Author: Roy Eassa
Date: 08:13:48 09/29/03
Go up one level in this thread
On September 29, 2003 at 10:36:10, Alexander Yates wrote:
>In Reply to: Re: Garry Kasparov - My Great Predecessors posted by Mark Young on
>September 28, 2003 at 09:34:43:
>
>
>On September 28, 2003 at 09:34:43, Mark Young wrote:
>
>>On September 28, 2003 at 04:38:08, Kurt Utzinger wrote:
>>
>>>On September 28, 2003 at 00:15:02, Mark Young wrote:
>>>
>>>>H. Pillsbury - Em. Lasker
>>>>After whites 17 move of f5
>>>>
>>>>[D]2r2rk1/pp3pp1/4bb1p/q2p1P1Q/3P4/2N5/PP4PP/1K1R1B1R b - - 0 1
>
>
>
>>>>
>>>>Here GM Kasparov gives 17..Rxc3 a (!!) mark. With the comment that the move
>>>>Rxc3!! is beyond the powers of even a strong computer.
>>>>
>>>>Is Rxc3 really beyond the powers of a strong computer? Or is the move Rxc3 not
>>>>worthy of a (!!) because it leads only to a draw. This is what the computers see
>>>>after blacks 17..Rxc3.
>>>
>>>
>>> I do not think that 17...Rxc3 will be played by
>>> any computer program. And by the way: I have still
>>> not found a win for Black in the following line:
>>>
>>> [Event "40'/40+40'/40+40'"]
>>> [Site "MyTown"]
>>> [Date "????.??.??"]
>>> [Round "?"]
>>> [White "Neue Partie"]
>>> [Black "?"]
>>> [Result "1/2-1/2"]
>>> [SetUp "1"]
>>> [FEN "2r2rk1/pp3pp1/4bb1p/q2p1P1Q/3P4/2N5/PP4PP/1K1R1B1R b - - 0 1"]
>>> [PlyCount "23"]
>>> [TimeControl "40/2400:40/2400:2400"]
>>>
>>> 1... Rxc3 2. fxe6 Ra3 3. bxa3 Qb6+ 4. Kc2 Rc8+ 5. Kd2 Qxd4+
>>> 6. Ke1 Qc3+ 7. Ke2 Qc2+ 8. Rd2 Qe4+ 9. Kd1 Qb1+ 10. Ke2 Qe4+
>>> 11. Kd1 Qb1+ 12. Ke2 Qe4+ 1/2-1/2
>>
>>I came to the same conclusion. I don't think there is a win after Rxc3 with best
>>defense. This is why computer will not play the move. Computers see Bd7 holding
>>a much better advantage. And it seems the computers are correct.
>
>
>The computers are not correct. They can't see deep enough to find Rxc3!!
>followed by Ra3!
>
>This game I'm sure has been well analyzed by many Grandmasters, not just
>Kasparov, who BTW is the highest rated player of all time!
>
>There are many combinations still too difficult for computers, and that has been
>demonstrated by various tests. Nolot comes to mind...#6 I believe. Regardless,
>we know some combinations are too deep or abstruse for the computer at present
>time.
>
>Even if a computer does find the move, it will take a very long time and it may
>not be for the correct reasons.
>
>1... Rxc3 2. fxe6 Ra3 3. bxa3 Qb6+ 4. Kc2 Rc8+ 5. Kd2 Qxd4+
>6. Ke1 Qe3+ 7. Be2 fxe6 8. Qh3! Bc3+! 9. Kf1 Rf8+ 10. Bf3 Qe4! -+
>
>There are other variations of course and the B retreat has merit so the machines
>most likely will stay with Bd7, instead of moving on to the more complex Rxc3!!
>
>Just for the record, Lasker made this sacrifice famous, and Rxc3 is now regarded
>as thematic.
>
I have spent a bit of time analyzing with several programs. I think you're
right that this is probably a win for Black. It is VERY deep with many
variations. Here are a few more:
1... Rxc3! 2. fxe6
[2. bxc3 Rc8! 3. fxe6
(3. Rc1 Qb6+ 4. Bb5 Bd7 5. a4 Bxb5 6. axb5 Qxb5+ 7. Kc2 Qa4+
8. Kd3 Bg5 9. Ke2 Qa2+ 10. Kf3 Bxc1 11. Rxc1 Qd2 -+)
(3. Rd3 Rc6 4. a4 Qxa4 5. Rd2 g6! 6. fxg6 Bg5 -+)
3... Qxc3 4. exf7+ Kf8 5.Qe2 Bxd4 6. Qe8+ Rxe8 7. fxe8=Q+ Kxe8
8. Rxd4 Qxd4 -+]
2... Ra3! 3. bxa3
[3.e7 Re8 4. bxa3 Qb6+ 5. Kc2 Rc8+ 6. Kd2 Bxd4! -+]
3... Qb6+ 4. Kc2 Rc8+ 5. Kd2 Qxd4+ 6. Ke1 Qe3+ 7. Be2 fxe6 8. Qh3
[8. Kf1 Bd4 9. Rxd4 Qxd4 10. g3 Rf8+ 11. Bf3 Qd2 12. h3 Rf5 13. Qg4 h5 -+]
8... Bc3+ 9. Kf1 Rf8+ 10. Bf3 Qe4! 11.Qg3
[11. a4 Bb4! 12. Qh5
(12. Kf2 Bc5+ 13. Kf1 Qc4+ 14. Ke1 Bb4+ 15. Kf2 Qxa2+ -+)
12... g6! 13. Qh3 Qc4+ 14. Kf2 Qc2+ 15. Kg3 Bd6+ 16. Kh4 Rf4+
17.g4 g5+ 18. Kh5 Kg7 {and Black forces mate}]
11... Qc4+ 12. Kf2 Bd4+ 13. Rxd4
[13. Ke1 Qc3+ 14. Kf1 e5 15. Rxd4 Qxd4 16. Qf2 Qc4+ 17. Qe2 e4 -+]
13... Qxd4+ 14. Ke2 Qb2+ 15. Ke1 e5! 16. Bxd5+ Kh8 17. Qd3 Qc1+ 18. Qd1 Qxd1+
[{More complex is} 18... Qc5 19. Qd3 Rd8 20. Rf1 Rxd5 21. Rf8+ Qxf8
22. Qxd5 -+]
19. Kxd1 Rd8 -+ {Black should win the endgame.}
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.