|Dec-11-07|| ||Jim Bartle: Can't believe I slept through this...|
|Dec-11-07|| ||shr0pshire: That was an amazing endgame by Karjakin. I don't know how he did it. It seemed like he didn't have any theoretical chances to draw around move 39. Well played though.|
|Dec-11-07|| ||Tacticstudent: Unbelieveble! I thought Shirov was totally won; but once again, Karjakiin proves his valor.|
|Dec-11-07|| ||scholes: The move sequence between 25 and 40 is unbelievable for a rapid game .Let the future visitors be known this was a rapid game .This game is much better classical games between carlsen and kamsky .Whats the point of having 2 hours if you spend all of it in 3 moves out of opening for finding a way to lose your rook in 5-7 moves and blitz rest of the game.|
|Dec-11-07|| ||Eyal: Starting with 49.Nxe4 we're in tablebase territory, where this position is declared as won for White in <208> moves with best play for both sides! (no joke). It's interesting to note that most of the time - until 93...Bxf3+, after which it's a draw - both players perform quite accurately according to the oracle, and at certain points Shirov is not so far from nailing it (at least theoretically - I'm not sure what's the status of the 50 moves rule in such cases). For example, 86...Kd4 would be a win in 25, but then Rh1 and Rh2+ slide Black back to a win in 131... The position after 93.Kg2, where Black turns it into a draw by the piece exchange, is declared to be a win in 39 moves with best play for both sides, starting with: 93...Re8 94.Nf1 Ke4 95.Ng5+ Kd4 96.Nf3+ Kd3 97.N1h2 Ke3 98.Nf1+ Ke2 99.Ng3+ Kd3 100.Nh4 Ra8 101.Ngf5 Ra6 102.Kf2 Rf6 103.Kg2 Ke4 104.Kg3 Be2 105.Kh2 Ba6 106.Kg3 Bc8 107.Ne7 Bd7 108.Nhg6 Rb6 109.Kh4 Be8 110.Nf4 Rb7 111.Nfd5 Kd4 112.Kg4 Bf7 and White loses one of the knights.|
|Dec-11-07|| ||whiteshark: <47...Rd3> was the mistake leading to a draw as it lost the last black pawn.|
|Dec-11-07|| ||Shadout Mapes: <Eyal> Very important correction: after 49.Nxe4 the position is won for <black> in 208 moves. After reading your post I stared at the position for a while in disbelief before checking the tablebase myself to see white's win!|
|Dec-11-07|| ||Eyal: <Shadout Mapes> Heh - yes, of course. And instead of 47...Rd3, Rd8 - defending the pawn by the threat of check on the e file - would probably make this endgame more humanly possible to win (though it still seems to be quite tricky).|
|Dec-11-07|| ||dabearsrock1010: but how can you win in 208 wouldnt it be drawn by 50 moves rule|
|Dec-12-07|| ||ganstaman: <dabearsrock1010: but how can you win in 208 wouldnt it be drawn by 50 moves rule>|
Yes, but tablebases ignore that. So the game is theoretically won, but technically drawn.
Out of curiosity, isn't there some situation(s) where the 50-move rule is extended for reasons such as this?
|Dec-12-07|| ||cotdt: The arbiter can suspend the 50-move rule in situations like this, technically. I don't know if it has ever been done before, though.|
|Dec-13-07|| ||acirce: <Out of curiosity, isn't there some situation(s) where the 50-move rule is extended for reasons such as this?> No. There used to be exceptions, but those were removed long ago.|
<The arbiter can suspend the 50-move rule in situations like this, technically.> No. Stop making things up.
|Dec-13-07|| ||cotdt: <acirce> then in that case, the rules of chess seem to be problematic. there are many endgames that are theoretically winnable, but the 50-move draw rule would prevent it. seems pretty adhoc rule to me. i'm sure that as tablebases expand, a lot of endings that look completely drawish are actually decisive... in hundreds of moves. usually when i play correspondence chess, both sides agree that a decisive tablebase result is automatic victory.|
FIDE does have provisions to make exceptions to almost any situation, but I won't get into an endless debate over it. There are always higher powers that the arbiters can look to.
|Dec-13-07|| ||Eyal: <9.3 The game is drawn, upon a correct claim by the player having the move, if
a. he writes his move on his scoresheet, and declares to the arbiter his intention to make this move which shall result in the last 50 moves having been made by each player without the movement of any pawn and without any capture, or |
b. the last 50 consecutive moves have been made by each player without the movement of any pawn and without any capture.> (http://www.fide.com/official/handbo...)
So, currently no exceptions mentioned.
<The rule has a long history, with Ruy López's 1561 text on chess including details of it. In the 20th century, with the discovery that certain endgames (see below) can only be won in more than fifty moves (without a capture or a pawn move) from certain positions, the rule was changed to include certain exceptions in which one hundred moves were allowed with particular material imbalances. The exceptions were later removed and all material combinations are now subject to the fifty move rule.
The exceptional positions (above) were:
1. Two knights versus one pawn (See Troitsky line)
2. Rook and bishop versus a rook
3. Rook and a rook pawn on its original square, versus a pawn blocking the rook pawn and a bishop on the same color as the opponent's pawn.
In 1928 FIDE enacted rules that in the rook and bishop versus rook endgame, 132 moves were allowed, since it was twice the 66 moves that were thought to be required at that time (the actual maximal number of moves needed is 59). In 1952 FIDE revised the law, requiring that players agree to an extension for these positions before the first move is made. FIDE rules allowed seventy-five moves for the rook and bishop versus rook.
At some point, the rule was changed to one hundred moves for such positions. Later more positions requiring more than fifty moves were found. FIDE included these endgames in the extended rule:
1. queen versus two bishops
2. queen versus two knights
3. two bishops versus a knight
4. two knights versus a pawn
5. rook and bishop versus a rook, and
6. a queen with a pawn on the seventh rank versus a queen.
The one hundred move extension was in force for a short time, and it was changed to seventy-five moves in 1988. In 1992 the rule was changed back to fifty moves for all positions.> (http://en.wikipedia.org/wiki/Fifty_...)
So apparently, unlike what I thought at first, the changes in the rules didn't have so much to do with the development of tablebases. And considering the state of current knowledge about endgames, that return to the fifty moves rule for all positions seems very problematic, if not downright silly.
|Dec-13-07|| ||Eyal: Btw, interesting to note that the case of rook and bishop versus two knights, which arose in the Karjakin-Shirov game, was never cosidered as one of the exceptions to the 50 moves rule.|
|Dec-13-07|| ||Jim Bartle: I'm interested in the earlier part of the game, when Shirov still had a pawn. Could he have won by saving that pawn?|
|Dec-13-07|| ||whatthefat: Well I think it's a matter of practicality. There are certain positions that take several hundred moves to win, and require many only moves to be made by the attacker along the way. And more than likely a human player will make inaccuracies that extend the win, so where do you draw the line? Do you go twice as far as the tablebase minimum number of required moves? If so, you need a set of rules for every material combination. And what about positions that have not been solved by tablebases yet?|
All in all, I think the 50-move rule is a practical solution.
|Dec-13-07|| ||cotdt: <All in all, I think the 50-move rule is a practical solution.> For OTB chess, you might be right. But here correspondence chess diverges from this rule, at least the way most people are playing it. The 50-move rule maybe also exist for correspondence chess, but nobody I know follows it. If there is a tablebase win, then it's a win regardless of whether there is a span of 50 moves of no capture. It's very peculiar that chess under different time controls are now being played under different rules.|
|Dec-13-07|| ||whatthefat: Which correspondence chess association doesn't apply the 50-move rule?|
|Dec-14-07|| ||TIMER: <cotdt> Makes sense that, if you are allowed to refer to tablebases- then tablebase wins are accepted as wins. However in this case it is not to do with the time control but access to computers, tablebases and reference materials that is the important difference.|
|Dec-14-07|| ||TheAlchemist: To continue what <Eyal> started, Tim Krabbe wrote the following:|
In the starting position, Shirov was 208 moves away from mate, assuming perfect play by both sides. And apart from that blunder, he did make steady progress, Karjakin defending well. Then on move 68 Karjakin blundered with 68.Nbd3 (going from a distance of 149 moves to 54), Shirov immediately repaying the compliment by letting him come back to 162 with Rh3+, instead of 68...Re8.
Generally, Shirov played better, more or less staying at the shorter distances to the win that Karjakin allowed him. From move 49 to 70, he steadily progressed from 208 to 160, then he was around 110 for a few moves, jumping to 84 at move 76, and to 43 at move 77. He even managed to get under 30 - until 86.Ne1, when at only 25 moves, he was closest. But precisely at that point, he fell back to 131 in just two bad moves.
Of course Shirov didn't know all that - but the invisibility of his progress, even the uncertainty of it, must have felt like one of those dreams where you try to run, but with each step you float back, ever further from your goal.
With 92.Kg3, Karjakin again allowed a huge shortcut: from 136 to a striking distance of 31, but at that point, it seems both players were through with this ending; Shirov's 92...Rf8 threw away 43 moves; 93.Kg2 threw back 35, and finally 93...Bxf3+ threw away the whole win.
All in all, in his 44 moves in this endgame, Shirov played the best one 24 times; Karjakin scoring 19 out of 43 - 'best' meaning staying on the fastest road to victory or the slowest one to defeat.
PS: In these rapid games, the players do not have to write down their moves, and as a result, cannot claim 50-move draws. It is therefore an academic matter that only at move 86, Shirov missed a (theoretical!) win within 50 moves of the last capture (49.Nxe4): 86...Kd4 87.Nc2+ Kd3 88.Ne1+ Ke2 89.Nc2 Rh1+ 90.Kb2 Kd3 91.Nb4+ Kc4 92.Nd5 Rh2+ 93.Ka3 Bc2 94.Ne3+ Kd3 95.Ng4 Rh5 96.Nf2+ Ke2 97.Nc3+ Kxf2 and mate in 13 move moves.>
|Sep-17-16|| ||whiteshark: btw the only ♖♗:♘♘ endgame in this database|
|Jan-02-19|| ||norami: To have a checkmate in 208 moves in a real game - is that the record? Has anyone heard of a longer mate in a real game?|