|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing > |
Mar-13-07
 |
| YouRang: <maxfrank> Again, you're probably right. I tend to be of the old school that says nobody ever won by resigning (or accepting a draw). In this Ribli-Spassky game, it seems to me that there is a chance for a blunder. For instance, if white had played on with 86. Kf5, it might continue: 86...Qf8+ 87. Kg5 (diagram:black to move)
click for larger view
And here, black might play the plausible-looking 87...Kh7, or perhaps 87...Qf7, but either of these moves gives white mate in 22! I wouldn't *expect* a GM to make such a blunder, but then again, I've seen worse. :-) |
 |
Mar-13-07
 |
| playground player: How much time was consumed by this game? You'd think White could have kept playing, unless he was exhausted or time was running out. |
|
| Mar-13-07 |
| Themofro: Originally i assumed that it was white to play and win, so i just played Qd4+ mating next move. Then i realized while this page was loading that it was ...85, which meant black to play, lol. Found it in a second though, if you are familiar with the idea then it's not challenging at all. The queen on the g file was a clue since then black only had to take away his escape squares on the h file and get rid of his queen... |
|
| Mar-13-07 |
| MiCrooks: After Kf5 Black has 9 moves that hold the draw! Kh7 or Q any of h1,h2,h7,f8,a6,b6,c6,d6 works. Ah, the marvels of tablebases! But since Ribli immediately agreed to a draw it must be a well know drawn endgame at that point. I had initially thought that there might be chances for White as well, since the King is not normally supposed to be over near the pawn in Q vs Q+P endings. |
|
| Mar-13-07 |
| maxfrank: <YouRang> In the position you show, ...Qh6 again is the consistent move which requires no think time, since it is only a repetition. At some level, playing on might be OK, but I think it is well below the master level for the game position when the draw was agreed. Anyway, the problem served its purpose. Cute tactic, plus some endgame study. |
|
| Mar-13-07 |
| Lifelong Patzer: Got Qxh6 right away! An 85 move game! Exhausting just to think about. |
|
Mar-13-07
 |
| al wazir: I still think white had a win on move 67. |
|
Mar-13-07
 |
| Billy Ray Valentine: <RandomVisitor: 85...Qd2+ also draws. 85.Qh3 does not work, it is a tablebase draw after 85...Qf2 or 85...Qc2. In fact, the game is drawn by tablebase after 80.Qxf3.> 80. Qxf3 is the first move where there are only 6 pieces left on the board, which allows a tablebase assessment. What move throws away the win for white? I have to say, Queen and Pawn endings are the most annoying endgames to play... and I say that as someone who likes endgames more than the opening or middlegame. |
|
| Mar-13-07 |
| Justawoodpusher: I found the drawing move very fast but then was not so sure if this is a perpetual. So I tried to save the draw against Fritz. It's possible, but in zeitnot I would probably make a mistake. <jahhai> I found that one should for expample stay with the black queen on the squares b1,c2,h7 when the white king wants to hide on h3. |
|
Mar-13-07
 |
| fm avari viraf: To be candid, I didn't like Spassky's position. In fact, Ribli could have grind his advantage into a win. But Spassky being a magician tactician found a resourceful move Qxh6+ & salvaged half a point. |
|
| Mar-13-07 |
| maxfrank: Actually, the game seemed resignable at move 81. Perhaps that was Spassky's intention. I'd guess, "I've hung on so far. I'll play on a little with the time control adrenaline still affecting Ribli, and (84 Kh4) - oops! he just screwed up!!!" |
|
Mar-13-07
 |
| mig55: I think ....Qd2+ draws too? |
|
Mar-13-07
|
| euripides: <max> for more on why the general principles in Fine/Benko are not to be trusted, see http://www.jeremysilman.com/book_re... |
|
Mar-13-07
|
| euripides: <max> thanks for the von Trier suggestion - I have seen the old Pasolini version many years ago, but not this - I will keep an eye out. |
|
Mar-13-07
 |
| Sneaky: Even if there are other ways to draw this position according to tablebase, there's no doubt about which move is the best move. For today, you either saw ...Qxh6+ or you get no credit, in my opinion. I saw it quickly only because it was Tuesday. If they gave me this on Friday I probably would give up without even trying, thinking it must be over my head! |
|
| Mar-13-07 |
| Dr.Lecter: Well, I thought there might be a perpetual, but couldn't find any. Conclusion: it took me over 10 seconds to find the answer. |
|
| Mar-13-07 |
| cionics: I feel really dense, but can someone help a neophyte! What happens after 86 Kxh6? I don't see why this is a draw. Shouldn't white be able to promote the pawn eventually? Also, what is the meaning of a "tablebase draw". Thanks. |
|
Mar-13-07
 |
| chancho: <cionics> After 86.Kxh6 the Black King has no moves.(stalemate) |
|
| Mar-14-07 |
| cionics: Duh! Thanks, Chancho. |
|
Mar-15-07
|
| YouRang: <cionics><Also, what is the meaning of a "tablebase draw".> A "tablebase" is a computer-generated exhaustive table of certain chessgame endings. For example, it is not hard to find an online tablebase that covers all endings with 6 or fewer pieces (where the king, pawns, and everything else counts as a "piece"). With such a tablebase, you can give it any legal position with 6 or fewer pieces (and tell it whose turn to move), and it will tell you the "state" of that position. Every such position is in exactly one of 3 states:
(1) white can force a win
(2) black can force a win
(3) the side to move can force a draw.
The tablebase will tell you what the state is for the given position, AND tell you what state you will obtain for each legal move that can be made in that position. Hence, the tablebase, used iteratively, will tell you how to force a win (if possible), or how to preserve the draw (if possible). If either side can force a win, it also tells you how many moves until mate. This position is a "tablebase win" or a "tablebase loss", depending on whose turn to move. If neither side can force a win (i.e. state 3 above), the position is a "tablebase draw". BTW, I don't think any tablebases allow castling -- they assume that castling rights are gone. Also, you can't give it an initial position where capturing en passant is possible. |
|
| Mar-16-07 |
| cionics: Yourang,
Thanks for the very complete answer. I always appreciate your posts, by the way. But, I must say, I often picture you as looking like your icon which is kind of scary! |
|
| Mar-24-07 |
| Fisheremon: <al wazir: I still think white had a win on move 67.> Sure as 67.a5 could do and seemed to be the last (theoretical) winning chance for White. |
|
Mar-30-07
 |
| gambitfan: A superb stalemate obtained as the game looks lost! |
|
| Oct-14-09 |
| crosscheck: What is wrong with 84. Qg7+. That seems to do it. That takes off the queens! Surely white can win the pawn ending? |
|
| Oct-15-09 |
| Eduardo Leon: <crosscheck>, unfortunately, after <84. Qg7+?>, black gets the opposition and, thus, the draw: <84. Qg7+ Qxg7 85. hxg7+ Kxg7 86. Kg4> (<86. Kh4 Kh6>) <86. ... Kg6>. |
|
 |
|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing > |
