Author: Christophe Theron
Date: 23:29:47 04/16/01
Go up one level in this thread
On April 17, 2001 at 01:46:34, Uri Blass wrote:
>On April 17, 2001 at 00:58:23, Aloisio Ponti Lopes wrote:
>
>>This is a game that was played just for fun (1 processor for 2 engines... :-(
>>don't have the money to buy my dual). Time is 2 min. + 12 sec/ move increment.
>>Moves were entered manually.
>>
>>[Event "Top-Ten"]
>>[Site "Ponti Chess Club"]
>>[Date "2001.04.17"]
>>[Round "1"]
>>[White "Chess Tiger 14.0"]
>>[Black "Fritz 6"]
>>[Result "1-0"]
>>[Opening "B23 Sicilian: Closed"]
>>
>>1.e4 c5 2.Nc3 g6 3.Nf3 Nc6 4.a3 Bg7 5.Rb1 e6 6.b4 b6 7.Bb2 d6 8.Bb5 Ne7 9.Nd5
>>O-O 10.Nf6+ Bxf6
>>11.Bxf6 Bb7 12.O-O cxb4 13.axb4 a6 14.Bd3 Nb8 15.Re1 Nd7 16.Bg5 Re8 17.Bh6 Qc7
>>18.Qe2 a5
>>19.Qe3 a4 20.Bb5 Bc6 21.Qc3 e5 22.Ng5 Rec8 23.Qf3 f6 24.Bc4+ d5 25.exd5 Bb7
>>26.Ba2 Qxc2
>>27.d6+ Qxa2 28.Qxb7 fxg5 29.dxe7 Qe6 30.Bxg5 Rcb8 31.Qe4 Re8 32.d4 a3 33.dxe5 a2
>>34.Ra1
>>Kg7 35.f4 h6 36.Bh4 Ra4 37.g3 Nc5 38.Bf6+ Kh7 39.Qc2 Qb3 40.Re2 Qxb4 41.Rxa2
>>Qd4+ 42.Rf2
>>Rc4 43.Qb2 Qd5 44.Rf1 Qc6 45.Rc1 Nd3 46.Rxc4 Qxc4 47.Qd2 Qc5+ 48.Kg2 Qd5+ 49.Kh3
>>Qe6+ 50.Kh4
>>Qf5 51.g4 Qe4 52.Kg3 Qd4 53.h3 Qg1+ 54.Qg2 Qd4 55.Qf1 Qe3+ 56.Qf3 Qd4 57.g5 Nc5
>>58.Kh2 hxg5
>>59.fxg5 b5 60.Qg4 Qxg4 61.hxg4 Ne6 62.Ra6 Nc7 63.Rd6 Kg8 64.e6 Kh7 65.Kg3 Kg8
>>66.Kf4
>>
>>I can't believe it is Fritz in Zugzwang... oh, I see, it's Tiger with the white
>>pieces... ;-)
>>
>>Nxe6+
>>67.Rxe6 Kf7 68.Rb6 b4 69.Rxb4 Ra8 70.Rb3 Re8 71.Rd3 Ke6 72.Rd8 Rxe7 73.Bxe7 Kxe7
>>74.Rd5
>>Ke6 75.Rd2 Ke7 76.Ke5 Kf7 77.Rd7+ Ke8
>>
>>[D] 4k3/3R4/6p1/4K1P1/6P1/8/8/8 w - - 0 78
>>
>>Very interesting, I could not see the reason why Tiger did not play Ke6 here.
>
>The reason is very simple.
>78.Kf6 is forcing a mate by tablebases or mate in 2 if black does not take the
>rook and when the program finds one mate it does not continue to calculate.
>
>Uri
Tiger continues to calculate one additional ply when it finds a mate, and if it
has not found anything better after this additional ply, it stops thinking and
plays the mate (only when time controls modes are active).
It gives sometimes suboptimal solutions, but practical fast end of games.
As a human player, when you have found a way to mate in 6 in a tournament game
after thinking for 10 seconds, you are not going to keep on thinking for 3
minutes in order to find a shorter mate. It is not a very elegant behaviour for
your opponent.
As funny as it sounds, I have tried to program a little bit of courtesy in my
chess engines... :)
Christophe
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.