Author: Robin Smith
Date: 15:47:03 03/19/05
Go up one level in this thread
On March 19, 2005 at 12:05:54, Eelco de Groot wrote: >On March 17, 2005 at 23:01:22, Robin Smith wrote: > >>Several years ago I posted the position below, which at that time no program >>could solve. I have discovered that Shredder 9 can now find the correct move in >>about 4 hours on my old Athlon MP2000+. How do other programs do? >> >>[D] r1qr2k1/pb2bpp1/1p2pn1p/2p4P/3P1B2/2PB1N2/PP2QPP1/1K1R3R w - - 0 15 >> >>15.Rhg1! >> >>-Robin > >This does seem a difficult tactical problem. I tried it yesterday with a >tactical style Q5T in Pro Deo, but it does not play Rhg1 on its own. Maybe after >4 hours or a faster machine it can find that move, I stopped after about 30 >minutes at ply twelve. After entering 1.Rhg1 Nxh5 the score goes up to +0.95 at >ply ten, 30 minutes on Celeron 500 MHz. I suppose 1.Rhg1 Nxh5 2.Bxh6 is the >key-line to be found? Yes. After 1.Rhg1 Nxh5? 2.Bxh6 black is completly lost, but the lines are quite deep and still not easy for computers to find. If black does not capture on h5 he might survive, but White's chances are still much better. > At least that is what Q5T comes up with after 1.Rhg1. >"Quiet" move 1.Ne5 that Q5T chooses also does not seem without chances though? I >haven't tried 1.Rhg1 b5 that Fritz gives. Fritz's 1...b5 is interesting. I have not looked at it deeply but I think White must still be much better since his pieces are generally better positioned for a quick attack. 1.Ne5 is also promising, however I am pretty certain that Rhg1 is stronger, since g1 is the most natural square for the rook, while it is not yet clear where the knight should reside; plus the fact that both Fritz and Shredder eventually settle on Rhg1. <snip> >After 1.Rhg1 Nxh5 <snip> >00:30:05.7 0,95 10 274919662 Bxh6 gxh6 Qe3 c4 Bc2 > Eelco You might try Pro Deo in the position after 2.Bxh6 gxh6 3.Qe3 c4 4.Bc2. I think it will quickly become clear that black is in big trouble. -Robin
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.