< Earlier Kibitzing · PAGE 4 OF 4 ·
Later Kibitzing> 
Aug2610   ganedi: How does white win after 1.h4 b4 2.Kd4 b3
3.Kc3 Ka3 

Aug2610   rapidcitychess: After looking at this game I have a slang for a knight: A rabbit.
I mean why in the world would you call that jumper a knight! Knights don't hop for crying out loud! Anyway on to the show.
Today's puzzle is like one of those break dance videos. White's knight starts the windmill with 56.Nh5! But black's tango is too slow, he's doing the robot instead of the kickup and white completes with a slide to c1, to cut the video at just the right time. 

Aug2610   YouRang: <ganedi> <How does white win after 59.h4 b4 60.Kd4 b3
61.Kc3 Ka3> (note: I used the game move #s) [diagram]
click for larger view
One way:
<62.Nc1!> (blocking black K from a2 and threatening Nxb3) If <62...b2> then <63.Kxc2 bxc1=Q+ 64.Kxc1> 

Aug2610   ZUGZWANG67: It's the endgame festival! White has the only piece (N) on the board but B is the one who threatens promotion on the short run. W must take immediate measures to prevent that from happening. Thus, he must sacrifice the N. This way of sacrificing is rather common in this kind of endgame so one must not be surprised. But how does one approach? Via e4, d5 or h5? The trap is subtle. I don't think it's worth investigating d5. I would go for h5. To me e4 is definitly to be avoided because it lacks flexibility. In fact it seems that the N must arrive at f4 at some point. From there it keeps the option of either going to d3 or e2, depending upon Black' strategy. For example if B promotes straight the N will go d3, forking the BK and the (new) BQ. But may Black decide to play the K at c3 at some point, then e2 controls c1 and checks the BK. For example, let's imagine 56.Ne4(?) c2 57.Nf2 Kc3! and draw. Or 57.Ng3?? c1Q loses.
For this reason I think 56.h5(!) is more to the point. After 56...c2 (56...Kb3 57.Nf4 and 58.Ne2) 57.Nf4 now: a) 57...c1Q 58.Nd3+;
b) 57...Kc3 58.Ne2! Kd2 59.Nd4;
This is a standard manoeuvre: The N heads for b3 while attacking c2. 59...c1Q 60.Nb3+ wins the Q for the N. After this, the win is obvious. Unless I'm an innocent victim of the "ChessGames trick of the day pattern"(lol!...) I think I'm ok. Time to check!

I'm a thursday survivor. 

Aug2610   Patriot: <ZUGZWANG67> <For example, let's imagine 56.Ne4(?) c2 57.Nf2 Kc3! and draw.> Did you mean "and black wins."? 

Aug2610   VincentL: <David2009>I managed to beat Crafty in the line in which the c pawn promotes to a knight, but only by advancing the a pawn first. Black had to give up the knight for the a pawn, and the K side advance was then easy. 

Aug2610   njchess: I'm familiar with this game so this wasn't much of a puzzle. The knight just does get back. Fischer had a draw through move 41 and should have settled for one. Eliskases was a fine positional player who was probably a bit too cautious against the top echelon of players, and routinely played for a draw. This game was no exception except Fischer tried too hard to win in an otherwise static, drawn game. After the dinner break of 1.5 hours following move 40, White sealed 41. Nb8. Fischer's reply of 41. ... Bc5 made absolutely no sense. 41. ... Bxa3 was the correct move to ensure a draw. Black need only to exchange the pawns to ensure the draw. 

Aug2610   SuperPatzer77: Hey, Chess folks!! After 56. Nh5 what if 56...c2, 57. Nf4 c1=N???? I mean that it is 57...c1=N instead of 57...Kc3. I wonder how we can solve Black's move of 57...c1=N. SuperPatzer77 

Aug2610   wals: Black's error:
(+2.09):41...Bc5. Only move,
Analysis by Rybka 4 x64: depth 27:
1. (0.71): 41...Bd6xa3 42.Nc8xb6[] Ba3xb2[] 43.Nb6xc4[] Bb2c1[] 44.f4f5 h6h5 45.Nc4e5 h5xg4 46.h3xg4 Kh7h6 47.Ne5f7+ Kh6h7 48.Nf7d6 g7g6 49.Nd6e4 Bc1b2 50.Kh1g2 Kh7g7 51.Kg2g3 Bb2e5+ 52.Kg3f3 Kg7f7 53.Kf3g2 Kf7g7 54.Kg2f2 Kg7f7 55.Kf2e3 Kf7g7 56.Ke3e2 g6xf5 (+4.96):51...Be7. Only other move,
Kxa4, +4.49.
(+#21):56...Kxa4. Other moves just as hopeless,
Analysis by Rybka 4 x64: depth 24:
1. (11.04): 56...c3c2 57.Nh5f4[] Kb4c3 58.Kd5e4[] c2c1Q 59.Nf4e2+[] Kc3c4 60.Ne2xc1[] Kc4c5 61.h3h4 Kc5d6 62.Nc1d3 Kd6e7 63.g4g5 h6xg5 64.h4xg5 Ke7e6 65.Nd3f4+ Ke6d7 66.g5g6 Kd7e7 67.Ke4e5 Ke7f8 68.Nf4e6+ Kf8e7 69.g6g7 Ke7f7 70.Ke5f5 Kf7e7 71.g7g8Q Ke7d6 2. (#25): 56...b6b5 57.a4xb5[] c3c2 58.Nh5f4[] Kb4c3 59.Nf4e2+[] Kc3d2 60.Ne2d4[] c2c1Q 61.Nd4b3+[] Kd2d1 62.Nb3xc1[] Kd1xc1 63.b5b6 Kc1d2 64.Kd5e4 Kd2c3 65.b6b7[] Kc3b4 66.b7b8Q+ Kb4c4 67.Qb8c7+ Kc4b5 68.Ke4d4 Kb5b4 69.Qc7b8+ Kb4a3 70.Kd4c3 Ka3a2 71.Qb8b2# 3. (#20): 56...Kb4b3 57.Nh5f4[] c3c2 58.Nf4d3[] Kb3xa4 59.h3h4 b6b5 60.g4g5 h6xg5 61.h4xg5 b5b4 62.Kd5c4 Ka4a3 63.g5g6 b4b3 64.Kc4c3 c2c1Q+ 65.Nd3xc1 b3b2 66.Kc3c2[] b2xc1Q+ 67.Kc2xc1 Ka3b4 68.g6g7 Kb4c5 69.g7g8Q Kc5d6 70.Kc1c2 Kd6e7 71.Kc2c3 Ke7f6 4. (#20): 56...Kb4xa4 57.Nh5f4[] c3c2 58.Nf4d3[] Ka4a3 59.Kd5d4[] b6b5 60.h3h4[] b5b4 61.Kd4c4[] b4b3 62.Kc4c3[] Ka3a2 63.Nd3c1+[] Ka2b1 64.Nc1xb3[] c2c1Q+ 65.Nb3xc1 Kb1xc1 66.g4g5 h6xg5 67.h4xg5 Kc1d1 68.g5g6 Kd1e2 69.g6g7 Ke2f3 70.g7g8Q Kf3f4 71.Qg8g6 Kf4f3 5. (#20): 56...Kb4a3 57.Nh5f4[] Ka3xa4 58.Nf4d3 c3c2 59.h3h4 b6b5 60.g4g5 h6xg5 61.h4xg5 b5b4 62.Kd5c4 Ka4a3 63.g5g6 b4b3 64.Kc4c3 c2c1Q+ 65.Nd3xc1 b3b2 66.Kc3c2 b2xc1Q+ 67.Kc2xc1 Ka3b4 68.g6g7 Kb4c5 69.g7g8Q Kc5d6 70.Kc1c2 Kd6e7 71.Kc2c3 Ke7f6 

Aug2610   CHESSTTCAMPS: I should be familiar with this game, but I am not. It seems that the young whippersnapper Bobby is giving the old master Eliskases a beating in this ending, with the black cpawn racing to promotion. If only white could get the knight to d3 or e2, the pawn would be stopped, but 2 tempos is so little time. The best I could come up with was 56.Ne4 c2 57.Nf2 Kc3 (c1=Q?? 58.Nd3+) and white is out of luck. I was stumped  also looked at 56.a5, 56.Kc6, and 56.Kd4 with no evidence of hope. On the verge of giving up, I took my 4th "fresh" look at the position and found it in a flash: 56.Nh5!!
As counterintuitive a move as one can imagine  with a central square available, the knight instead moves to the edge and away from the action. Just yesterday I saw a kibitzer use the old aphorism "a knight on the rim is dim" and it ran through my head that there are a number of exceptions to this rule of thumb in openings, middlegames, and endings. But this did not clue me in here. A) 56... c2 57.Nf4 Kc3? 58.Ke4! Kd2 (c1=Q 59.Ne2+ wins) 59.Nd3 and white's gpawn will promote first. A.1) 57... c1=Q? 58.Nd3+ Kxa4 59.Nxc1 wins
But with that bpawn, black still has chances, too:
A.2) 57... Kxa4 58.Nd3 b5 59.h4 b4 60.Kc4! (60.g5 hg 61.hg b3 seems to lead to a draw with both sides queening) b3 61.Kc3 Ka3 62.Nc1! b2 63.Kxc2 and black can resign. A.2.1) 59... Ka3 60.Kd4 b4 61.Kc4 b3 62.Kc3 transposes to A.2. A.2.2) 59... Kb3 60.Kd4 b4 61.g5 hg 62.hg Ka2 63.Kc4 b3 64.Nc1+ followed by 65.Nxb3 wins. B) 56... Kxa4 57.Nf4 b5 (c2 58.Nd3 transposes to A.2) 58.Nd3 Kb3 (b4 59.Kc4 stops the pawns) 59.h4 b4 60.g5 hg 61.hg c2 62.Kd4 and black's pawns can do nothing. Time to see if Eliskases pulled off the tricky win... 

Aug2610   sethoflagos: <<SuperPatzer77>: Hey, Chess folks!! After 56. Nh5 what if 56...c2, 57. Nf4 c1=N????> Maybe 58 Kc6 Ka5 59 Nd5 Kxa5 60 Kxb6 Nd3 61 h4 Ne5 62 g5 hxg5 63 hxg5 Ng6
1/21/2! 

Aug2610   DarthStapler: I didn't get it 

Aug2610   YouRang: <SuperPatzer77: Hey, Chess folks!! After 56. Nh5 what if 56...c2, 57. Nf4 c1=N????
I mean that it is 57...c1=N instead of 57...Kc3. I wonder how we can solve Black's move of 57...c1=N. > It looks like just pushing qside pawns will work. For example: 58.h4 Nb3 59.g5 hxg5 60.hxg5 Nc5 61.g6 Nd7 62.Kd6 Nf6 63.Nd5+! nicely forking a N with a N.
click for larger view 

Aug2610   VincentL: <YouRang><SuperPatzer77>After 56. Nh5 c2 57. Nf4 Kc3 58. Ne2 Kd2 59. Nd4 c1=N my effort was: 60. Kc6 Nd3 61. Kxb6 Nb4 62. Nf5 Kd3 63. a5 Ke4 64. a6 Nxa6 65. Kxa6 Kf4 66. Nxh6 Kg5 67. Nf5 winning for white. click for larger view 

Aug2610   YouRang: <VincentL> That works, except it's not the case that <SuperPatzer77> was referring to. The case in question has black playing 57.c1=N (not 57.Kc3). 

Aug2610   CHESSTTCAMPS: <gofer: I take back what I said yesterday about's "knights on the rim...".> But thanks for the setup for my post!
<David2009 <...my suggested 58 Nd3 is an inaccuracy jeopardising the win. 58 Ne2!...>> Your 58.Nd3 is just fine. In preparing my post, I spent a lot of time on the decision Nd3/Ne2, choosing the former on general principles supported by a lot of analysis from the puzzle position. In fact, there are a number of resources for white to stop the dangerous looking pawns. From your first Crafty link, I won as follows: 58.Nd3 c2 59.h4 (a waiting move is useful to get black to commit the bpawn) b4 60.Kc4 Ka3 click for larger viewNow white wins easily with 61.Nc1 Kb2 62.Nb3. A different path is 59... Kb3 60.Kd4 b4 61.g5 hg 62.hg and white can win the black pawns without bothering to queen: 62...Ka3 (Ka2 Nxb4+) 63.Kc4 b3 (Ka2 64.Nc1+) 64.Kc3 click for larger viewNow 64... b2 65.Kxc2, 64... c1=Q 65.Nxc1 b2 66.Kc2, or 64... Ka2 65.Nc1+ are all losing. 

Aug2610   SuperPatzer77: Welldone, <YouRang>. Thanks for an excellent analysis. White sure wins if 57...c1=N!?. White's strong reply to 57...c1=N!? is 58. h4!! as you mentioned. SuperPatzer77 

Aug2610
  patzer2: For today's Thursday puzzle solution, we were treated to one of Bobby Fischer's rare loses  even if it was one of his earlier games as a young teenage GM. In this case, he drops the point to 47yearold Austrianborn Erich Eliskases, who after winning several championships in his youth in Europe went to South America at the outbreak of WWII and become a citizen of Argentina. The game between 17yearold Fischer and the older Eliskases could have ended in a peaceful draw with the rather obvious 40...Bxf4! or the more subtle 40...Bxa3! recommended by Mednis. Instead, Bobby pressed too hard for the win and Eliskases responded to the challenge and gave us a real treat with the winning endgame tactical surprise 56. Nh5! It's not clear to me whether Bobby anticipated 56. Nh5! or not. However, it is clear that it is the only way for the Knight to stop the dangerous passed cpawn. Indeed, 56. Nh5! is not only the winning move, it is White's only move that doesn't lose. The alternative 56. Ne4? c2 57. Nf2 also anticipates the use of a Knight Fork to gain a tempo to stop the pawn. However, the attempt fails miserably after 57...Kc3 58. Ne4! After 56. Nh5! c2 58. Nf4! White also holds the passer with the fork threat, but the difference is he also has the option of playing Nd3 with tempot to capture the passed cpawn. Here's some movebymove analysis of 56. Nh5! c2 58. Nf4! with Fritz 10: 56. Nh4! c2 57. Nf4! Kb3 (57... Kxa4 58. Nd3 Kb3 59. h4 ) (57... Kc3 58. Ne2+ Kd2 59. Nd4! c1=N 60. Kc6 Nd3 61. Kxb6 ) 58. Nd3 Kc3 59. Ke4 Kd2 60. h4 . 

Aug2710   EXIDE: Very nice! I saw the fork, however did not see the importance of e2 and hence missed h5. 

Aug2710   FISCHERboy: ReikiMaster: 36..Bxf4 is quite drawish as is 40..Bxf4. 41..Bxa3 42.Nxb6 Bxb2 43.Nxc4 gives white a nominal advantage. But then looking for a draw is not how you win three consecutive World Championship Candidate matches in a row with a perfect total score of 2000! I agree, Sir! 

Feb1513   Eduardo Bermudez: Eliskases Centenary !! 

Jul0314   zydeco: I have the sense that Fischer was spiritually inclined towards the Ragozin Variation, which was featured in his favorite book (Lipnitsky's Questions of Modern Chess Theory), but couldn't quite make it work (+23=0 in Wade's index of Fischer games). He gradually dropped it after the early '60s. Mednis, in How to Beat Bobby Fischer, claims that Eliskases was basically playing for the draw (especially starting with 26.Qh5 instead of the stronger 26.Qc4), and Fischer had it for the asking (e.g. 34....Bxf4 35.Nf8+) but was hellbent on a win. 41.Nc8 was the sealed move, which makes Fischer's blunder with 41.....Bc5 especially incomprehensible (he had a forced draw with 41....Bxa3). Mednis says that Fischer, in his adjournment analysis, planned to trap the knight with 45.....b5 followed by 46....Kd7 but missed 47.Kd5! when the knight escapes. In Fischer's defense, it wasn't a real adjournment, just an hourandahalf dinner break, in which the players had to analyze quickly while eating. 

Jul0314   Petrosianic: <He gradually dropped it after the early '60s.> Yeah, but he still seemed to be trying to steer into it as late as Game 1 of the Spassky match. 

Sep0718
  plang: 12..e5 at once would have been playable as 13 Nd5..Qd8 (13..Qe8? 14 Nf6+!) 14 dxe..Nxe5 15 Nxe5..Bxe5 and Black is OK (if 16 Nb6..Bxh2+). 31 exf+..Qxf7 32 Qxf7+..Kxf7 33 Nxb7..Ke6! would have been good for Black whose active king would have been excellent compensation for the pawn. 34..Bd6 avoiding the perpetual with Ne8+ showed that Black was playing for a win. Fischer had apparently counted on trapping the knight with 45..b5 46 axb..Kd7 missing that White would have had the winning reply 47 Kd5!. 

Nov2422
  Messiah: Fantastic endgame, I love it! 



< Earlier Kibitzing · PAGE 4 OF 4 ·
Later Kibitzing> 