Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF(Junior 6 - MCP8)AMD K6-2 450, 2.5-.5, ended 25-15.

Author: Chessfun

Date: 01:51:33 07/01/00

Go up one level in this thread


On June 30, 2000 at 08:03:48, blass uri wrote:
>
>This is a problem in the learning function.
>I found that it could draw some games in the B87 variation for example it drew
>game 30 but could not repeat the same line in game 38.
>
>game 30
>
r4r1k/6pp/3Q1b2/4p3/8/4B3/q5PP/3R1RK1 w - - 0 1
>
>mchesspro8 drew by 26.Bc5 and in game 38 in the same position mchess played
>26.Qb4 and lost
>
>I also do not understand the reason for repeating the same line that give white
>a position with a pawn down.
>
>The question is if it can happen without the autoplayer(I do not 100% trust the
>autoplayers and if I see programs do something stupid one of the possible reason
>that I think of is the autoplayer).
>
>Uri



Game 14 =    26. Bc5
Game 18 -    26. Bc5
Game 26 =    26. Bc5
Game 28 =    26. Bc5
Game 30 =    26. Bc5
Game 34 =    26. Bc5
Game 38 -    26. Qb4

The interesting thing is that the 26. Bc5 variation
comes up a number of times and that even after losing
game 18 it stayed with it in Game 26. And it was infact
Junior that varied on it's 28th move after winning the previous
game 26. It seems more surprising that MCP8 did not vary with
15. Rab1 as it lost the previous game and on mine thats the move
it wants to play.

In this B87 opening MCP8 is out of book after Black Juniors 14....Nd7
it lists 1 book moves for black Qc7. Fritz 6 gen.ctg lists
2 choices preferred Qc7 and lists Nd7 as !?. though at the
end of Fritz's line after Nd7 it scores as equal. And it has
only a minimum number of games along this line.

Now it gets interesting.
Junior in Junior 6 jbook.ctg only lists one move and that move
is 14....Qc7 not as played in all these games 14....Nd7.
But it's book only lists one game so in analysis it chooses
14....Nd7. as when book is optimized minimum is 2 games.

At this point in time after 14....Nd7 my MCP8 will not play
15. f5 but after 600 secs it chooses Rad1 I watched the eval
in both play mode and analysis mode and it never considered f5.
It scores this 15. Rab1 as -1.03 and scores f5 as -1.18.
Though Junior as can be seen by it's eval expected 15. f5.

Thanks.




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.