Author: Dann Corbit
Date: 21:51:00 01/06/02
Go up one level in this thread
On January 07, 2002 at 00:37:55, Richard A. Fowell wrote: >After a brief look (2 hours @ 867 MHz, 5 billion nodes) >MacChess thinks pretty much the same, as Tiger from 11 to 16 ply. >The MacChess PV matches the Tiger PV for the first 13 ply. > >[D]r1k4r/pp1n2bp/2p1Q1p1/1q4B1/3P4/2P2N1P/PP3PP1/R5K1 w - - 2 17 > >The MacChess continuation at 16 ply: > >1. Qf7 Bf6 >2. Bxf6 Qf5 >3. Ne5 Nxf6 >4. g4 Nxg4 >5. hxg4 Qxf7 >6. Nxf7 Rf8 >7. Re1 h5 >8. gxh5 gxh5 >9. Nd6+ Kc7 >10. Ne4 > >Log file follows. > >Richard A. Fowell > >==================================================== >05/22|00:00:00| 100901| -276| Re1 Rf8 Qe8+ Rxe8 Rxe8+ Kc7 Rxa8 Qxb2 etc. >05/34|00:00:00| 131718| -254| Qf7 Bf6 Bxf6 Qf5 d5 Nxf6 dxc6 bxc6 >06/01|00:00:00| 181192| -254| Qf7 Bf6 Bxf6 Qf5 d5 Nxf6 dxc6 bxc6 >07/01|00:00:00| 288548| -301| Qf7 Bf6 Bxf6 Rf8 Qxh7 Qxb2 Re1 Rxf6 >07/02|00:00:01| 726355| -270| b3 Rf8 Qe7 Rg8 Qf7 Qd5 Qxd5 cxd5 >07/23|00:00:02| 1356968| -264| Re1 Qd5 Qxd5 cxd5 Bh4 Bf6 Bxf6 Nxf6 Ne5 >08/01|00:00:04| 2621714| -263| Re1 Qd5 Qxd5 cxd5 Re7 Rg8 a3 h6 Bf4 >09/01|00:00:07| 5004796| -263| Re1 Qd5 Qxd5 cxd5 Re7 Rg8 Bf4 h5 Ne5 Bf6 >10/01|00:00:14| 10011393| -263| Re1 Qd5 Qxd5 cxd5 Bf4 Rg8 Re7 h5 Be5 Bh8 b3 >11/01|00:01:06| 45879433| -244| Re1 Qd5 Qxd5 cxd5 Re7 Rg8 Bf4 h5 Ne5 Nb6 etc. >11/02|00:01:21| 56292719| -169| Qf7 Bf6 Bxf6 Qf5 Ne5 Nxf6 g4 Qf4 Qe6+ Kc7 >Qf7+ Kd8 Qxb7 >12/01|00:02:06| 87445375| -130| Qf7 Bf6 Bxf6 Qf5 Ne5 Nxf6 g4 Nxg4 hxg4 Qxf7 >Nxf7 Rf8 Nd6+ Kc7 Ne4 >13/01|00:04:33| 189896369| -125| Qf7 Bf6 Bxf6 Qf5 Ne5 Nxf6 g4 Nxg4 hxg4 Qxf7 >Nxf7 Rf8 Re1 h5 Nd6+ Kc7 >14/01|00:12:57| 540920201| -130| Qf7 Bf6 Bxf6 Qf5 Ne5 Nxf6 g4 Nxg4 hxg4 Qxf7 >Nxf7 Rf8 Re1 h5 Ne5 Re8 >15/01|00:38:35| 1580408463| -132| Qf7 Bf6 Bxf6 Qf5 Ne5 Nxf6 g4 Nxg4 hxg4 Qxf7 >Nxf7 Rf8 Re1 h5 gxh5 gxh5 Nd6+ Kc7 >16/01|02:05:58| 5029329783| -131| Qf7 Bf6 Bxf6 Qf5 Ne5 Nxf6 g4 Nxg4 hxg4 Qxf7 >Nxf7 Rf8 Re1 h5 gxh5 gxh5 Nd6+ Kc7 Ne4 This is the exact pattern we expect to see with a really good positional move. A steady, rapid increase in the ce with increasing depth. This position is very difficult and volatile. The analysis by crafty in the other thread that I posted showed that the players did not always follow the predicted pv, and when they deviated the analysis showed their choice to bet better at times than the predicted move. I have found a good way to look ahead on positions like this is to set the time control to 2 hrs per move, set the board to the position in question, and then tell yace the "auto" command. After running all night, you will get a pretty good idea if it is sound or not. Of course, to really know about the game, you will have to repeat this task for every position which forms a question in your mind. Rather time consuming.
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.