Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Star Diamond's battery life data and a guess as to its rating constant

Author: J. C. Boco

Date: 10:05:07 03/11/04

Go up one level in this thread


On March 10, 2004 at 09:22:09, Peter Berger wrote:

>On March 10, 2004 at 08:29:58, J. C. Boco wrote:
>
>>Oh, By the way, I'm a bad Class C player.  I have played 40 moves in 2 hours.  >I
>>have set the computer to play mostly at 40 moves in 2 minutes, but sometimes it
>>had up to 5 minutes and sometimes as low as 40 seconds.  My record:
>>
>>0 = 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0
>>
>>Also interesting, if I use as a constant the number 100 in the Log equation
>>determining playing strength for a computer (100 points per doubling), then
>>based on the various time controls I have played my TPR for all these games
>>would have been 1380, only about 35 points lower than my real USCF rating.
>
>I think that's really a _very_ fine performance for a 1400 player against the
>computer. Did you by chance save the moves of your two wins? It would be
>interesting to be able to have a look at the games.
>
>Peter

I always record my games. If you really want, I'll type in one of the wins (only
1 as you'll see...)

The earned victory came with the Star Diamond playing a Dragon by transposition,
and since it was by transposition there was a meaningless move the computer had
inserted which didn't follow the Dragon theme.  It "sacrificed" a piece for a
harsh attack.  Or, you could say it blundered a piece because I navigated
through the attack and emerged up 2 pawns.  Anyway you look at it, I was up
material and I won the game.  Anyone can win 2 pawns up, right?  Well......

One of my losses "should" have been a win.  I was black in a Gruenfeld and
something didn't look quite right.  I spent 20 minutes trying to see how my
pieces could work together and found a move that won material, found only
because I insisted that I take advantage of the fact the computer played b3 yet
in capturing one of my pawns the bishop was not fianchettoed by on the f4
square.  A very proud moment for me.

Then despite that fact I was up 2 pawns, I wanted to adopt a new philosophy in
my play:  more dynamics.  I purposely dropped a pawn so that several moves down
the road I'd have a rook on a semi-open file.  I did so.  I then sunk a knight
into a weak square (supported by that rook) and right in front of a backward
weak pawn.  My knight stood on c3 and all was right with the world.  God was in
his heaven, the birds in the air and the snail on its thorn (the saying goes
something like that).

I ended up losing the game because I was so proud of my knight there I didn't
realize that to make progress it had to trade itself off for the very very bad
bishop to my rook could blast through.  A very great lesson learned in that
game.

To even things out -- chess justice as it were -- to balance out loss which
should have been a win, my other won game came from a dead draw.

3 pawns each on the Kingside (locked up), and I had 1 pawn on the queenside
directly in front it the Diamond's doubled pawns.  I was a pawn down.  In fact
being a pawn down is how I won the game!  My king was in the center and the
computers pushed back.  It was a simple case of my not allowing his king to come
out.  Very simple, although it seemed a little hard after 2 hours of play and
the clock ticking...  It kept repeating moves, and then when repitition wouls
have occured it moved it's king to another square.  Still a draw.  It repeated
moves again.  Then, finally, I'm sure the computer "thought" that since it had a
positive evaluation, it shouldn't repeat moves and go to zero.  So it moved it's
king to another square which a 10 ply search would have seen loses.

I always play with permanent brain on.  If I spent time thinking it would have
searched 10 ply.  But all I had to do was shuffle my king back and forth so I
didn't give any of my time for the computer to think.

I consider this my only psychological victory over a computer!!!



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.