.b5! dxc4 54.a6! bxa6 55.bxa6! c3! 56.a7! c2! 57.a8Q! c1Q! 58.Qd5+! Kg4 59.Qd7+! (59.Qg2+? Kh5 60.Qh3+ Kg5 61.Qg2+ Kf6 and White is out of checks 62.Qh2) b) 46.g5 Ke6 47.Kf2 Kf5 48.h4 Ke4 49.Ke2! f3+ 50.Kf2! b6 51.a4 Kd3 52.Kxf3! Kc3! 53.Ke4 Kxb3! 54.Kd3 and White has horizontal opposition 54...Kb4 55.Kd4! Kxa4 56.Ke4!
After that long digression, lets return to the game 43.Rd2 Rc5 44.Kf2 White could hold up ...b5 with 44.a4 Ke6 but actually has to be a little careful as now the pawn ending after 45.Kf2 d5 46.cxd5+ Rxd5 47.Rxd5? (47.Rc2) 47...Kxd5 48.Ke3 Kc5 49.Ke4 Kb4 is winning for Black! 44...b5 45.cxb5 Rxb5 46.Ke3 Ke6 47.Rb2 Rc5 48.Kd3 It is simpler to get the passed pawns rolling with 48.b4 Rc3+ 49.Kd2 Ra3 50.b5 Kd7 51.Rc2 48...d5 49.Rc2 Again, the passed pawns should be pushed. 49.b4 Rc1 50.b5 Kd7 51.b6 Kc8 52.a4 Ra1 53.Rb5 49...e4+ 50.Kd2 Ra5 50...Rb5 51.Rc8 51.a4 d4 52.Rc6+ Ke5 53.Rc7 h5 54.Re7+ Kf5 55.Rd7 e3+ 56.Ke2 56.Kd3 Re5 56...Ke5? 56...Ke4! 57.Re7+ Re5
57.Re7+? 57.Kd3! Rd5 58.Rc7! puts Black in a lot of trouble. His pawns are immobile While White's can run free. Putting the rook on the cfile cuts off the Black king from the White pawns while the White rook can still help control Black's pawns from c4 or c2 58...Kd6 59.Rc4! (59.Rc2 Rc5!) 57...Kd6 58.Re4 Rd5 59.Kd3 Kc5 60.Re8 Kb4 61.Rb8+ Ka3
62.a5? The epawn had to be kept under control with 62.Re8 Kxb3 63.a5 Rxa5 (63...Kb4 64.a6 Rd7 65.a7 Rxa7 66.Kxd4) 64.Kxd4= Ra2 65.Rxe3+ 62...Re5! 63.Ke2 d3+ 64.Ke1 Rc5 65.Kd1 e2+ 65...Kb2 is a more immediate mate 66.Kd2 Re5 01
Lessons from this ending: 1. Passed pawns must be pushed. I could have saved myself alot of trouble by getting the queenside pawns rolling. 2. Pawn endings are all about calculation. Relying on intuition can easily lead you astray. 3. Always maximize the utility of your pieces. The dual ideas of cutting off the Black king while simultaneously attacking/controlling the Black pawns with 58. Rc7! perfectly illustrate this theme.
The Hungarian theoreticians Peter Lukacs and Laszlo Hazai are longtime collaborators and contributors to the New In Chess Yearbook series. It is very rare to have an edition of the Yearbook without a survey by them in it and Volume 128 was no exception. While their surveys are always worthwhile, their examination of a variation in the antiBerlin Ruy Lopez didn't fit in very well for my Olympiad series as there were no games played with the specific line they looked at. So for this post I selected a short game in the variation that featured a somewhat similar pawn structure and piece arrangement. This was Board 3 of the match between Azerbaijan and England with Arkady Naidiitsch as White against David Howell.
1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3 Bc5 5.Bxc6 5.c3 00 6.00 Re8 7.Nbd2 a6 8.Bxc6 dxc6 9.Nc4 was the survey variation 5...dxc6 6.00 Qe7 7.Nbd2 Bg4 8.Nc4 Nd7 9.h3 Bh5 10.Bd2 f6 11.b4
This new move is Komodo's choice. Deep Fritz and Stockfish prefer to prepare it with 11.a3 and 11.Rb1 respectively. None of the engines give White very much here. 11...Bxb4 12.Bxb4 Qxb4 13.Rb1 Qe7 14.Rxb7 Nb6 15.Na5 00 16.Nxc6 Qc5 17.Nxa7 Qa3 18.Nc6 Qd6 19.Na7
19...Qa3 19...Na4 with the idea of Qa6 give Black a big edge. White will have to struggle for a draw with 3 pawns for the piece. Instead, Howell decides to repeat moves. Note that 19...Nd7 isn't as good since 20. Nb5 Qa6 21. Rxc7 Qxb5 22. c4 forces the Black queen to abandon defense of Nd7. 20.Nc6 Qd6 21.Na7 Qa3?! ½½ I wonder if Howell would have played 21...Na4 if it wasn't a team event where sometimes the strategy is to hold with Black and press with White. It didn't work out well here as the Azeris swept the remaining boards to with 3.5:0.5.
I've posted my game with Gabriel Schwartzman from the penultimate round of the 1994 US Open to the GM games section. I was having a pretty good tournament with 7.5/10, just behind the leaders and was actually paired down in this game as Schwartzman was only at 7/10. I ended up in a pretty passive middle game position and tried to avoid creating further weaknesses. I was rewarded when he allowed me some exchanges that freed my position somewhat, but I almost immediately handed him back the advantage with an illtimed pawn exchange. A very interesting queen and bishop ending arose, where I think Black may be drawing. However, in the game he kept the pressure on and I eventually blundered and lost.
Schwartzman was a chess prodigy. At the time, it was very rare to be a teenage grandmaster. However, it looks like he stopped playing chess in 2001.
One of the legendary chess composers is Richard Reti. I would expect most chess players to know his famous king and pawn study by the keymove (1.Kg7) alone. Reti was quite prolific with over 100 endgame studies to his name. Of course, composing that many studies in the precomputer age means some of them would eventually get cooked. That is the case with BCE162, which was one of several similar studies published in 1922. This one was cooked by Igor Bondarevsky in 1955. Benko did not include this position in the revised edition of BCE. However, Benko later corrected Reti's study by moving the starting postions of the kings.
1.h4! 1.Kd2? Bc6! 2.h4 Be8 1...Kg2 2.Kd2! 2.Kd3? Kf2 3.b5 Be2+ 2...Kg3 2...Kf2 3.b5 Be2 4.a6 bxa6 5.b6 3.Ke3! Now we are back in the main line of Reti's study. 3...Bg4 4.b5! Kxh4 5.b6! Bc8 6.Kf4 10
Thus, the study is saved, but I didn't really like the unnatural starting position of the Black king on h1. It isn't really clear how the king would have ended up there in a game. I fooled around with the position a little and came up with the following alternatve.
I had some versions with a dark squared bishop for White, but I didn't think they were fully satisfactory, so I settled on this one with the rook. 1.Rc5! c1Q 2.Rxc1! Kxc1 3.Ke3! Kd1 3...Bb5 4.h4! Be8 5.Kd4! 4.h4! Ke1 5.Ke4! 5.Kf4? Kd2; 5.Kd4? Kf2 6.Kc5 Kg3 (6...Ke3 7.Kb6 Bf3! 8.b5 (8.h5 Bxh5! 9.Kxb7 Be2 10.a6 Kd4! 11.b5 (11.a7 Kc4) 11...Bxb5! 12.a7 Kc5) 8...Kf4) 7.Kb6 5...Kf2 5...Kd2 6.Kd4! Kc2 7.Kc5! Kd3 8.b5! Ke3 (8...Ke4 9.a6! bxa6 10.b6! and the Black king is blocking the long diagonal) 9.a6! bxa6 10.b6! Bf3 11.h5! and wins 6.Kd4 6.Kf4 also wins, which is a small flaw. I chose the text as the main line since it transposes back to Reti's study. 6...Kg3 7.Ke3! Bg4 8.b5! which is back to Reti's main line 8...Kxh4 9. b6!
Continuing my Olympiad/Yearbook 128 series, the next survey on the list was by Robert Ris on the Fianchetto Benoni. While the variation examined by Ris was not played in the Olympiad, the game between Boris Markoja of Slovenia and Daniel Dardha of Belgium was an exciting game in another theoretically important line of the Fianchetto Benoni. 1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nc3 exd5 5.cxd5 d6 6.g3 g6 7.Bg2 Bg7 8.Nf3 00 9.00 Re8 10.Bf4
This position is getting a lot of scrutiny these days. The Quality Chess Grandmaster Repertoire series has basically become the main line of theory in recent years. There has been some back and forth in the Fianchetto Benoni. Initially, Boris Avrukh recommended 10.Nd2 for White in Volume 2. Marian Petrov came up with some resources for Black in Volume 12 on the Modern Benoni, so Avrukh changed his recommendation to the bishop move in Volume 1A. 10...Ne4 Ris' survey explored 10...h6 but no one in the Olympiad went for this, but the text is a very important line for the theory of this variation. 11.Nxe4 Rxe4 12.Nd2 Rxf4!?
Black sacrifices the exchange for a pawn, gaining the bishop pair and a strong queenside pawn majority in the process. However, White gets a strong central pawn position. Petrov prefers 12...Rb4 with the idea that after 13.a3 Rxf4 14.gxf4 Bxb2 White has to put his rook on an awkward square. 15.Ra2. Avrukh counters 12...Rb4 with 13.Rb1 which isn't examined by Petrov 13...g5!? 14.Be3 Bf5 15.a3 Rxb2 16.Rxb2 Bxb2 17.Qb3 is the main line from there when the engines favor White, but the two games I found in the database were both drawn. The position definately needs further practical tests. 13.gxf4 Bxb2 14.Rb1 Bg7 15.Nc4 Another crossroads where Petrov considers this position as better for White, whereas Avrukh prefers 15.e4 15...b6 Both Leko and Bologan played the immediate 15...Na6 here. That may be more flexible as Black might sometimes be able to play ...b5 in one go. 16.a4 16.Qa4 is more popular here, but there haven't been many games. 16...Na6 17.e4 The first new move. ProhaszkaPaschell, Budapest 2009 saw the players duck the fight after 17.Re1 Nb4 18.e4 Ba6 19.Qb3 ½½ 17...Nb4 18.e5 Ba6!? Sharpening the play. An alternative would be to precede this with 18...dxe5 19.Nxd6 Bxf1 20.Kxf1 Qh4
21.Qf3 21.Nxf7!? is an interesting sacrifice to get the central pawns rolling 21...Kxf7 22.e6+ Kg8 23.d6 looks scary for Black. 21...Rd8 22.Rd1 Rxd6 22...Bh6 feels more consistent. 23.exd6 Qd8 24.Re1 Bf6 25.Qe4 Qxd6
Again, time to take stock after an exchange sac. Black still has a pawn, but now he lacks the bishop pair and his queenside pawns aren't very mobile. The knight also seems a bit stranded on b4. On the plus side, the White pawns are very weak and Black's pawn structure is relatively solid. Overall, White should have the edge, but is it enough to win? 26.h3 Kg7 27.f5 a6 28.fxg6 hxg6 29.h4 Qh2 30.Re3 b5 grabbing the hpawn looks too dangerous as it lets the dpawn run free 30...Qxh4 31.Qxh4 Bxh4 32.d6 31.axb5 axb5 32.Qe8 Nc2 The coldblooded engine says Black should grab the pawn now 32...Qxh4 33.d6 (33.Rh3 Qc4+ 34.Kg1 Qc1+) 33...Qd4 with equality 33.Rg3 Bxh4 34.Rf3 Qc7 35.d6 Qa7 36.Qe5+ f6 37.Qe7+ Qxe7 38.dxe7 Kf7
39.Rc3 39.Rh3! provokes 39...g5 which fatally weakens the e8h5 diagonal 40.Rc3 Nb4 41.Re3 Ke8 42.Bf3 and White wins 39...Nb4 40.Rxc5 Kxe7 41.Be4 f5 42.Rxb5 Na6 Black has an immediate draw with 42...fxe4 43.Rxb4= 43.Rb7+ Kf6 44.Rb6+ Kg5 45.Bxf5 Black should also hold the draw after 45.Bc6 Nc5 46.Be8 Nd3 47.Rxg6+ Kf4 45...Kxf5 46.Rxa6
46...g5?
Dvoretsky
opens his section on Rook and Pawn vs. Bishop and Pawn with a blue section
(those are the most important ones) that begins
One should not protect the pawn by placing it on a square of the bishop's color.
Almost all of these postions are lost.
The current game is no exception.
Instead, Black should set up a barrier with 46...Bf6
when his fortress cannot be taken
47.Rh6 Kf4 48.Kg2 Kg4 49.Rh8 Kf4 50.Rf8+ Kg4 51.f3+ Kh5 52.Re8 Kg6 53.Kh3 Bf2 54.Kg4 Bc5 55.Re5 Bd4 56.Rxg5+ Kf6 57.Rb5 Be3 58.f4 Kg6 59.Kf3 10
Happy New Year! Looking at my backlog of BCE corrections, I should be able to keep posting a postion each Wednesday well into 2019, including many triple plays of backtobacktoback posts with a common theme. This week I'll complete the troika of positions featuring the first World Champion, Wilhelm Steinitz. Today's example comes from his 1866 match with Adolf Anderssen.
There is some debate if this was a world championship match. Fine calls it such, but the general consenus seems to be that the title did not come into being until 20 years later when Steinitz beat Zukertort. However, these were 2 of the top players of the day. In August 1866 when the match finished, Sonas has Steinitz as #2 and Anderssen as #4. The difficulty in rating players from that era, is that there weren't many international competitions. The rest of Sonas' list consists of players I have never heard of: #1 Berthold Suhle, #3 Cecil de Vere, #5 Hans von Minckwitz, #6 Philipp Hirschfeld and #7 Gustavus Reichhelm. Maybe I'll look into the games of these players at some time in the future. I did find that Shule drubbed Hirschfeld +7 0 =2 in an 1865 match, which may account for his top rated position.
The AnderssenSteinitz match was a crazy match by the standards of any era. After a game one loss, Steinitz came back with 4 straight wins. Undaunted, Anderssen hit back with 4 consecutive wins of his own. Then, Steinitz took back over winning 4 of the last 5 to take the match 86 with no draws!!
The BCE position comes from the 3rd game. After 40. Nb5 Steinitz launched a combination that led to a pawn up knight ending.
40...Re6 41.Qxg6+ Rxg6 42.Nxa7 Rxg2+ 43.Kh1 Rxh2+ 44.Kxh2 Nf3+ 45.Kg2 Nxe1+
46.Kf1! Nd3 Since Fine's defense relied on playing f3, I thought 46...Nf3!? was an interesting try to prevent that, but White is just in time with 47.Kg2 (47.Nc6 Kg6! 48.Kg2 Nh4+ 49.Kf1 Kf5 50.Ke2 Ke4) 47...Nh4+ 48.Kf1 f3 49.Ke1 Kg6 50.Kd2 Ng2 51.Kd3 Kg5 (51...Kf5 52.Nb5) 52.Nc6 Kh4 53.Nd4 Kxh3 54.Nxf3! and holds 47.Nc6 Kg6 48.Ke2 Nc5 49.Kf3 reaching the starting point of BCE129.
Incredibly, Fine had to defend this very ending a few years after the publication of BCE in his match with Miguel Najdorf in 1949. I hadn't previously known about this match between two of the top players in the Americas at the time. Fine won the first two games of the eight game match, but Najdorf won the next two and the final four were drawn. The third game was the only one in which Fine opened with 1. e4, but Najdorf did not try for his namesake variation and answered 1...e5. The knight ending was reached after 44. Kxe2
There are numerous conflicting stories about possible side wagers between the players on the outcome of this endgame. Don Miguel was a great storyteller, but there is probably truth somewhere in all the stories.44...Kf6 45.Nd6 Ke6 My database gave this move as 45...Ke5, which simply drops the f7 pawn. I checked some other databases online and they also had 45...Ke5. Finally, I found a printed game score in the February 1949 issue of the Swedish magazine Tidskrift för Schack which has the move 45...Ke6. I assume the move was incorrectly keyed into a database originally and the error has been copied ever since. I don't really know how to correct something like that. 46.Nc4+ Kd5 47.Nb2 Kd4 48.Nd3 f5 49.Nb4 Nf4+ 50.Kd2! 50.Kf2? Nd3+ is a winning pawn ending for Black 50...Ne6 51.Nc6+ Kd5 52.Ne7+! White's knight gets on a bad track after 52.Nb4+? Kc4! 53.Nc2 f4! now the knight can't go too far afield because of ...Nd4 so the White king eventually has to give ground 54.Ne1 (54.Na3+ Kb3 55.Nb5 (55.Nc2 Ng5 56.Ne1 Kc4! 57.Ke2 Ne6! 58.Kd2 g5 59.Kc2 h5 60.Kd2 Nd4 61.h3 (61.Kd1 Kc3) 61...h4 62.Kd1 Kc3 63.Kc1 Ne2+ 64.Kd1 Ng1) 55...Kb4) 54...g5 55.Kc2 52...Ke5 53.Ke3 Kf6 54.Nd5+ Kg5 55.Kf2 Kh4 56.Kg2 g6 57.Ne7 Kg5 58.Kg3 f4+ 59.Kf2 Nd4 60.Nd5 Nc6 61.Nc3 Kh4 62.Kg2 Ne7 63.Ne4 Nf5 64.h3?
Ironically, establishing the pawn formation he recommended for Anderssen turns out to be the losing move. Numerous annotators have pointed out that 64.Nf2 would allow White to hold the draw The Tidskrift för Schack version of the wager story is that Fine offered a draw here, which Najdorf countered with an offer of a $200 bet that was met with silence. 64...Ne3+! Heading for e1, this move gains a crucial tempo versus the other route 64...Nd4? 65.Nf2 Nc2 66.Nd3! and White holds 65.Kh2 Nc2 66.Kg2 Ne1+ 67.Kf2 Kxh3 68.Kxe1 Kg2! 69.Ke2 Edward Lasker's suggested 69.Nd6 sets the trap 69...Kxf3? (Black wins similar to the game by rolling the hpawn 69...h5) 70.Nf7 h5 (70...g5 71.Nxh6 Kg3 72.Nf7 g4 73.Ne5 f3 74.Nxg4) 71.Ne5+ Kg2 72.Nxg6 f3 73.Nh4+ Kg3 74.Nxf3 Kxf3 75.Kf1 = 69...h5 70.Ng5 h4! 71.Ne6 g5 01 Black wins the pawn ending after 72.Nxg5 h3 73.Nxh3 Kxh3 74.Kd3 Kg2! 74...Kg3?? 75.Ke4 + 75.Ke2 Kg3+