Shakmaty Bereolos - The Official Chess Site of Peter Bereolos

10/31/18 - BCE-45 Flohr-Ragozin, Moscow 1936

I'll complete the set of examples from the games of Salo Flohr with a pawn ending from his game versus Viachaslav Ragozin in the third international tournament held in the Soviet Union. Fine has the date as 1935, and while both players were also in that tournament, the BCE example is from the 1936 event. This one is a minor correction as one of the variations points to BCE-45a, which has an incorrect evaluation.

Flohr was unable to repeat his success from the 1935 event when he tied for first with Botvinnik. While he did finish clear third, his +1 score was well distant of Botvinnik at +6 and a dominant Capablanca who went undefeated +8 =10.

In the game versus Ragozin, Flohr was pressing, but Ragozin managed to hold a rook ending thanks to his active rook after 39...bxc4

40.bxc4 40.Rxf7+ Ke6 41.Ra7 (41.Rxh7 Rxd2 42.Rh6+ Kf5 43.Rxa6? c3!) 41...cxb3 42.Rxa6+ Kf5 43.axb3 Rxd2 is equal 40...Rxd2 41.Rxa6 41.Rxf7+ Ke6 42.Rxh7 Rxa2 again doesn't offer White much 41...Rc2 42.Rf6 Ke7 43.Rf4 g3

doubling White's pawns before capturing on a2. Ragozin awards the text an excamation point in the tournament book giving the following variation 43...h5 44.a4 Rc1+ 45.Rf1 Rxc4 46.Ra1 Rc6 47.a5 Ra6 48.Kf2 Ke6 49.Ke3 Ke5 50.g3

with the conclusion и постепенно белые завоевывают пространство королем, which Google translates as and gradually White conquers the king. I'm not sure if the end of that should translate as White wins or something more along the lines of White invades the kingside. In any case, the implication is that White is winning, and at first blush it would seem that he has everything going for him with rook behind the pawn and an apparent path to enter with his king. Yet, even here it looks like Black can keep the balance. 50...f5 51.Ra4 f4+ 52.gxf4+ Kf5 53.Ra2 h4 54.Ra4 Re6+ 55.Kd3 Ra6

and White does not seem to have a way to make progress. For example, 56.Kc4 g3 57.hxg3 hxg3 58.Ra2 Kxf4 with a draw 44.hxg3 Rxa2 45.g4 Rc2 46.Kh2 f6 47.Re4+ Kd6 48.Kg3 Rc1 49.Kf4 Rf1+ 50.Ke3 Rg1 51.Kf2 Rc1 52.Kg3 Kd7 Ragozin suggests 52...Rd1 as simpler; 52...Rc2 tying White to the g2 pawn also looks solid 53.Kf4 Rf1+ 54.Ke3 Rg1 55.Rd4+ Ke6 56.Rd2 Ke5 57.Rc2 57.Kd3!? to activate the king 57...Kd6 58.Kd4 Rc1 59.Rd3 Rg1 60.Rf3 (60.Rd2 Rf1) 60...Rxg2 61.Rxf6+= 57...Re1+ 58.Kf2 Ra1 Ragozin awards this an exclam, but 58...Rb1 or 58...Rd1 will come to the same thing 59.c5 Ra7 60.c6 Rc7 61.Ke3 Kd6 62.Kf4 Rxc6! 63.Rxc6+ Kxc6! 64.Kf5 Kd5! 65.Kxf6 The BCE-45 position 65...Ke4! 66.Kg5 Ke3 67.Kh6 Kf4 68.g5 Kf5 69.Kh5 Kf4 70.Kh4 Ke3 71.Kg3 Ke4 72.Kg4 Ke3 73.Kf5 Kf2! 74.Kf6 Kg3! 75.Kg7 Kg4 76.Kh6 Kh4 1/2-1/2

10/30/18 - Duda-Karjakin, 2018 Batumi Olympiad

My next selection in my Olympiad/NIC Yearbook series comes from the Round 4 match between Poland and Russia. This was the match that really announced Poland as a contender as they upset the highly ranked Russian team. However, the match didn't start that great for Poland as Sergey Karjakin held an easy draw with Black on board 1 against Jan-Krzysztof Duda. 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 d5 5.cxd5 exd5 6.Bg5 h6 7.Bh4 c5

Jan Timman's survey in Yearbook 128 mainly focused on the newer 7...g5 8.Bg3 Nc6!? but no one went for this riskier approach in the Olympiad. 8.dxc5 g5 While it had been played a few times before, this move really came into fashion when Nigel Short used it in his 1993 World Championship match with Garry Kasparov. 9.Bg3 Ne4 10.e3 Qa5 11.Nge2 In the fifth game of the match Kasparov played 11.Be5 but Short equalized with 11...0-0 and the game ended soon thereafter in a draw with Short only consuming 11 minutes for the entire game. The text was Kasparov's choice when Short repeated the variation in game 9. 11...Bf5 12.Bxb8 Kasparov's 12.Be5 remains the main choice here, but capturing the knight scores about the same in the database. 12...Rxb8 13.Nd4 Bd7 14.Nb3 The other line here is 14.Bd3 when no Black player has entered the complications of 14...Nxc3 (since 14...Ba4; or 14...Bxc3+ are simpler paths to equality) 15.0-0 14...Bxc3+ 15.bxc3 Qxc3+ 16.Qxc3 Nxc3 17.Bd3 Ke7 18.Kd2 Na4 19.Rhc1

The first new move 19.Rac1 and 19.h4 had been tried before. The engines have it all as equal. 19...Rhc8 20.Rc2 Rc7 21.Rac1 Rbc8 22.f3 h5 23.Na5 Nxc5 24.Nxb7 Rxb7 25.Rxc5 Rxc5 26.Rxc5 Rb2+ 27.Rc2 Rxc2+ 28.Bxc2

The only weakness in the position is the isolated black d-pawn on the same color as the bishops. However, Black easily defends it. The two players put all their other pawns on dark squares and soon agree to repeat moves. 28...Kd6 29.Kc3 Kc5 30.g3 h4 31.gxh4 gxh4 32.a3 a5 33.f4 f6 34.Bd3 Bg4 35.Bc2 Bd7 36.Bd3 Bg4 37.Bc2 1/2-1/2 It looks like it is back to the drawing board for White in this line.

10/28/18 - Bereolos-Gurevich, 1992 Indiana State Championship

I've added my 1992 game against Dmitry Gurevich to the GM games section played in the final round of the 1992 Indiana State Championship. The 51st edition was a bit unusual as the organizers put up a big prize fund, which attracted a lot of strong out of state players. There were 17 masters including 7 players over 2400.

Despite the strong field, I still had a chance to retain the title I had won the previous year as I was a half point clear of the other in state participants. However, I got an awful position out of the opening against Dmitry. I sacrificed the exchange to get some lines for my bishops, but he returned it to reach a winning major piece ending a pawn up. He transitioned to a rook ending and I did have some chances to resist, but did not find the toughest defense and went down.

10/26/18 - Svetushkin - Kosic, 2018 Batumi Olympiad

Back to e4 with a survey by Alexandr Predke on the Steinitz Variation of the French and an exciting game between Dmitry Svetushkin and Dragan Kosic on board 3 of the Moldova-Montenegro match 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.e5 Nfd7 5.f4 c5 6.Nf3 Nc6 7.Be3 Be7 8.Qd2 0-0 9.Be2

9..a6 9...b6 10.h4!? was the main focus of Predke's survey. He does discuss the pros and cons of b6 versus a6 mentioning that after 9...a6, White will likely castle kingside as 10. 0-0-0 runs into 10...c4 followed by a pawn storm. 10.0-0 b5 11.Nd1 11.a3 along with the text are the two major branches here. 11...cxd4 11...b4 is the main line here 12.Nxd4 Bb7 Now we are in much rarer territory with the main move being 12...Nxd4 although there is a Morozevich game with the bishop move 13.c3 13.Nxc6 Bxc6 14.Bd4 f5 15.exf6 Nxf6 16.Qe3 Bd6 17.Qxe6+ Kh8 18.Ne3 Qc7 19.Qf5 Ne4 20.Qh5 Bxf4 21.Bd3 Rae8 22.Rae1 Re6 23.g3 Bxe3+ 24.Bxe3 Rxf1+ 25.Rxf1 Nf6 26.Bd4 Qe7 27.Qh4 Re1 28.c3 Rxf1+ 29.Kxf1 Qf7 30.Bxf6 gxf6 31.Qf4 Qe7 32.Qh6 Kg8 33.Qh4 Be8 34.Qg4+ Kf8 35.Kf2 h5 36.Qd4 Qe5 37.Qc5+ Kg8 38.Qa7 h4 39.Qh7+ Kf8 40.Qxh4 a5 41.Qf4 b4 42.Qh6+ Ke7 43.Qg7+ Kd6 44.Qf8+ Kc6 45.Bc2 bxc3 46.bxc3 Kb6 47.Bd3 d4 48.c4 Qe3+ 49.Kg2 Bc6+ 50.Kh3 Qxd3 51.Qb8+ Ka6 52.Qc8+ Bb7 53.Qe6+ Ka7 54.c5 Qh7+ 55.Kg4 Qg6+ 0-1 (55) Paravyan,D (2528)-Morozevich,A (2683) Sochi 2016] 13...Rc8 14.Nf2 Nc5 15.Nxc6 Bxc6 16.Bd4 Bd7

The first new move, there had been one previous game with 16...Qc7 from d7 the bishop can help defend against the f5 break. 17.Qe3 This seems to be the choice of all the engines giving White about a half pawn plus, which seems about right because of the extra space and Black's bad bishop. White's last move covers e4 and it looks like he has the plan b4 and Bd3 pointing everybody at the Black king 17...Qc7 18.b4 Na4

19.Ng4 Ra1 didn't seem to be doing much, so I don't know why White didn't cover his pawn with 19.Rac1 19...Nxc3 20.Bd3 White's intention may have been 20.Nf6+ but Black seems to hold with 20...Bxf6 (20...gxf6 21.exf6 Nxe2+ 22.Qxe2 Qc2 23.Qxc2 Rxc2 24.fxe7 Re8 25.Bf6 offfers White the better chances) 21.exf6 Nxe2+ 22.Qxe2 Qc2 20...Ne4 The black knight didn't have great prospects, so Black gives back the pawn to exchange off White's attacking light-squared bishop. 20...Bxb4 looks a bit too greedy 21.Bxh7+ Kxh7 22.Qh3+ Kg8 23.Nf6+ gxf6 24.exf6 with mate soon to follow 21.Bxe4 dxe4 22.a3 Kh8 23.Rad1 White is in no hurry to capture e4, which would open the long diagonal for Black's now unopposed light-squared bishop. 23...Qb7 24.Qg3 Rfd8 24...e3 is positionally desirable, but Black prefers to not waste a tempo. The pawn can become valuable later. 25.Ne3 Bf8 26.Qh4 Bc6 27.f5 Bd5 28.f6 g6 29.Bc5

29...Rxc5 Black must maintain his dark-square bishop to defend his king 29...Bxc5 30.bxc5 Rxc5 31.Rf4 g5 (else 32.Qh6 Rg8 33.Qxh7+ Kxh7 34.Rh4#) 32.Qxg5 Rg8 33.Qh6 Rg6 (other moves are met by 34.Rh4 mating on h7) 34.Qf8+ picks up Rc5 30.bxc5 Bxc5 31.Qh6 Bf8 32.Qh3 Qc7 33.Qg3 Qc5 34.Qf4 Qb6 35.Kh1 a5 35...Bxa3? 36.Nxd5 exd5 37.Rxd5 with the point 37...Rxd5? 38.Qh6 and Black no longer has ...Bf8 36.h4 b4 37.axb4 axb4 38.h5 gxh5 39.g4 h4 40.Rd2 Perhaps a bit premature. 40.g5 looks better. White can decide how best to deploy his rooks later 40...Bc5?! 40...Bc6 41.Rxd8 Qxd8 and White's attacking prospects are greatly diminished, while Black's passed pawns are very dangerous 41.Nxd5 exd5

42.Qg5 since the Black bishop left f8, White again had the trick 42.Rxd5 42...Bf8 43.Rh2 Qa6 44.Qf5 Qd3 45.Rxh4 h6

46.Qf4 Again, the direct 46.g5 looks much more dangerous 46...e3 now Black's queen can defend along the b1-h7 diagonal 47.Rh3 d4 48.g5 b3 49.Rg1 Qg6 49...b2 50.g6 threatening Rxh6+ with a mating attack 50.Rh2 Rc8 51.e6 fxe6 52.gxh6 Qf5 53.h7 Rc1 54.Qg3 Qe4+ 55.Rhg2 Rxg1+ 56.Kxg1 Qb1+ 57.Kh2 Qxh7+ 58.Kg1 Qb1+ 59.Kh2 Qh7+ 60.Kg1 Qb1+ 61.Kh2 Qh7+ 1/2-1/2

10/24/18 - BCE-161 Thomas-Flohr, Hastings 1935/36

This week's BCE correction is from another game by Salo Flohr, this time against Englishman George Thomas at the annual Hastings tournament 1935/36. Flohr had many successes at Hastings, winning outright in 31/32, 32/33, and 33/34 before tying for first with Euwe and Thomas in 34/35 (but ahead of both Capablance and Botvinnik). In 35/36 he finished second by a point to Fine, who defeated Flohr in their individual encounter.

The BCE example is a bit of a mess. Fine cites it as Flohr-Thomas, reversing the colors. He does say that Flohr sacrificed a knight, which is true, but reads strangely if Flohr was White since White is a piece up. He follows the game continuation, but at the end has Black resigning instead of White. There also might have been one further move played. Then, to top it all off he includes analysis of the position if it were White to move instead of Black and makes another mistake there. Benko did add the drawing line for Thomas in the revised edition and eliminated the White to move analysis. However, he left all the other mistakes in the text, so it really didn't fix the example.

Here's the game from a few moves before BCE. It was kind of a race in a B vs. N ending after 43.Kxe5

43...Kb4 44.Kd6 Kxb3 45.Kxc6 Kc4 46.Kd6 Kd3 47.Bc1 Kxd4 48.c6 Kc4 48...a5 49.g4 Ke4 50.g5 Nd4 51.c7 Nb5+ 52.Kc6 Nxc7 53.Kxc7 is also a draw 49.c7 Nxc7 50.Kxc7 The starting position for BCE-161 50...d4 51.Kd6? 51. Bd2! is the missed draw 51...a5 52.g4 a4 53.g5 d3 54.Ke7 a3 55.Bxa3 d2 56.Kxf7 d1Q 57.Be7 Kd5 58.g6 Qf3+ 59.Bf6 Qf5 60.g7 Qe6+ Fine ends here with and Black resigned. Of course it is White who is resigning. My database continues one more move. 61.Kg6 Qg8 0-1

10/23/18 - Fuller-Bereolos, 2013 Louisville Open

Yet another example with R+2 connected passers vs. R+1 comes from my game against William Fuller in the 2013 Louisville Open. It is amazing how this type of ending can be so difficult and subtle. We reached the position of interest after 51.Rxh6

Thanks to his active pieces, White should be able to hold. 51...Rg4+ 52.Kd5 this is probably best 52.Kd3? offers the least resistance, Black can quickly achieve his ideal position with his king supporting the pawns while the rook controls White's pawn from behind. 52...Rf4 53.Rh5 Kf6 54.Rh6+ Ke5 55.Re6+ Kd5 56.Rf6 c4+ 57.Ke3 Re4+ 58.Kd2 (58.Kf3 c3) 58...Rd4+ 59.Kc2 Kc5 and Black will methodically advance his pawns; 52.Ke3? is a much tougher nut to crack, but Black can still win 52...Rd4! 53.Rg6 (53.Re6+ Kd7! 54.Rf6 Kc6 55.Rf8 Rd1 56.Ke2 Rd5 57.f6 Rf5 58.f7 Kd5 59.Kd3 Ke6) 53...Rd1! 54.Ke4 Kd7 55.Rg7+ Kc6! 56.f6 d5+! (56...Rf1? 57.f7! and Black can't make progress 57...c4 (57...d5+ 58.Ke5! and White is ready to build a bridge with Rg6+ and Rf6) 58.Kd4 d5 59.Rh7 Kd6 60.f8Q+ Rxf8 61.Rh6+ Ke7 62.Kxd5=) 57.Ke5 Re1+! 58.Kf5 Rf1+! 59.Ke5 d4 60.f7 d3! (60...Kb5? 61.Kd5 d3 62.f8Q Rxf8! 63.Rb7+! Ka6 64.Rb1 holds) 61.Rg6+ Kb5! 62.Rf6 Rxf6! 63.Kxf6 d2! 64.f8Q d1Q! and Q+c vs. Q is winning for Black.; 52.Kf3 also draws, but White has to play accurately 52...Rd4 53.Re6+! Kf7 (53...Kd7 54.Re1! and White will create counterplay with his f-pawn) 54.Ke3 Rd1 55.Kf4 c4 56.Rh6! c3 (56...d5 57.Ke5 d4 58.Rh7+ Kf8 59.Rc7 c3 60.Rc8+ Kf7 61.Rc7+ Ke8 62.Rc4 Rd2 63.f6 c2 64.Rc8+! Kf7 65.Rc7+! Kg6 66.Rg7+! Kh6 67.Rg1! d3 (Black even loses after 67...Rd1? 68.f7!) 68.Ke4) 57.Rh7+! Kf6 58.Rc7! Rd3 59.Ke4 Rh3 60.Rc6 Ke7 61.f6+ 52...Rd4+ 53.Kc6 c4

54.Re6+? 54.Kb5 is still a draw 54...d5 (54...c3 55.Rh7+ Kf6 56.Rc7) 55.Rc6 Rd3 56.f6+ Kf7 57.Kb4 Kg6 58.Rd6! c3 59.Kb3! d4 60.Kc2! Rd2+ 61.Kc1 d3 62.Rc6! Rc2+ 63.Kd1! Kf7 64.Rd6! and without his King, Black can't make progress. 54...Kf7 55.Kb5 c3 now the White King is cut off and his rook can no longer attack the pawns from behind, so Black is winning. 56.Re3 c2 57.Rc3 Rd2 57...Rd5+ 58.Kb4 Rc5-+ is simpler 58.Kb4 Kf6 59.Kb3 Kxf5 60.Kb2 Ke4 61.Kc1 Re2 62.Rc4+ Kd3 63.Rc3+ Kd4 and Black won 0-1

The primary lesson from this ending is whether on the attacking or defending side, the king needs to participate. When the White king got cut off in the game, the win became relatively easy. In a couple of the drawing lines, we saw Black unable to make progress because his king was cut off.

10/20/18 - Piorun-Kryvoruchko, 2018 Batumi Olympiad

The first survey on the closed games in Yearbook 128 was by Jose Vilela in the Ne5 variation of the open Catalan. The exact line he looked at did not appear in any Olympiad games, but I've chosen a game in that variation that had an interesting ending and was also part of a prominent match.

After 5 rounds of the Olympiad, only 4 undefeated teams remained: Azerbaijan, Ukraine, Poland, and Czech Republic. The game I'm going to look at today was between Kacper Piorun and Yuriy Kryvoruchko on board 3 of the 6th round match between Poland and Ukraine.

1.d4 Nf6 2.c4 e6 3.g3 d5 4.Nf3 Be7 5.Bg2 0-0 6.0-0 dxc4 7.Ne5 Nc6

The topic of Vilela's survey was 7...Qd6 which Mamedyarov tried against Kramnik in a must win situation in the final round of this year's Candidates tournament. Almost all the games in the Olympiad continued with the text except for one with 7...c5 8.Bxc6 8.Nxc6 was more popular at the Olympiad played in 4 of the 6 games that reached this position. The bishop capture is recommended by Hilton and Ippolito in the first volume of their books on Alexander Wojtkiewicz's repertoire. My database shows Wojo at +5 =1 -0 from this position. 8...bxc6 9.Nxc6 Qe8 10.Nxe7+ Qxe7 11.Qc2 Wojo preferred 11.Qa4 which gives White the aditional option after 11...c5 of 12.Qa3 11...c5 12.Qxc4 cxd4 13.Qxd4 White is a pawn up, but has some light square weaknesses around his king 13...e5 14.Qh4 Qb7 The only game Wojo drew in this line was in the 2005 US Championship against Yuri Shulman who played 14...Rb8 15.Nc3 Nd5

This looks like the first new move, but I don't think it was preparation since I didn't find any other games by Piorun in this line and Kryvoruchko deviates from the top engine line on the very next move. In the most recent high level game in this line, Adams preceded the text with 15...Be6 against Radjabov in the 2017 Gashimov Memorial 15...Rd8 has also been played. Both of these alternatives have scored 50% in the database. 16.Ne4 The engines prefer 16.Bd2 by a small amount, but then Black can regain his pawn 16...Qxb2 17.Nxd5 Qxd2 18.Qe4 with only a small pull for White. However, the text looks like a natural reaction as the knight eyes the squares c5, d6, and g5 16...Qa6 the queen looks a bit awkward on a6 as the two natural squares to develop the bishop to are b7 and e6, both of which can now be hit with Nc5 trading N for B [The engines like 16...Qc6 which prepares a battery on the long diagonal with Bb7 and also keeps an eye on e4 and f6, which puts ...Nf4 ideas in the air; the immediate 16...Nf4 is well met by 17.Nf6+ 17.Rd1 Bb7 18.Nc5 Qc6 19.Nxb7 Qxb7 20.Qe4 with B vs. N and an extra pawn, White is clearly better.

20...Rad8 21.Rd3 Qb4 22.Qxb4 Nxb4 23.Rd2 f6 24.a3 Nc6 25.b4 Nd4 26.Kg2 a5 27.bxa5 Nb3 28.Rxd8 Rxd8 29.Rb1 Nxa5 30.Be3 Rc8 31.Bd2 Ra8 32.Bb4 Nc6 33.Bc5 Ra5 34.Rc1 Ra6 35.Kf3 Kf7 36.Rc4 Ke6 37.e4 Ra8 38.Ke3 Rb8 39.Rc3 Rb2 40.Rd3 Kf7 41.Rd7+ Kg8 42.Kd3 perhaps White should start with 42.h4 getting a pawn off of the second rank. Now, Black finds a way to create some problems.

42...Nd4 43.Bxd4 exd4 44.a4 Rxf2 45.Rxd4 Rxh2 46.Kc4 Kf7 47.a5 Ke6 48.a6 Ke5

49.Kb3! This looks odd, but seems to be the only way to win 49.Rd5+ Kxe4! 50.Ra5 Rc2+! 51.Kb5! Rc8 52.a7 (52.Kb6 Kf3 53.Ra3+ Kg4! 54.Kb7 Re8 55.a7 g5) 52...Ra8 53.Kb6 Kf3=; 49.a7 Ra2! 50.Rd7! Kxe4 51.Rxg7 h5!= 49...Kxd4 50.a7 Rh3 51.a8Q Rxg3+ 52.Kc2 Rg5

This type of queen versus rook ending is similar to those I looked at in May. Black doesn't quite have a fortress and White still has a pawn remaining. Nevertheless, it is a very tricky ending to play for both sides. The main method in May was to cross the king over to attack the pawns from behind. The presence of the White pawn gives rise to some additional possibilities such as sacrificing the queen for the rook to breakthrough and queen the pawn or exchanging the pawn in a manner that leaves Black with split pawns. 53.Kd2 Ke5 54.Qc6 A variation that would engineer a favorable pawn exchange for White was 54.Qd5+ Kf4 55.Qd4 Kf3 (55...Re5 56.Qe3+ Kg4 57.Qg1+; 55...Kg3 56.e5) 56.Qe3+ Kg4 (56...Kg2 57.Qxg5+) 57.e5 Rxe5 (57...fxe5 58.Qe4+) 58.Qg1+ 54...Kf4 55.Kd3 h5 56.Qd6+ Kg4 57.e5?

Now Black ends up with 2 connected pawns and should be able to hold the draw. White should have continued 57.Qh2 or 57.Ke3 in order to squeeze the Black king. I'm not going to go too in depth over the rest of the game since the ideas were covered in my May post. I imagine the players were down to the increment here, so I'll mostly just point out blunders and only moves. 57...fxe5 57...Rxe5? 58.Qd7+! and White should win against the split pawns 58.Ke4 Kh3 59.Qd3+ Rg3? 59...Kh2! is one of those hard to find moves in time pressure. The fast glance would reject that move because of 60.Qd2+ but Black holds after 60...Rg2! 60.Qf1+! Kh4 61.Qe1 Kh3 62.Qh1+ Kg4 63.Qc1? 63.Qd1+! 63...h4 64.Qc8+ Kg5 65.Qh8 g6 66.Qxe5+ Kh6? 66...Kg4! if the Black king can get to h3 he will find shelter. 67.Qf6? 67.Qh8+! Kg5 68.Qd8+ Kg4 69.Qd7+! Kg5 70.Qd6 stops the king from advancing and White can start bringing his king forward after 70...Kh5 71.Ke5! 67...Kh5! 68.Qh8+ Kg5? 69.Qf8? Kh5 70.Qe8 Kg5? 71.Qc6? Kh5? 72.Qa8? Here, and over the next few moves White again misses opportunities to advance his king 72.Ke5! 72...Rg4+? 73.Kf3? 73.Ke5! 73...Rg3+? The only drawing move was 73...Rg1! which again is hard to find because of fear of leaving the rook on an unguarded square. 74.Kf2? 74.Kf4! 74...Kg5 breaking the streak of 8 straight moves that changed the evaluation back and forth from win to draw. 75.Qh8 Kg4 76.Qf6 Kh3 finally the Black king has reached a safe haven and he is able to advance his pawns far enough to secure the draw. 77.Qe6+ Kh2 78.Qd5 g5 79.Qe5 Kh1 80.Qa1+ Kh2 81.Qe5 Kh1 82.Qe1+ Kh2 83.Qe4 Rg1 84.Qe5+ Rg3 85.Qb8 Kh1 86.Qxg3 1/2-1/2

10/17/18 - BCE-576 Flohr-Keres, Kemeri 1937

The third member of the winning triumvirate at Kemeri 1937 was Salo Flohr. This was a name I knew, but I didn't know how strong he really was. Sonas ranks him #2 in the world behind Alekhine for almost a full year in 1936 and 1937. He drew matches with both Euwe and Botvinnik and was poised to be Alekhine's challenger in 1937 when world events began to intercede. After finishing last at the AVRO tournament in 1938 he remained one of the world's top players, but was never again a contender for the highest title.

His first round game at Kemeri against Paul Keres is one of the more complicated BCE corrections. On my correction pages I put the evaluation of the position before the diagram. This is the first case where I have been unable to make a definite conclusion, so I have put Draw? as the evaluation. This is an incredibly hard position to reach an absolute assessment on because both kings have no place to hide. The position has too many pieces for the tablebases to solve. Engines suffer from the horizon effect here because of all the possible checks by both sides on many different squares. There are also captures on e4, g4, h2, and e3 that reset the clock on the fifty move rule. I tried various attempts against the engine, but always ended up eventually getting to 0.00. However, letting the engine ponder for a long time from the starting position, its evaluation hovers a bit above 1, which is usually the sign that the position is a draw, but the machine can't go deep enough to drop the evaluation to 0.00.

Fine follows the game continuation, and shows that Flohr missed a win. However, he concludes two earlier notes as reaching winning positions for White, when in both Black clearly has perpetual check. The first variation can be fixed, but the second one remains uncracked, so I think the book is reopened on the evaluation of this ending.

10/15/18 - Bereolos-Georgiev, 2006 Spring North American FIDE Invitational

I've added my 2006 game against GM Vladimir Georgiev to the GM games section. I've only added a few additional comments on the endgame to my original notes from my tournament report on the 2006 Spring North American FIDE Invitational. That tournament remains the strongest that I've played in. The game against Georgiev was typical of my games in that tournament, which was the first I ever played with the FIDE time control. I got a decent position, but then didn't cope well once I got into the perpetual time pressure with the 30 second increment, and finally made a terrible howler that lost outright.

10/10/18 - BCE-64b Bergs-Petrovs, Kemeri 1937

One of the other co-winners at Kemeri 1937 was Latvian Grandmaster Vladimir Petrovs. Earlier this week Chessbase had a feature on his life, so this week's BCE post is timely. Kemeri 1937 was the peak of his chess career which got upended by the events leading to World War II.

Petrovs' 10th round game with Black against his countryman Teodors Bergs is the subject of this post. (Fine cites this game as Berg-Petrov, but most sources I looked at have an s at the end of both players names, so I'm going with that. After 38...Rcc1

39.a4 White could have immediately played the deflection he employs on the next move 39.Rxe4 Rb1+ 40.Kc2 Rxe4 41.Kxb1 Rxh4 with equality 39...e3 40.Rxe3 Rb1+ 41.Ka2 Rxe3 42.Kxb1 Rxb3+ 43.Rb2? White holds with 43.Kc2 Ra3 44.Rd5 Rxa4 45.Kb3 Ra3+ 46.Kxb4 Rg3 47.Rd6 Rxg2 48.Rxa6 Rg4+ 49.Kc3 Rxh4 50.Kd3 My database ends the game here. 0-1 Some sources carry it another move or two. 43...Rxb2+! 44.Kxb2 a5! reaches the starting BCE-64b position. Apparently, the game was adjourned and Bergs resigned. The ending is certainly tricky, so it would have been reasonable to play on.

10/6/18 - Fawzy-Cheparinov, 2018 Batumi Olympiad

The Olympiad is in the books with a 3-way tie for first between China, the US, and Russia. The medals went in that order by the Olympiad tiebreak system. By game points the US would have come out on top thanks to Li Chao's first round loss.

I had mentioned that I was trying to think of a way to look at some opening theory at the Olympiad. The method I've come up with is to look at games played in variations discussed in the most recent New In Chess Yearbook. New In Chess publishes the Yearbook every three months. Each issue features a number of opening surveys. Volume 128 came out shortly before the Olympiad and featured 26 surveys. I thought it would be interesting to see how topical this issue was by looking at games featuring the openings in question. I'll look at them in a different order than in the book to give some variety from post to post. I'll alternate between open and closed games and jump from opening to opening within each.

The first survey was by former FIDE champ Ruslan Ponomariov looking at an alternative in the Bg5 Najdorf 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bg5 Nbd7

This move was popular in the 50's and 60's with such players as Petrosian, Polugaevsky, and Tal using it. There was also a famous win by American GM William Lombardy with it versus Spassky in the World Student Olympiad. It has undergone a revival in the past 10 years as Black seeks to stay flexible and avoid the labyrinth of theory after 6...e6. Black maintains the option of the classical Najdorf move ...e5 as well as possibilities of transposing to Dragon positions with ...g6. Ukrainian GM Anton Korobov seems to be the biggest proponent of the move. In the Olympiad it was played 8 times with an even score +2 =4 -2.

Ponomariov does discuss the old main move 7. Bc4, which was seen in a couple of games in the Olympiad, as were 7. f4, 7. Qd3 and 7. Be2. However, the focus of the survey is on the move 7. Qe2 which is awarded an exclam in Negi's e4 repertoire book. (Ponomariov recommends this book as well as Kevin Goh Wei Ming's book on the Bg5 Najdorf for further study of this variation). The move was played a couple of times in the Olympiad by Egyptian IM Adham Fawzy. The game I'm going to look at was in round 10 against long-time Bulgarian #2 Ivan Cheparinov who is now representing Georgia and played board 3 for the host country in Batumi. Cheparinov long served as Veselin Topalov's second and is well known for being prepared in sharp theoretical lines. 7...h6 8.Bh4 g6 9.f4 e5 Cheparinov plays the most principled move 9...Qc7 was Svidler's choice against Carlsen in Biel earlier this year. 10.fxe5 10.Nf3 is another alternative. It scores about the same in the database and gets Ming's approval. Negi prefers the text. 10...dxe5 11.0-0-0 This tactical shot was introduced by Hou Yifan against Vishy Anand at Wijk aan Zee in 2013. 11...Qc7 Black gets smashed after 11...exd4? 12.e5; Anand played 11...Be7 when both Negi and Ming suggest 12.Nb3 as an improvement on Hou's 12.Nf3 12.Nb3 b5 In the final round match between Egypt and Peru, Fawzy-Martinez saw 12...Bg7 but White already had a nice position after 13.Qc4 and went on to win.13.Rxd7!?

The primary games in the survey look at Negi's recommended 13.Nd5 but Ponomariov also covers the text, calling it a risk-free way for White to play. Cheparinov had faced 13.Qf3 earlier this year in the Chinese team championship winning against Yunguo Wan. 13...Nxd7 14.Nd5 Qd6 Ponomariov considers the best play in this position to be 14...Qb8 15.Bf6 Nxf6 16.Nxf6+ Ke7 17.Nd5+ Ke8 18.Nf6+ with a draw by perpetual check, but a short draw is not what Black wants in a team tournament where he outrates his opponent by almost 300 points. 15.Na5 Ra7 this seems to be a novelty, with Black ready to give back the exchange on c7. Ponomariov only looked at 15...g5 16.Qf2 this move looks logical hitting the rook, eyeing f7 and clearing the diagonal for Bf1 to move followed by Rf1. The engine prefers 16.Qf3 keeping the diagonal open for Bh4-e1-b4 16...Rc7 17.Be2 Qc5 18.Bg4 g5 18...Qxf2?? 19.Nxc7# would be an abrupt finish 19.Rf1 Qxf2 19...gxh4 20.Nxc7+ Qxc7 21.Qxf7+ Kd8 22.Bxd7 Qxd7 (22...Bxd7 23.Qf6+) 23.Rd1 Bd6 24.Qf6+ and White emerges the exchange and a pawn ahead 20.Nxc7+ Kd8 21.Bxf2 Kxc7 22.Bg3 f6 Black can't cover both the e and f pawns, so he forces White to exchange on d7 hoping that the bishop pair will provide compensation. 23.Bxd7 Kxd7 24.Rxf6 Bd6 25.c4 h5 26.b4 h4 playing to force the exchange of White's active rook. However, the line 26...Bxb4 27.Nc6 Rf8 looks tenable for Black 27.Be1 Rf8 28.Rxf8 Bxf8

29.Kc2 White may have feared that Black could construct a fortress after 29.c5 However, now the game opens up a little more for the bishops. 29...bxc4 30.Nxc4 Ke6 31.Bc3 Bd7 32.h3 White is trying to play patiently since the weakness on e5 isn't going anywhere but after the text the White pieces get a bit tangled, so perhaps he should have played 32.Nxe5 Ba4+ 33.Kd3 Bxb4 34.Nf3 32...Bc6 33.Nd2 Bb5 34.Kd1 White's king goes to cover f1 in order to free up the knight. 34...Be7 35.Ke1 Bd8 36.Kf2 Bb6+ 37.Kf3

37...Ba4?! It's always a difficult decision to part with one of the bishops in this kind of ending, but 37...Bd4! 38.Bxd4 exd4 seems to solve all of Black's problems 38.Kg4 Be3 39.Nf3 If White carelessly hits the bishop with 39.Nc4?? or 39.Nf1?? he is mated by 39...Bd1# 39...Bd1 40.Bxe5 Bd2 41.Bd4 Be2 41...Bxb4 42.Be3 wins both of Black's kingside pawns 42.a3 Bd1 43.Kh5 Bc1 44.a4 Be2 Again, Black loses his kingside after 44...Bxa4 45.Nxg5+ 45.Kg6 g4 46.hxg4 h3 47.Bg1! stopping the pawn and preventing 47...hxg2 because of 48. Nd4+ 47...Ke7 48.Nd4 Bxg4 49.gxh3 Bxh3 50.Nf5+ Ke6 51.Bd4 Bd2 52.b5 axb5 53.axb5 Bf1 54.b6 Ba6

55.Ng7+! Kd6 56.Kf6? White could immediately roll the e-pawn home with 56.e5+! Ke7 (56...Kd5 57.e6 Kd6 58.Kf6) 57.Bc5+ 56...Kc6 57.e5 Bc4? 57...Ba5! 58.e6 Bc8! 59.e7 Bd7 and Black is ready to take the b-pawn 58.Nf5? White can again win with the direct 58.e6 but the variation is a little tricky to calculate since White must sacrifice his knight back with check to achieve a winning B+P vs. B ending. 58...Bxe6 59.Nxe6! Ba5 60.Nd8+! Kb5 61.b7! Bxd8+ 62.Ke6! Bc7 63.Kd7! Bh2 64.Kc8 Kc6 65.Ba7 Be5 66.Bb8 Bd4 67.Bf4 Ba7 68.Be3; 58.Ne6 preparing Nd8+ is a less calculation intensive way to win. 58...Ba5! 59.Nd6 Ba6 59...Bd5 looks like a surer way to draw, keeping an eye on the e6 square 60.b7 (60.Nc8 Kb7 61.Ne7 Bc4 and 62...Bxb6 ) 60...Bd8+ 61.Kg7 Bc7 62.Kf8 Kd7 60.e6 This seems to be the best chance 60.Nf7 Bxb6; 60.b7 Bc7; 60.Be3 Bc3; and 60.Kg7 Kd5 don't make any progress

60...Kxd6? 60...Bxb6! 61.e7 Kxd6! 62.e8Q Bxd4+! 63.Kg5 and the tablebase proudly annouces Mate in 70. However, with best play by Black it takes 60 moves before White can win one of the bishops so the game would be drawn before then by the 50 move rule!! I wonder what the time situation was here. The winning variation for White is not difficult to calculate after 60...Kxd6, so by process of elimination it would seem that Black should go for the Q vs. 2B position. But perhaps he used process of elimination the other way and knew 2B vs. Q is losing so tried 60...Kd6 instead. 61.e7? The final slip. White wins with 61.Be5+! Kc6 62.e7! Kd7 63.Kf7! Bc4+ 64.Kf8! Bb4 65.b7! 61...Kd7! 1/2-1/2 Now the pawn is stopped after 62.Kf7 Bc4+ 63.Kf8 Bb4 64.b7 Bxe7+ 65.Kg7 Bd6

10/3/18 - BCE-352a, Reshevsky-Apscheneek, Kemeri 1937

This week's BCE post completes the trilogy of examples featuring Sammy Reshevsky. This time a tactical example in a R+BP+RP vs. R ending. Like many of the BCE positions from actual games, the events leading up to the BCE position are also of interest.

Kemeri 1937 was a big success for Reshevsky, who tied for first with Petrovs and Flohr half a point clear of Alekhine (this was between his World Championship matches with Euwe, so at that point he was ex-World Champion) and Keres. The field also included notables such as Tartakower and Fine.

One of the things I was curious about from the BCE position was how the Black pieces landed on such awkward squares. It almost looked like a study and is even more amazing that they got there when you look at the position after 51. Rxe5

51...Kf6 This is position 364a in BCE, which Fine uses to show the defender holding this type of position 52.Re4 Rb3 53.Ra4 Rb6 54.Kg3 Rb3 55.Ra6+ Kg5 56.Ra5+ Kf6 57.Rf5+ Kg6 58.Rc5 Ra3 59.Rd5 Kf6 60.Kf4 Ra4+ Fine concludes 364a at this point with the comment no real progress could be made 61.Ke3 Ra3+ 62.Rd3 Ra6 63.f4 Re6+ 64.Kf3 Ra6 65.Kg3 Ra1 66.Rd6+ Kg7 67.Kf3 Re1 68.Ra6 Rg1 69.Re6 Kh7 70.Re7+ Kh8?!

After trying several defensive setups it looks like Apscheneek saw the tactical sequence f5 h5 fxg5 Rg5 and decided White had no other way to make progress after f5. There was nothing to fear from the more normal 70...Kg6 here. 71. f5+ Kf6! 72. Rh7 Rh1! and White has nothing 71.Rd7 White isn't ready yet for 71. f5 h5 72. gxh5 Rg5! 73. f6 because of 73...Rf5+ 71...Kg8 72.Ra7 Kh8 I'm guessing that somewhere around here Sammy figured out what Black was planning and then found the way to cross him up 73.Ra4 Kh7 Black probably avoided 73...Kg7 because of 74. f5 h5?! (White has nothing after 74...Kf6) 75. gxf5 Rg5? (75...Kf6 is still a draw) 76. Rg4! and White wins 74.Rd4 Here on 74. f5 h5 75. gxh5 Rg5 76. f6 Rxh5! holds because the Black king has access to both g8 and g6 (76...Rf5+? 77. Rf4! and the f-pawn goes through) and 77. Rg4 cutting off the Black king can nowrap be met with 77...Rf5+ 78. Rf4 Rxf4+ 79. Kxf4! Kg6 74...Kh8?! Black stubbornly clings to his defensive idea. Again, 74...Kg6 is a surer way to draw 75.f5 Baiting the hook 75...h5?! 76.gxh5 Rg5?

The BCE position. Black could still hold a theoretical draw with 76...Rg1+, but who would put his king in the corner and sacrifice a pawn in order to perform a difficult defense? The game concluded 77.f6! Ra5 Black must have figured out White's trap now, but it is too late, see the BCE page for the details of 77...Rxh5. Neither 77...Kg8 78. Rg4! nor 77...Kh7 78. Rf4 offer any hope either. 78.Rd8+ Kh7 79.Kg4 sidestepping Rf5+, White is now ready to push f7 79...Ra7 Black could have tried for stalemate tricks with 79...Rg5+ when 80. Kxg5? is stalemate and 81. Kh4? Rxh5 draws with the same idea, but 80. Kf4! would end all resistance. 80.Kf5 Ra5+ 81.Ke6 Ra6+ 82.Ke7 Ra7+ 83.Rd7 Ra1 84.Ke8+ Kh8 85.f7 Re1+ 86.Re7 Rf1 87.Re6 Dashing Black's last hope 87. f8Q+? Rxf8+! 88. Kxf8 stalemate 1-0

A similar blunder occured on Board 1 in the recent US Masters. In round 2, the top seed Jeffery Xiong had conducted a difficult defense with Black against Magesh Panchanathan in a R+4 vs. R+3 ending. He had reached a drawn position after 62. Ke4 when he went in for a tactical operation

62...h5? The path to the draw is to attack the g-pawn from behind 62...Ra1 63.Kf5 Rf1+ 64.Ke6 Rg1 65.Rc8+ (65.Rh7 Rxg4 66.Rxh6 Ra4; or 65.Rc4 Rg2 66.Rd4 Rg1 67.Kd7 Ra1 In both cases, Black has sufficient checking distance.) 65...Kg7 66.Kd7 Rxg4 67.e6 Rd4+! 68.Ke8 h5 69.Rd8 Re4 70.e7 h4 71.Rd3 Kf6 72.Rf3+ Kg5 and each player will have to give his rook for the opponent's passed pawn. 63.gxh5 Rh6 64.Kd5! Rxh5 65.Kd6! Rh1 66.Rc8+! Kg7 66...Kf7 67.e6+ Kf6 68.Rf8+ Kg7 69.e7 67.e6 Rd1+ 68.Ke7 Ra1 69.Ke8 1-0

Xiong shrugged off this loss and still managed to be a part of the 8-way tie for first. In the Armegeddon game for the title, he pressed John Burke for a long time with White, but in the end only managed a draw

10/1/18 - al Ghamdi - Bekker-Jensen, 2018 Batumi Olympiad

Being such a big upset, the game between Aithmidou and Li got a lot of attention from the major chess sites with anlysis showing that in addition to the opportunity he later had, the Chinese grandmaster also missed an opportunity before transitioning to the rook and pawn ending to reach a draw after 67.Rxg7

67...Ra3+ 68. Kf2 Now, instead of the game continuation 68...Kxd6, he could have inserted 68...Ra2+! If the king retreats to the first rank, the ending after 69...Kxd6 is drawn, but after 69. Kf3 Ra3+! White can't advance 69. Kg4? Rg3+ drops the rook, 69. Kf2 Ra2+! repeats the position and remarkably retreats to e2 or g2 are met by 69...Kxd6 which is also a draw. The latter cases show how narrow the margin was in this ending.

Today, I want to look at another first round ending that I haven't seen discussed anywhere yet. This game between Ahmed al Ghamdi and Simon Bekker-Jensen was board 1 in the match between Saudi Arabia and Denmark. I was quite surprised to see that Black managed to win from the position after 57...Nxf5

Black doesn't even have a passed pawn here. We saw in the game Men-Bereolos that even two passed pawns in a N+2 vs. B+1 ending doesn't guarantee success. 58.Kf3 It looks odd to give ground here. I'll guess White wanted to play 58.Be2 but was afraid of the knight fork 58...Ng3+ The pawn ending after 59.Ke3 Nxe2 60.Kxe2! is a draw, but I don't think many people would play this way. 58...h5 59.Kf2 Again preparing Be2, which would have lead to a lost pawn ending on this move 59.Be2 Nd4+! 60.Kf2 Nxe2! 61.Kxe2 Ke6! with the opposition 59...Ke5 60.Be2 h4 61.Bg4 Nh6 62.Bc8 Ke4 63.Be6 Kf4 Here, White should realize that there is no danger from ...g4 and that the only way for Black to make progress will be to check with his knight on d3 or e4. Consequently, White should transfer his bishop to the b1-h7 diagonal with 64.Bb3 and 65.Bc2 64.Bc8 Ng8 65.Be6?! Nf6 66.Bc8 Ne4+ 67.Kg2 Ke3 White's position is starting to get critical. Black's next step is to transfer the knight to f4 then penetrate with his king to f2 or f3-g3. To cover both squares White will need to answer the check with Kf1 or Kg1 to prevent ...Kf2 and have his bishop on the d1-g4 diagonal to prevent ...Kf3. The knight on f4 will also be attacking h3. Therefore, the only square for the bishop is g4 and he must stay in touch with that square. 68.Bd7 Nf6 69.Bc6? stopping ...Nd5, but Black has another route to f4 and the bishop can't go to g4 anymore. 69.Bc8 Nd5 70.Bg4 Nf4+ 71.Kf1 and Black can't make progress.

Note that this is not a zugzwang position as White can also play Kg1 69...Nh5 70.Bd7 Nf4+ 71.Kh2 Kf2 Now Black methodically squeezes White out. The plan is to transfer the knight to f3 which forces the White King to h1 after which Black plays ...Kg3, transfers the knight back to f4 and wins the h-pawn. 72.Bc8 Ne2 73.Bd7 Nd4 74.Bc8 Nf3+ step 1 75.Kh1 Kg3 step 2 76.Bd7 Ne5 77.Bf5 Nc4 78.Kg1 Ne3 79.Bc8 Nd5 and step 3 as 80...Nf4 and 81...Nxh3 follow. 0-1

This ending was reminiscent of one from this year's US Masters between Mark Dejmek and Charlotte Chess Center Executive Director Peter Giannatos. Comparing the position after 33...Nxc5

to that from the Olympiad game, it would seem that Black has even fewer chances here. Not only are the pawns even, but the Black pawns are fixed on light squares where they can be attacked by the bishop. Still, there is no danger for Black and there are White weaknesses on e3 and g3, so he plays on. 34.Kg2 Kd6 35.Bb5 Ne4 36.Ba6 Kd5 37.Bc8 Nd6 38.Ba6 Kc5 39.Kf2 Ne4+ 40.Kg2 Kb4 41.Bc8 Nd6 42.Be6 Kc3 43.Kf2 Kd3 44.Bd7 Ke4 45.Be6 I would prefer 45.Bc6+ when Black has to retreat or repeat the position with 45...Kd3 46.Bd7 45...Nb5

46.Bd7? Black's plan is ...Nc3-d1xe3, so White should transfer his bishop to cover d1 46.Bb3 Nc3 (46...Kd3 47.Be6 forces Black back.) 47.Bc2+ 46...Nc3! 47.Bc8 47.Ke1 Kf3 48.Bxf5 h6 49.Kd2 Nb5 50.e4 Nd6 51.Kd3 Kxg3 and the passed g-pawn will be decisive. 47...Nd1+! 48.Ke2 Nxe3! 49.Bb7+ Kd4 50.Bc8 f4 51.gxf4 Ke4 52.Ba6 Kxf4 Unlike the 2 vs. 1 position in the Olympiad game, Black already has a passed pawn and White's h-pawn is completely doomed 53.Bd3 h6 54.Bb5 Nf5 0-1