Author: Robert Hyatt
Date: 19:03:41 11/16/03
Go up one level in this thread
On November 16, 2003 at 21:29:52, Christopher A. Morgan wrote:
>Bob,
>
>As you posted in one of the replies below, Crafty analyzed the ending position
>in game three of the X3D match, after white’s Rb3, about +2.6, and even before
>the end that white had a far superior position. Uri, I think it was, posted
>analysis of other engines and most were +1 or less. Results I got for depth of
>12 ply were as follows: Crafty 19.03 - +2.68, Shredder 7.04 - +0.95, Fritz 8 -
>+0.44, CT14 - +1.24, and for depth of 15 for Junior 7 - +1.05. I think we have
>to assume that the new deep Fritz running in a 4X2.8 configuration also came up
>with a similar analyses that it was lost, probably +2.5 or worse. Tony Rook on
>Chess FM radio said 4-5 moves before the end that Mathias had said Fritz would
>not resign unless the analysis was that much in favor of GK (white).
>
>There was no doubt among the grandmasters that Fritz (black) was toast, with GM
>John Fedorowicz on Chess FM radio even going so far to wonder whether GK would
>be getting upset that Fritz was playing on in a completely lost position.
>Obviously your Crafty was one of the few engines to analyze the position
>properly. I congratulate you on Crafty’s fine analysis. What do you attribute
>Crafty’s analysis superiority in this position to compared to the top commercial
>programs? Thanks.
>
>Chris
I'm not sure I can answer. I know what Crafty does internally. I have no
idea what the "top commercial programs" do. Remember, I've been tuning Crafty
on ICC for _years_ to try to stop the "draw masters" that play there. They
fight for just such positions as Kasparov achieved. Fritz didn't have to
allow it. But it seems to not understand the concept of "blocked pawns",
or coversely, it doesn't seem to evaluate potential pawn levers to make sure
that it has a "useful break" rather than a blocked position where the opponent
gets to call the shots.
Note that for most of the game my eval was only +1. But at the end, here is
what I was seeing in the position where Fritz resigned:
14 38.81 1.54 45. ... Ne8 46. Nb4 Rbb8 47. Rba3 Rxa3
48. Rxa3 Nc8 49. h4 g6 50. Ba6 Ne7
51. Kb1 Bg7 <HT> (s=5)
14-> 1:25 1.54 45. ... Ne8 46. Nb4 Rbb8 47. Rba3 Rxa3
48. Rxa3 Nc8 49. h4 g6 50. Ba6 Ne7
51. Kb1 Bg7 <HT> (s=3)
15 1:35 1.60 45. ... Ne8 46. Nb4 Rbb8 47. Rba3 Rxa3
48. Rxa3 Nc8 49. h4 g6 50. Ba6 Ne7
51. Qe2 Bg7 52. f4 Ra8 53. Rb3 (s=2)
15-> 4:35 1.60 45. ... Ne8 46. Nb4 Rbb8 47. Rba3 Rxa3
51. Qe2 Bg7 52. f4 Ra8 53. Rb3 (s=15)
16 5:47 -- 45. ... Ne8
16 6:43 2.03 45. ... Ne8 46. Nb4 Rd8 47. Rba3 Nf6
48. Ra7 Rxa7 49. Rxa7 Qe8 50. Ba6 Nd7
51. Rc7 Nb8 52. Be2 Rd7 53. Ra7 Bg5
(s=14)
16 18:12 2.02 45. ... g6 46. Nc7 Rc8 47. Ba6 Rbxc7
48. bxc7 Rxc7 49. Rb8+ Nfg8 50. Ba5
Ra7 51. Bb6 Rxa6 52. Rxa6 Qxh3 53.
Kb2 Qf5 54. Ra7 (s=10)
16 28:34 1.95 45. ... Rd8 46. Nb4 g6 47. Kb2 Qf5
48. g4 Qe6 49. Ba6 Rbb8 50. Kc1 Bg5
51. Kd1 h5 52. gxh5 gxh5 53. Bf1 Rb7
(s=6)
16-> 40:27 1.95 45. ... Rd8 46. Nb4 g6 47. Kb2 Qf5
48. g4 Qe6 49. Ba6 Rbb8 50. Kc1 Bg5
51. Kd1 h5 52. gxh5 gxh5 53. Bf1 Rb7
(s=7)
17 47:25 2.03 45. ... Rd8 46. Nb4 Nfg8 47. Kb1 Qf5
48. f4 Qe6 49. g4 g6 50. Qf2 Rbd7 51.
Be2 Rb7 <HT> (s=6)
That's not a place where _I_ would have resigned, for the record...
This page took 0.01 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.