Author: Christophe Theron
Date: 11:13:32 12/20/01
Go up one level in this thread
On December 20, 2001 at 13:47:00, Scott Gasch wrote:
>>>
>>>Commercial engines are not using any "unknown" board representations or search>>techniques.
>>
>>How do you know?
>
>By "unknown" search techniques I was replying to the question of the original
>poster. He specifically asked if pros were using alpha-beta or a variation
>thereof. As far as I know every professional engine is using a..b.
>
>>1) I think my board representation is original. I have never found it anywhere
>>else. It's not 100% original, but what makes it really effective is definitely
>>something I have never found elsewhere.
>
>My board representation is orignial too... but its a lot like 0x88. My point is
>that there is no magic bullet board representation. There is no "secret" board
>representation that gives the pros an edge. I'd argue that what board
>representation you use, as long as it's not totally slow/dumb, has relatively
>minimal impact on your engine strength. Everyone is going to roll their own
>board representation or put their own spin on an existing one.
OK, I understand what you mean.
I partially agree with you. The gain from a specific board representation can be
significant, but not huge.
>>2) My program as well as other commercial engines are using search techniques
>>that have never been published.
>>
>>>Perhaps some are using forward pruning techniques that are not
>>>published anywhere. The degree to which this affects their playing strength >>is debatable.
>>
>>No it's not. It makes commercial programs clearly stronger.
>>
>>>I'd be surprised if there was another technique like nullmove out
>>>there but it wouldn't be the first time I was wrong...
>>
>>There are plenty.
>>
>>Null move is to chess what MacDonalds is to food. Quick and easy, not too bad,
>>but definitely not the final say.
>>
>
>So you think that the strength of the pros comes from the pruning techniques
>they use? Meanwhile you have people like Vincent who will tell you that forward
>pruning is the ultimate evil.
And you believe him? You should be more careful in choosing your information
sources.
> And you have people like Bob who have a ton of
>computer chess experience and just stick with nullmove.
...and is constantly 100-150 elo behind the top commercial programs.
> So this is what I mean
>by issue of forward pruning being debatable.
I don't find it debatable because I know it is possible to do better.
Christophe
This page took 0.07 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.