Thursday, October 23, 2014


Regularly on this blog I disclose flaws of our best engines. Sometimes to compare engines. At other moments to proof that some ideas discovered by humans, can still not be coded in an algorithm. Despite the flaws few will doubt about the usefulness of the engines. Even my team-captain Robert surprised me a few days ago by including some engine-evaluations in his report which a few years earlier was simply unthinkable. We can't deny the fact that the engines have surpassed us largely in strength. Only some time ago Carlsen stated in an interview: "I can't beat the best computers".

Quality Chess recently published a nice example of how big the difference in strength is today between human and computer. The writer asked a + 2650 player how many big mistakes (direct wins or an important tactical sequence) he made in 24 games (some opponents had only 2200 elo). An engine could quickly find 10 mistakes despite the moderate opposition. As 2300 player I obviously make many more of those big mistakes. Some of those are surely avoidable but I believe there are also mistakes which can'be prevented via an optimal time-consumption or tactical trainings.

In round 6 of Open Gent I and my opponent missed a beautiful tactic. It was the only flaw in otherwise a pretty smooth victory. Just before going to bed after the play-day  I discovered it with an engine. I can't resist the temptation to check immediately just finished games with an engine despite the late hour, often after midnight. Some players surely won't agree with such attitude as they find it bad for the night's rest.
[Event "Open Gent 6de ronde"] [Date "2014"] [White "Brabo"] [Black "Stuer, M."] [Result "*"] [ECO "B92"] [WhiteElo "2333"] [BlackElo "2130"] [SetUp "1"] [FEN "2rrn1k1/1q4p1/p2p3p/Pp1Pppb1/nPP5/1Q2BP2/2RNB1PP/2R3K1 w - - 0 25"] [PlyCount "13"] 25. cxb5 $6 {(A small flaw in the game as I miss here a devilish combination. Bxg5 would have kept full control.)} Nc3 $1 { (A sacrifice creating an interference. It is something known pretty well in chess-problems but in OTB it is pretty rare so also easy to miss.)} (25... Rxc2 $6 {(Black continued with this exchange so also did not notice it. After the move black did not get anymore chances to turn the tide and had to resign 15 moves later. )}) 26. Rxc3 Bxe3 27. Kf1 Bxd2 28. bxa6 {(The engines manage to find anyway a big advantage for white but this is clearly a coincidence.)} Rxc3 29. axb7 Rxc1 30. Bd1 Nc7 31. Ke2 $16 *
A fantastic move which initially the engines evaluate as equal for black but when giving more time they anyway find a big advantage for white in the complications. The piece-sacrifice creates an interference between queen and bishop. Such type of interferences are really rare in practice but it is well-known in the world of chess-compositions. There exists a bunch of themes on (Holzhausen) interferences. In the next paragraphs I will explain in a nutshell the different possibilities as this is an interesting and beautiful little niche of chess.

The Grimshaw was already covered in my article chess-compositions as it was multiple times used in my problem Loyds organ pipes. A Grimshaw is a mutual interference of 2 pieces. The simple example below demonstrates this theme.
Grimshaw-problem White mates in 2 moves
When the mutual interferences of the 2 pieces happen due to a piece-sacrifice then we call it a Novotny. In problem-chess this is rather ordinary so it is in most cases included in a bigger concept. I selected a rare example from board-practice which explains in its simplicity well the idea.
[Event "Rowy Jantar op"] [Site "Rowy"] [Date "2000.06"] [White "Berg, Emanuel"] [Black "Zezulkin, Jurij"] [Result "*"] [ECO "B33"] [WhiteElo "2432"] [BlackElo "2510"] [SetUp "1"] [FEN "7k/7p/p2p3Q/4br2/4pqN1/6R1/5P1P/7K w - - 0 40"] [PlyCount "3"] 40. Nf6 $1 {(A Novotny between bishop and rook. White played Qe6 in the game and after quite some adventures the game ended in a draw.)} Bxf6 (40... Qxh6 41. Rg8# ) (40... Rxf6 41. Qg7#) 41. Qf8# *
If the Novotny is created by pieces moving in the same direction (diagonal or vertical) then we call it a Plachutta. A simple but clear example from the American top-composer William Anthony Shinkman can be found below.
Plachutta-problem White mates in 3 moves
My game-fragment is not a Plachutta despite there was a piece-sacrifice and the queen + rook move rectilinear. There is only 1 interference instead of 2. The queen can't take the knight because the queen will be captured and not because the rook will be interfered. Finally when we have a Plachutta on one and the same line then we talk about an Anti-Bristol. This last one is the most complex theme and it is much harder to find a good self-explaining example.
Anti-Bristol White mates in 3 moves
The theme Anti-Bristol originates from the Bristol-clearance which I already covered in my article problem-moves. So many different types of interferences exist. You don't need to know this to play chess well but I do find it entertaining. Aren't we playing chess first and foremost to enjoy, right?


Grimshaw-problem: 1.Qb1 (menaces Qb7#)
1...., Bb2 2.Qh1# The bishop interferes the rook.
1...., Rb2 2.Qf5# The rook interferes the bishop.

Plachutta-problem: 1.d5 (menaces Ra8# en Rg8#)
1...., Bxd5 2.Rg8+ Bxg8 3.Ra8# The bishop interfered the queen.
1...., Qxd5 2.Ra8+ Qxa8 3.Rg8# The queen interfered the bishop.

Anti-Bristol: 1.c3 (tempo)
1...., Qc6 2.Pd6+ Qxd6 3.Ne3# The queen interfered the rook.
1...., Rf6 2.Be6+ Rxe6 3.Na3# The rook interfered the queen.

Tuesday, October 7, 2014


If there is today one player capable to win games with a minimalist approach then surely it is the reigning world-champion Carlsen. Time over time he proves that playing solid moves is often sufficient to let the opponent crack. Not everybody is enthusiast about such style of playing which e.g. can be read in the reaction of the German honorary-president Robert von Weizsacker. Soulless, boring, this has nothing to do about who is the better player but just who can sit the longest concentrated at the board (sitzfleisch) , are some of the harsh reproaches.

I don't want to discuss here if Carlsens games are attractive or not as there is no accounting of taste. However what I do want to extract is that most players even in equal positions aren't able to maintain the balance. On my level this is of course even more relevant. My article my most beautiful move discuss this aspect in detail. My opponents play much less stable than Carlsen opponents whereby it often is sufficient to avoid big errors to win a game. By applying such cautious approach my games often last longer than average. I must admit that the simul which recently I gave in Veurne due to this approach drifted into an exhausting marathon. My apologies for those being upset by this.

So I find Stevens reaction on my previous article a bit exaggerated. Games really don't end quickly in a draw because on 1 or 2 moments not the most critical move was played. Besides before complaining about the number of draws, we should first check if we apply sufficiently Sofia rules. It has little meaning to demand more aggression if too easily premature draws are made. On the other hand I do have to admit that the reaction also includes some truth. Sometimes the opponent neither makes clear mistakes and more than a draw can't be achieved without taking some risks. This also happened in my game against Andrew Stone.

That is also the reason why at move 42 I suddenly made the decision to sacrifice a piece for 2 connected white pawns. With less than 5 minutes remaining on the clock I must admit that it was mainly intuitively but soon it became clear that the sacrifice was fully sound. After the game I told Kara that I only found the idea very late in the game when I figured out that playing solid moves is not sufficient to win. Kara replied that the sacrifice is well known from the King's Indian. Nowadays I am reading Kasparov series of My Great Predecessors to improve my knowledge of chess-history. It is a must for me with my limited repertoire but I hadn't encountered before this sort of sacrifice in the few King's Indian games which I already replayed.

Following Kara's advice I replayed also all the won games in the King's Indian of former wc-finalist David Bronstein and the French grandmaster Igor Nataf both specialists in this opening but I could not find examples of sacrifices on g4. Besides g4 looks weird in the King's Indian so maybe there is a confusion with the standard sacrifices on h3. There are many types of sacrifices. I can distinguish 3 big categories of light piece sacrifices for 2 connected pawns.

A first category is the sacrifice in the endgame. A light piece is sacrificed to obtain 2 connected and far advanced pawns which overpower the opponents extra piece. I remember an anecdote of the reigning Belgian champion Geert Van der Sticht whom after a painful defeat against former world-class-player Michail Gurevich ventilated his emotions.
[Event "Metz op 20th"] [Site "Metz"] [Date "2002.04.17"] [Round "5"] [White "Gurevich, Mikhail"] [Black "Van der Stricht, Geert"] [Result "1-0"] [ECO "E32"] [WhiteElo "2641"] [BlackElo "2420"] [SetUp "1"] [FEN "r7/pn2kpp1/1pn1p2p/1N6/1P2P3/PK3P2/5BPP/3R4 b - - 0 25"] [PlyCount "50"] 25... Rd8 {(Geert underestimates or misses the piece-sacrifice on a7 otherwise something else would have been played.)} 26. Rxd8 Nbxd8 27. Nxa7 {(Elementary according to Mikhail. It is likely not winning but the defense is absolutely not easy.)} Nxa7 28. Bxb6 Nc8 29. Bd4 g6 30. a4 Kd7 31. b5 f5 32. exf5 exf5 33. Be3 Ne6 34. a5 f4 35. Bf2 Nd6 36. b6 Kc6 37. a6 g5 38. Kc2 h5 39. Kd3 g4 40. b7 Kc7 41. Bd4 Nxd4 42. Kxd4 Nf5 $2 {(The previous moves also contain mistakes but this was blacks last chance to save the game.)} (42... gxf3 $1 43. gxf3 Nf7 44. Ke4 Ng5 45. Kxf4 Ne6 {(Black arrives in time to capture the pawns on the queen-side and return to stop the pawns on the king-side.)}) 43. Ke4 Ne3 44. fxg4 hxg4 45. Kxf4 Nd5 46. Kxg4 Nb4 47. h4 Nxa6 48. h5 Nc5 49. h6 Nd7 50. Kf5 1-0
After the game Mikhail made a witty remark something like "Elementary, my dear Watson" which obviously wan't very pleasant for Geert. The piece-sacrifice was surely not winning but gives very dangerous practical chances. Now a player of the strength like Mikhail knows many more bricks than most players so he was right. A variation of this theme can be found in the famous game Capablanca - Lilienthal which is covered in My Great Predecessors part 1.
[Event "Moscow"] [Site "Moscow"] [Date "1936"] [White "Capablanca, Jose Raul"] [Black "Lilienthal, Andor"] [Result "1-0"] [ECO "A12"] [SetUp "1"] [FEN "5k2/rp3ppb/1np1p2p/N7/1PP5/6P1/4PPBP/3R2K1 b - - 0 34"] [PlyCount "40"] 34... Ke8 $2 {(Black has already a difficult position but this allows a winning piece-sacrifice. )} 35. Nxb7 Rxb7 36. Bxc6 Rd7 37. c5 Ke7 38. Bxd7 Nxd7 {(Normally exchanging a rook and 2 pawns for 2 pieces is acceptable but here the connected pawns are too strong.)} 39. c6 Nb6 40. c7 Bf5 41. Rd8 e5 42. Rb8 Nc8 43. b5 Kd6 44. b6 Ne7 45. Rf8 Bc8 46. Rxf7 Nd5 47. Rxg7 Nxb6 48. Rh7 Nd5 49. Rxh6 Kxc7 50. e4 Ne7 51. f3 Kd7 52. h4 Ke8 53. Rf6 Ng8 54. Rc6 1-0
A second category is well known as it initiates a king's attack. The pawn-shield in front of the king is dismantled after which the king is under fire. As a special example I selected a King's Indian game with a piece sacrifice on g4 of which I earlier stated not have found any examples. Well the difference exists in the fact that white and not black sacrificed.
[Event "Tel Aviv ol (Men) qual-A"] [Site "Tel Aviv"] [Date "1964"] [Round "1"] [White "Keres, Paul"] [Black "Walther, Edgar"] [Result "1-0"] [ECO "E93"] [SetUp "1"] [FEN "r2q1rk1/pp6/3p3b/3Ppb1p/6pP/2N3P1/PP2QRPN/5R1K b - - 0 23"] [PlyCount "26"] 23... Bg6 $4 {(Qd7 guarantees black a clear advantage but black obviously missed whites piece sacrifice on g4. Blacks king is now dead meat.)} 24. Nxg4 hxg4 25. Qxg4 Kh7 26. h5 Bd3 27. Rxf8 Bxf8 28. Rf3 Bc2 29. Ne4 Kh8 30. Rf7 Qe8 31. Nxd6 Qa4 32. Qg5 Qa6 33. Qxe5 Kg8 34. Rf6 Qd3 35. Kh2 Qh7 36. Qe6 1-0
Finally we arrive at the most difficult category and those are the more positional piece-sacrifices. There is no direct king's attack and no immediate promotions are threatened but the opponent is mainly restricted in activity. A beautiful example is without doubt Bronstein - Botvinnik which we can find back in My Great Predecessors part 2.
[Event "World Championship 19th"] [Site "Moscow"] [Date "1951.04.24"] [Round "18"] [White "Bronstein, David I"] [Black "Botvinnik, Mikhail"] [Result "1/2-1/2"] [ECO "D15"] [SetUp "1"] [FEN "r4nk1/1bqn2bp/2p1p1p1/1pPpNp2/1P1P1P2/2NBP2P/1B4P1/RQ4K1 b - - 0 26"] [PlyCount "64"] 26... Qc8 $6 {(Botvinnik underestimates or misses whites next move as he gets into serious troubles.)} 27. Bxb5 {(A pure positional sacrifice. Promotions or a kings attack is not immediately possible but black lacks a proper plan.)} Nxe5 28. fxe5 Bh6 29. Bc1 cxb5 30. Nxb5 Nd7 31. Nd6 Rxa1 32. Qxa1 Qa8 33. Qc3 Bf8 34. b5 Bxd6 35. exd6 Qa4 36. Qb2 Kf7 37. Kh2 h6 38. e4 f4 39. e5 g5 40. Qe2 Kg7 41. Qd3 $2 {(The adjourned move but this throws the win away. Sufficient for the win were e.g. c6 and h4.)} Nb8 42. h4 Qc4 43. Qh3 Qxb5 44. hxg5 hxg5 45. Qxe6 Qd3 46. Qf6 Kh7 47. Qf7 Kh8 48. Qf6 Kh7 49. Bxf4 gxf4 50. Qf7 Kh8 51. Qe8 Kg7 52. Qe7 Kh8 53. Qe8 Kg7 54. Qe7 Kh8 55. Qf8 Kh7 56. Qf7 Kh8 57. Qxb7 Qg3 58. Kh1 1/2-1/2
I believe you can find the 3 different sort of bricks in my game with Stone. When I execute the piece-sacrifice, we already deal mainly with an endgame. However earlier in the game the piece-sacrifice was also possible and the conditions different. Below my analysis discuss the different possibilities in detail.
[Event "Open Gent 5de ronde"] [Date "2014"] [White "Stone, A."] [Black "Brabo"] [Result "0-1"] [ECO "A10"] [WhiteElo "2200"] [BlackElo "2333"] [SetUp "1"] [FEN "r1b2rk1/1p3q2/3b1n2/p1p1p1pp/P1PpPpP1/3P1P1P/1PNB2B1/1R1Q1RK1 b - - 0 22"] [PlyCount "77"] 22... Bd7 $6 {(At this stage I was not thinking yet about sacrificing a piece on g4. With Bd7 I try to provoke b3 so the square is not available anymore for the other white pieces but this is too prophylactic. Stronger was Be6 followed up by Qd7 to introduce the option of sacrificing a piece on g4.)} (22... Be6 $1 23. Na3 Qd7 24. Nb5 (24. Kf2 $5 Kg7 25. Rh1 $5 Rh8 26. Qb3 $5 {(Because of this type of moves I played Bd7 but at the other wing blacks attack already breaks through.)} hxg4 27. hxg4 Bxg4 $17) 24... Be7 $1 25. Kf2 Kg7 $5 26. Rh1 Rh8 27. Qc2 $1 hxg4 $5 28. hxg4 Bxg4 29. Rbg1 $17 {(Accepting the piece sacrifice would give black a blistering kings attack.)}) 23. b3 Kg7 $5 { (An interesting alternative for black is Qh7 but it is unclear if it is any better than the more logical continuation of the game.)} (23... Qh7 $5 24. Na3 Be7 25. Nb5 Rfc8 26. Qe2 $1 Kg7 27. Rf2 $15) 24. Na3 $6 { (More important is to evacuate the white king out of the danger-area.)} (24. Kf2 $1 Rh8 25. Qe2 $1 Rh6 $5 26. Rg1 $1 Rah8 27. Bxa5 hxg4 28. hxg4 Rh4 29. Ke1 Nxg4 { (Even now the sacrifice is still possible but it does not have the same impact anymore.)} 30. fxg4 Rxg4 31. Kd2 Rh2 32. Rbf1 Rg3 33. Kc1 Bh3 34. Ne1 Bxg2 35. Rxg2 $15) 24... b6 $6 {(I still was not considering the piece-sacrifice on g4 so b6 looked to me a safe move before challenging the h-file. Extensive analysis however show that directly Rh8 is stronger.)} (24... Rh8 $1 25. Nb5 Qe7 $1 26. Kf2 hxg4 $1 27. hxg4 Rh4 $1 28. Rh1 Rah8 29. Bxa5 $1 {(Capturing the pawn is whites best choice for some counterplay. Here we see the disadvantage of no b6.)} Bxg4 $1 30. Rxh4 Rxh4 31. Qd2 $5 Bxf3 $1 32. Bxf3 $1 b6 $1 33. Rg1 Rh2 34. Rg2 Rxg2 $17) 25. Rf2 $2 {(The rook must have the possibility to be transferred to the h-file and this move does not help at all. Nb5 was much better deteriorating blacks coordination and buying time for white to organize the defense.)} Rh8 26. Nb5 Qe7 27. Kf1 $5 {(I spent a lot of time analyzing Rf1 to rectify whites 25th move but the extra tempos are too important.)} (27. Rf1 $5 hxg4 $1 28. hxg4 Rh6 $1 29. Kf2 Rah8 30. Rh1 $5 Rxh1 31. Bxh1 Rh4 32. Qg1 Qf8 33. Bg2 $5 Be7 $1 34. Rf1 Nxg4 {(Sometimes it is better to sacrifice the bishop while in other cases it is better the knight.)} 35. fxg4 Bxg4 $1 36. Ke1 Qh8 37. Rf2 $19) 27... Rh6 $2 {(If I would have considered the piece-sacrifice then I surely would have played hxg4 followed up by Rh4.)} (27... hxg4 $1 28. hxg4 Rh4 $1 29. Ke1 Nxg4 $1 {(Here again better than Bxg4.)} 30. fxg4 Bxg4 31. Bf3 Bh3 32. Qe2 g4 33. Kd1 gxf3 34. Qxf3 Qg5 $1 35. Kc2 $19) 28. Nxd6 $2 {(As demonstrated some earlier analysis, this bishop sometimes plays a role in the kings attack after the piece-sacrifice. However here there was no such danger. White loses an important tempo which he should have spent to evacuate the king out of the danger area.)} (28. Ke1 $1 Rah8 $5 29. Qe2 hxg4 30. hxg4 Rh4 31. Kd1 {(White escaped just in time as none of the sacrifices on g4 seems to guarantee something substantial. Practically it is likely still generating good chances.)}) 28... Qxd6 29. Ke2 Rah8 30. Qf1 Kf7 31. Kd1 Ke7 32. Kc2 hxg4 33. hxg4 Rh2 $2 {(At about this move I started for the first time to consider the piece-sacrifice. Because I was not able to calculate everything and putting all the heavy pieces on the h-file looked attractive, in the end I rejected the sacrifice.)} (33... Bxg4 $1 34. fxg4 Nxg4 35. Re2 Qf6 $1 36. Kb2 Nh2 $19 {(White has no good answer against the steam-roller.)}) 34. Qe2 Qb8 35. Rg1 Qf8 36. Bf1 Qh6 {(Objective achieved but now I realized this does not bring any dividends for black.)} 37. Be1 Rxf2 38. Bxf2 Rh7 39. Bg2 Qf8 40. Bf1 Rh2 41. Bg2 Qc8 {(Yes now I understood the position but white can easily remove the threat.)} 42. Kc1 $2 {(I was very surprised by white not playing Bf1. He still had sufficient reflection-time but maybe he gambled on my time-trouble.)} Bxg4 {(With less than 5 minutes remaining this is my last chance to win the game.)} (42... Nxg4 {(I doubt if this sacrifice wins.)} 43. fxg4 Bxg4 44. Qf1 Qh8 45. Rh1 Kf6 46. Rxh2 Qxh2 47. Qh1 $1 Qxh1 48. Bxh1 Kg6 49. Kd2 Kh5 50. Bg2 Bc8 51. Ke2 $1 g4 52. Bg3 { (A fantastic move. Both engines still evaluate the position as winning for black but it is a fortress.)}) 43. fxg4 Nxg4 44. Bh3 Rxh3 45. Qxg4 Qxg4 46. Rxg4 {(Now we liquidate to a pure endgame in which the pawns easily ensure the win.)} Kf6 47. Kd2 Kg6 48. Ke2 Kh5 49. Rg1 g4 50. Be1 Rh2 51. Kf1 Kg5 52. Rg2 Rh3 53. Rd2 g3 54. Kg2 Kg4 55. Bxg3 fxg3 56. Rd1 Rh2 57. Kg1 Kf3 58. Rf1 Rf2 59. Rb1 g2 60. Kh2 Rf1 0-1
The analyses took a lot of time especially because today's engines are still pretty helpless in planning such piece-sacrifices. To learn chess consists of studying a large amount of such bricks and engines aren't the best teachers hereby. I realize more and more that it is vital for the own chess-development that we have to study the classical masterpieces from our rich chess-history.