Author: Albert Silver
Date: 06:40:50 10/28/04
Go up one level in this thread
On October 28, 2004 at 02:39:46, Thorsten Czub wrote:
>I think that albert mainly tries to change PD with using all the nice search
>switches ed gives. i have made the experience that the strength of PD is not the
>search, it is its weakness and i think that PD plays lousy with all those search
>switches ON. it makes blunder moves, giving away a good game for nothing.
I tend to really disagree on this comment. I don't think the problem is the
switches. If the switches improve its speed, then it should improve its play. My
view of settings is that it is all a question of balance. A small extra
somewhere can be enough to make it a winner or a loser.
My previous settings lost to List 5.12, my only previous test against it, by
17.5 - 22.5. The difference between that version and this is so small it is hard
to believe, but I'll tell you:
In this version I added the [MISC_39] switch and I reduced the [Pawn Formation]
from 109 to 108. That's *it*. Everything else was left as is. It reeks of black
magic if you ask me, but I think it just struck a balance. Anyhow, the results
are exactly those I've reported, and I'd be happy to share the full games with
anyone requesting.
>From all those programs s8 is the fear-opponent of PD.
It is the feared opponent of pretty much EVERY program. Still, go look at Kurt's
results, and you'll see that List 5.12 was very much PD's nemesis. I also think
that List 5.12 would do very well on the SSDF list if it were to be tested.
>but it is possible to get a good style against s8.
>
>i doubt that the default or alberts style works against s8.
Well, I don't have Shredder 8 (I do have 7.04) so cannot test against it. Also,
when I reported this result, it is because I knew exactly how well (or badly)
the default settings did against the engine in question, so that I had something
to compare with. There would be little point in reporting this result if I then
found out that the default settings did even better. :-)
Still, I'm aware it is only one match, and have already begun a similar match
against Junior 8, which in my personal Nunn 2 tests beat PD default 24 - 16.
Albert
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.