Author: Robert Hyatt
Date: 08:55:37 02/06/03
Go up one level in this thread
On February 06, 2003 at 05:18:21, Dave Gomboc wrote: >On February 05, 2003 at 23:23:39, Robert Hyatt wrote: > >>On February 05, 2003 at 22:03:02, Will Singleton wrote: >> >>>On February 05, 2003 at 21:01:52, Robert Hyatt wrote: >>> >>>>On February 05, 2003 at 19:31:34, Will Singleton wrote: >>>> >>>>>On February 05, 2003 at 19:03:27, Drexel,Michael wrote: >>>>> >>>>>>On February 05, 2003 at 18:40:37, Will Singleton wrote: >>>>>> >>>>>>>On February 05, 2003 at 18:29:14, Martin Giepmans wrote: >>>>>>> >>>>>>>>On February 05, 2003 at 18:05:27, Will Singleton wrote: >>>>>>>> >>>>>>>>>On February 05, 2003 at 18:00:11, Martin Giepmans wrote: >>>>>>>>> >>>>>>>>>> >>>>>>>>>>Of course my engine thinks that it is winning big time after 16 g3 >>>>>>>>>>(instead of Bxh7+). >>>>>>>>>>Is there any engine out there that doesn't agree? >>>>>>>>>> >>>>>>>>>>Martin >>>>>>>>> >>>>>>>>>The position: >>>>>>>>> >>>>>>>>>[D]rnb1r1k1/pp3ppp/2p5/3p4/3P1Pnq/P1NBPK2/1PQBN1P1/R4R2 w - - 0 16 >>>>>>>>> >>>>>>>>>After analyzing with Ruffian, the following moves seem to be the best: >>>>>>>>> >>>>>>>>>16.g3 Nh2+ 17. Kf2 Ng4+ 18.Ke1 Qh3 19.Rg1 Nh2 20.Bxh7+ Kh8 21.Bd3 and white is >>>>>>>>>much better. >>>>>>>>> >>>>>>>>>(btw, if 18.Kg2 Qh2+ Kf3 g6! =+) >>>>>>>>> >>>>>>>>>So, perhaps there's something wrong after 18.Ke1... >>>>>>>>> >>>>>>>>>Hmmm.. Ruffian doesn't see anything after 14 ply, still has the main line above >>>>>>>>>at over 1 pawn for white. >>>>>>>>> >>>>>>>>>I can't see it. Anyone have a clue? >>>>>>>>> >>>>>>>>>Will >>>>>>>> >>>>>>>>My engine also doesn't see it (neither do I, but that's not amazing ;)) >>>>>>>>16. g3 wins, period. >>>>>>>>Or is there something hidden in the deep? >>>>>>>>Or was kasparov a little bit afraid? >>>>>>>>Or is Rolf right after all? >>>>>>>> >>>>>>>>Martin >>>>>>> >>>>>>>Actually, I think my analysis was wrong. After 16.g3, black plays Qh2 instead >>>>>>>of Nh2+, and then it's far from clear. Black looks like it will at least >>>>>>>equalize, and with white's king stuck in the center, black could win it. >>>>>>> >>>>>>>It appears GK can see far more than your average run-of-the-mill GM commentator. >>>>>>> >>>>>>>Will >>>>>> >>>>>>After 16...Qh2 white could play 17.Bxh7+ in case 17.f5 17...h5 or even 17...Nd7 >>>>>>18.Kxg4 Qg2 is too dangerous for white. >>>>>>White should be safe after 17.Bxh7+. >>>>>> >>>>>>Michael >>>>> >>>>>Yes, it looks like you're right. >>>>> >>>>>Will >>>> >>>> >>>>When Kasparov chose Bxh7, he took a long time. Crafty was analyzing the >>>>game live on ICC and turned in this 16 ply search result: >>>> >>>> 16-> 33:34 1.59 16. g3 Nh2+ 17. Kf2 Ng4+ 18. Ke1 Qh3 >>>> 19. Rg1 Nd7 20. Kd1 Nf2+ 21. Kc1 Nb6 >>>> 22. Re1 Nxd3+ 23. Qxd3 Bf5 24. Ng1 >>>> Qg4 25. e4 Bxe4 26. Nxe4 dxe4 27. Rxe4 >>>> Rxe4 28. Qxe4 Qxg3 >>>> >>>>After the Bxh7 move, I instantly got this: >>>> >>>> 10-> 2.86 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 >>>> 11 4.41 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 >>>> 11-> 5.46 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 (s=2) >>>> 12 9.92 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 >>>> 12-> 15.30 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 (s=2) >>>> 13 29.72 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 >>>> 13-> 39.29 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 (s=2) >>>> 14 1:09 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 >>>> 14-> 1:39 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 (s=2) >>>> 15 2:47 0.00 16. ... Kh8 17. Ng3 Nh2+ 18. Kf2 Ng4+ >>>> 19. Kf3 >>>> >>>>I would have preferred to see the g3 line myself... It looked (to me anyway) >>>>that white had a lot of life left there and was a piece up... The attack might >>>>have looked scarey, but it really didn't appear to be so to the computer... >>>> >>>>I think he is simply not up to playing more games vs the computer. If I were >>>>to venture a really outlandish opinion, rather than saying "DJ is maybe as >>>>strong as DB" I now might say "DB might really be as weak as DJ". It appears >>>>that Kasparov just doesn't have what it takes to play a 6 game match vs the >>>>machines... Not having been there playing for a million bucks, I can't claim >>>>to understand this. But I have played _many_ games against the machines (mine >>>>and others) and I must say "I don't get it" myself... >>> >>>Maybe you misunderstood. The apparent best line is not 16.Bxh7+, but rather >>>16.g3 Qh2 17.Bxh7+... so yes, g3 appears better than what GK played, but only >>>by a small margin. Crafty doesn't see that 16...Nh2+ is perhaps weaker than >>>Qh2. >>> >>>Will >> >> >>OK... I was only considering 16. g3 vs 16. Bxh7. >> >>I ran crafty after Qh2 rather than Nh2+... >> >>It (so far) thinks white is significantly better off after Qh2 than after >>Nh2+. Here's the output so far: >> >> 13-> 5:54 2.15 2. f5 h5 3. Rae1 Nd7 4. e4 Qh3 5. Nf4 >> Nge5+ 6. dxe5 Nxe5+ 7. Kf2 Qh2+ 8. >> Ke3 Qxg3+ 9. Kd4 Nf3+ 10. Kc5 Nxe1 >> 11. Rxe1 >> 14-> 23:40 1.96 2. f5 Nd7 3. Kxg4 Qg2 4. e4 Nf6+ 5. >> Kf4 dxe4 6. Bxe4 Nxe4 7. Nxe4 Qxe2 >> 8. Rae1 Rxe4+ 9. Qxe4 Qxd2+ 10. Qe3 >> Qxe3+ 11. Rxe3 Bd7 (s=2) > >Crafty is very overoptimistic in some lines, though. For instance, play through >the variations in http://www.talkchess.com/forums/1/message=282135 and notice >that for the first few moves in some crafty gives +2 but follow the line and >crafty's score drops bigtime. > >Dave I would never argue that a computer is infallible. However, the above is enough to convince me that it at _least_ needs a lot of analysis before tossing away that kind of apparent advantage and playing into a forced draw.
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.