< Earlier Kibitzing · PAGE 13 OF 13 ·
|Aug-07-11|| ||DrMAL: <David2009> I have tried your link on the Crafy Endgame Trainer a few times. It's a fun thing to play with but, unfortunately, it simply does not take enough time to compute and often comes up with incorrect answers. Why not simply use a standard engine like Crafty interactively on a standard GUI like Arena and this way you can wait to see how its answers change in time (eventually to the right one or at least a better one). Just a suggestion, cheers.|
|Aug-08-11|| ||LIFE Master AJ: <<<< Aug-07-11 patzer2: Fascinating Sunday puzzle. Never remotely considered the insane deflection sacrifice 36. Rc5!! The computers indicate a win for white after 36....dxc5? as <LIFE Master AJ> observes, even though the line looks difficult to this amateur.> |
I wonder if Black might hold after 36. Rc5!! Qa6 37. Rxb5! Bxc6! According to <David 2009> that is Crafty's choice, and I can't find anything with Fritz to force a win for White.>
In any event, this was one to enjoy this Sunday. The tactics involved after 36...dxc5? and the neat mate-in-four or five with the dancing Knights after 41. Ng4+! makes today's puzzle more than worth the price of admission.>>
<patzer2> All valid points.
|Aug-08-11|| ||DrMAL: <Jack Kerouac> Call it karma then, because Tal went +0 -4 =3 against Fischer thereafter (blitz games included).|
|Aug-08-11|| ||CHESSTTCAMPS: I favored 36.N3xe5 like many of the others, but could not work it through. Very neat key move from Tal!|
|Aug-08-11|| ||sevenseaman: <CHESSTTCAMPS> <36.N3xe5 > may be a natural looking move and it was my choice on 14 Sep 2010. But it didn't succeed. Only Tal got a win with <36. Rc5>. I have tried my best against Crafty and it too did not work. It does not happen in the game but Crafty defends with <37...Bxb6> and I am yet to cope with that. Very frustrating!|
|Aug-11-11|| ||Pawn and Two: My choice was 36.Nfxe5 dxe5 37.Qxe5+, with the thought of positioning my queen on e7. However, I did not try to calculate this variation any deeper.|
So what is the correct continuation in this position?:
click for larger view
It takes Fritz a while, 20 ply in fact, before it could confirm that Tal's move 36.Rc5!!, is the best choice. The continuation 36.Nfxe5 dxe5 37.Qxe5+, was Fritz's 2nd choice.
Here is Fritz's analysis: (1.24) (25 ply) 36.Rc5!! Qa6 37.Rxb5 Bxc6 38.dxc6 Qxc6 39.Ra5 Rc8.
In this variation, Kasparov in OMGP gives the inferior continuation 39...Rxa5? 40. bxa5 Nhf6 41.a6 Nd7, indicating Black has a difficult, but still defensible position. However, after 40.bxa5, Fritz shows White to be clearly winning: (3.51) (25 ply) 40...Nhf6 41.a6 Nc7 42.a7 Nd7, ( or 41...Nd7 42.a7 Nc7), 43.Qb8 Nxb8 44.axb8Q.
A much better defense for Black is 39...Rc8.
After 39...Rc8, a tempting continuation is: (1.61) (28 ply) 40.Ng5 Kh6 41.Nf7+ Kg7 42.Ra7 Qc1+ 43.Qxc1 Rxc1 44.Kh2 Rc7 45.Rxc7 Nxc7 46.Nxd6 Kf6. I did not find a convincing winning continuation for White in this variation.
However, after 39...Rc8, White still has excellent winning chances with: (1.60) (28 ply) 40.b5! Qc3 41.Qd2 Qb3 42.Ra7+, (1.97) (26 ply) 42...Kh8 43.Ng5 Nhf6, (2.30) (24 ply) 44.Nf7+ Kg8 45.Kh2 Kf8, (2.14) (25 ply) 46.f3 Rc3 47.Bb1 Rc7 48.Rxc7 Nxc7 49.Qxd6+ Kxf7 50.Qxc7+ Ke6 51.Qc6+ Ke7 52.Qc5+ Kd8 53.Qxe5, is winning for White.
After 40.b5!, Black could also try 40...Qc1+ 41.Qxc1 Rxc1+ 42.Kh2, (2.02 (26 ply) 42...Rc3 43.Bf1, (2.31) (26 ply) 43...Kh6 44.b6 Rc8 45.Bb5 Rb8 46.Ra6 Neg7 47.Bc6 Ne6 48.b7 Nc5, (2.91) (30 ply) 49.Ra8 Rxb7 50.Bxb7, or (2.99) (30 ply) 49.Ra7 Nf6 50.Nd2, are winning for White. An alternative choice to 43...Kh6 is: (2.42) (26 ply) 43...Rb3 44.Ng5 Nhf6 45.Ra7+ Kh6 46.Nf7+ Kh5 47.Bc4 Rb2 48.Ra2 Rxa2 49.Bxa2, is also winning for White.
After 40.b5! Qc3 41.Qd2, Black could also try 42...41...Qxd2 42.Nxd2, (1.80) (25 ply) 42...Rb8 43.Nc4 Kf8 44.b6 Nhg7 45.Ra7 Ne6 46.Bc2 Nd8, (2.11) (25 ply) 47.f3 Nb7 48.Ba4 Nf6 49.Kf2 Nd8 50.Rc7 Ne6 51.Nxd6 Nxe4+ 52.fxe4 Rxb6 53.Rc8+, are winning for White.
Fritz's 2nd choice, 36.Nfxe5, gives White an advantage, but a draw appears to be the likely result: (.69) (24 ply) 36.Nfxe5 dxe5 37.Qxe5+, (.96) (23 ply) 37...Kh6, (not 37...Kg8? 38.Bc2! Nhf6 39.Qe7 Qc7 40.Bb3), 38.Qe7 Bxc6 39.dxc6 Qc7, (.88) (26 ply) 40.Qc5, (.63) (26 ply) 40...Nhg7 41.Qxb5 Rb8 42.Qc5 Ne6 43.Qd5 Qe7 44.Qd7 Qxd7 45.cxd7 Nf6 46.Rc6 Nxd7 47.Bxe6 Rxb4.
|Aug-18-11|| ||CharlesSullivan: <Pawn and Two> After 37...Bxc6 38.dxc6 Qxc6 39.Ra5 Rc8,
click for larger view
I first thought that 40.Ng5 leads to a win for White. However, I am swayed by your analysis of 40.Ng5 -- a draw is the likely outcome.
But is the suggested 40.b5 better? For example: 40...Qc3 41.Qd2 Qb3 42.Ra7+ Kg8(!) 43.Ng5 Nhf6 44.Kh2 h6 45.Nf7 Kf8 [giving up a pawn] 46.Nxh6 Rc7 47.Rxc7 Nxc7 48.Qa5 Ne6 49.Qa7 Nc5 50.Bc2 Qc4 51.b6 Ncd7 52.b7 Qc7 53.Qa8+ Nb8
click for larger view
and Black has constructed a form of "fortress." An overnight computer think [Zappa Mexico II, 6 hours, +2.53] says best play is now 54.Ng4 Nxg4+ 55.hxg4 g5 56.Bb3 Ke7 57.Bd5 Kd7, and it seems that White cannot improve his position.
Any further thoughts?
|Aug-19-11|| ||Pawn and Two: <CharlesSullivan> Thanks for your input on this fantastically complicated game. |
Fritz reviewed your suggested "fortress" position for a while. It did not see a way to defeat it.
However, Fritz did suggest an improvement for White in the 40.b5 line, after your suggested 42...Kg8.
In this line, after 42...Kg8 43.Ng5 Nhf6 44.Kh2 h6 45.Nf7 Kf8, Fritz indicates White is winning with the following continuation: (3.34) (25 ply) 46.Nh8 Rc7 47.Rxc7 Nxc7 48.Nxg6+ Kf7, (4.27) (25 ply) 49.Nf5 d5 50.exd5 e4 51.Be2 f3 52.Nxh6+ Ke7 53.gxf3 exf3 54.Nf5+ Kd7 55.Nd4 Qxd5 56.Bxf3.
Fritz indicates that 45...h5, instead of the pawn sacrifice 45...Kf8, may be a little better continuation for Black: (3.07) (26 ply) 45...h5 46.Ng5 h4 47.Nf3 Ng7 48.b6 Qxb6 49.Qa2+ Kh8 50.Nxh4 Rc7 51.Rxc7 Qxc7 52.Nxg6+ Kh7 53.Nxf4, but White is still winning.
|Aug-22-11|| ||CharlesSullivan: <Pawn and Two> Thanks for the quick turnaround analysis. |
Your refutation of the 37...Bxc6 38.dxc6 Qxc6 39.Ra5 Rc8 40.b5(!) Qc3 41.Qd2 Qb3 42.Ra7+ Kg8 43.Ng5 Nhf6 44.Kh2 h6 45.Nf7 Kf8 variation with 46.Nh8(!) Rc7 47.Rxc7 Nxc7 48.Nxg6+ Kf7 is sound, although 49.Nf5 is an illegal move at this point. One line that leads to at least a 2-pawn (winning) advantage is 49.Nh4 Nfe8 50.Nf5 Ke6 51.Qe2 Kd7 52.Nxh6.
So 42...Kh8 is better, with your mainline being 43.Ng5 Nhf6 44.Nf7+ Kg8 45.Kh2 Kf8 46.f3 Rc3 47.Bb1 Rc7 48.Rxc7 Nxc7.
click for larger view
Here 49.Qxd6+ Kxf7 50.Qxc7+ Ke6 51.Qc6+ Ke7 52.Qc5+ Kd8 53.Qxe5 does not convince me because 53...Nd7(!) pretty much forces White to give back one of his two extra pawns, leading to a (probably) drawn ending with pawns only on the kingside; play might proceed 54.Qh8+ (nor did I find a win after 54.Qa1) 54...Kc7 55.b6+ Qxb6 56.Ba2 Qa5 57.Bc4 Qb4 58.Ba6 Qb6 59.Qc8+ Kd6 60.Bc4 Qf2 61.Qa6+ Qb6 62.Qa8 Kc7 63.Qh8 Qb4 64.Qg8 Qe7 65.Ba6 Ne5 66.Qc8+ Kd6 67.Kh1 Qf6 68.Qb7 h6 69.Bc4 Nc6 70.Bb5 Ne5 71.Ba4 Ke6 72.Qc7 Qe7 73.Bb3+ Kf6 74.Qc1 g5, and White cannot break through.
More likely to win is 49.Ba2 Nxe4 (a quicker White win follows 49...Qa3 50.Nxd6) 50.fxe4 Qg3+ 51.Kg1 Nxb5 52.Nxd6 Qc3 53.Qxc3 Nxc3 54.Bd5 Ke7. (At this point, Zappa prefers 55.Nc8+.) Rybka3 continues 55.Nc4 Kf6 56.Nb6 Ke7 57.Kf2 Nb5 58.Nc4 Kf6 59.g4 g5 60.Nb6 Ke7 61.Ba2 Nd6 62.Nd5+. At this point, Rybka shows +4.81 and Black is lost.
It is hard to known for certain amidst such complications, but it does seem as if Black cannot save himself with 37...Bxc6.
|Aug-23-11|| ||Pawn and Two: <CharlesSullivan> Thanks for catching my error. I had omitted a move in the 42...Kg8 variation.|
In this line, I had intended to show: 42...Kg8 43.Ng5 Nhf6 44.Kh2 h6 45.Nf7 Kf8, (3.34) (25 ply) 46.Nh8 Rc7 47.Rxc7 Nxc7 48.Nxg6+ Kf7 49.Nh4 Nce8, (4.27) (25 ply) 50.Nf5 d5 51.exd5 e4 52.Be2 f3 53.Nxh6+, with a winning position for White.
In the 42...Kh8 variation, you noted a critical point, after the moves, 43.Ng5 Nhf6 44.Nf7+ Kg8 45.Kh2 Kf8 46.f3 Rc3 47.Bb1 Rc7 48.Rxc7 Nxc7, and now instead of 49.Qxd6+?, you correctly noted the winning move for White, 49.Ba2!.
Fritz agrees with your indicated variation. Had I let Fritz recheck the position at move 49, it too would have clearly preferred 49.Ba2!.
After 49.Ba2!, (2.36) (26 ply) 49...Nxe4 50.fxe4 Qg3+, (2.49) (25 ply) 51.Kg1 Nxb5 52.Nxd6 Qc3 53.Qxc3 Nxc3 54.Bd5 Ke7, (2.67) (27 ply) 55.Nc4 Kf6 56.Nb6 Ke7 57.Kf2.
After 57.Kf2, Fritz shows White to be winning: (3.18) (30 ply) 57...Nb5 58.g3 g5 59.g4 Nd6, (3.85) (29 ply) 60.Nc4 Nxc4 61.Bxc4 Kd6 62.Ke2.
Another winning variation for White is: (3.36) (30 ply) 57...h6 58.g3 g5 59.g4 Nd1+ 60.Ke2 Ne3, (4.58) (30 ply) 61.Nc4 Nxc4 62.Bxc4.
I think the evidence is fairly conclusive, Tal's move 36.Rc5!!, was a winning idea.
At one time you expressed an interest in the New York 1927 tournament. I know you indicated you did not care to work with the German edition of the tournament book. However, as you may know, the long overdue English edition has recently been published. I highly recommend it, you will not be disappointed with this book!
|Aug-26-11|| ||CharlesSullivan: <Pawn and Two> I do have the English version of the New York 1927 tournament. Even before I got the book, I had analyzed about half the games. (I'll get around to the rest of the games one of these days.) |
The quality of Alekhine's analysis is about the same as his other books -- usually very good, but the computer catches him out quite frequently. Since he was such a great blindfold player, I have a pet theory that he probably analyzed the games without using a board and pieces; possibly that accounts for some of his analytical oversights.
|Nov-15-11|| ||dagano: seeing moves like Rc5 make me feel like i'm just playing a totally different game than tal did.|
|Jan-23-12|| ||Whitehat1963: Amazing save!|
|May-30-12|| ||DanielBryant: This game had an immense effect on me as a young player - not so much the tactical blaze at the end (which was largely lost on me at the time), but the long posturing and the amazing journey of the knight (especially 26.Nxa1).|
|Jul-28-12|| ||Poisonpawns: This is the type of game that will make you quit chess, or make you will study your behind off.|
|Nov-23-12|| ||Abdel Irada: What a knight!|
|Jan-21-13|| ||Tigranny: This game looks way prettier than Tal vs Larsen. Not at all overrated in my view.|
|Jan-21-13|| ||dehanne: Hard to believe Hjartarson didn't see what was coming when he played Ra1.|
|Jan-21-13|| ||andrewjsacks: Well, for one thing, it was probably the last move before the time control...|
|Mar-09-13|| ||sfm: <dehanne: Hard to believe Hjartarson didn't see what was coming when he played Ra1.>|
Well, the pretty 41.Ng4+ 42.Nh6+ 43.Ng8+ 44.Ng5# is not that easy to see if time is short.
Maybe Black checked on the natural 41.Nxd7+?? where he wins.
And Black is beaten anyway. He can avoid instant mate, e.g. by 40.-,RxR, but now 41.Nxd7+,K.. 42.Nxb8 leaves him in a hopeless position.
So, as a "fast nice try", 40.-,Ra1 was quite an obvious move.
|Jan-07-14|| ||PJs Studio: While 36.Rc5 is a fantastic move, I can't fathom 39.Ncxe5!! After 39...Qd1+ and 40...Ra1 white is mated if he can't see the subsequent four knight checks ending in mate. (I couldnt see 43.Ng8+ on the move! Yet Tal HAD to see it at move 39. (!)) Tal's attack was brilliant.|
I agree with sfm, I think Harjartarson couldn't see the mate. Might have thought he had a won game even?
It's games like this that make me feel like a wood pushing monkey.
|Feb-23-14|| ||SpiritedReposte: Let it be known Tal was a beast. Absolute boss. Exciting, razor sharp chess that bamboozled the best in the world!|
|Mar-18-15|| ||Alex Schindler: I was following, it all made perfect sense, I was momentarily curious about the decision nxa1 instead of moving a rook onto the file, but OK, a knight was making its way from outpost to outpost so the move seemed justified, certainly more useful than possible rook swaps... |
And then 36Rc5!, followed by ncxe5!! shortly thereafter and I realized, I'm hopelessly trying to understand the mind of a wizard.
|Jun-05-15|| ||SimplicityRichard: The Knight in shining armour!#|
|Oct-30-15|| ||mikealando: Tal the magnificent|
< Earlier Kibitzing · PAGE 13 OF 13 ·