Computer Chess Club Archives




Subject: Re: Question about static exchange evaluation

Author: Robert Hyatt

Date: 07:51:36 11/14/98

Go up one level in this thread

On November 13, 1998 at 15:53:45, Bruce Moreland wrote:

>On November 13, 1998 at 11:19:19, Larry Coon wrote:
>>On November 12, 1998 at 19:24:04, Bruce Moreland wrote:
>>>On November 12, 1998 at 16:47:22, Larry Coon wrote:
>>>>Yeah, but I'd hate to spend a lot of time developing and
>>>>benchmarking an approach that fails, only to find that
>>>>there's a big, obvious drawback that I'd overlooked.
>>>You are not going to like programming chess, if this kind of thing bothers you
>>:-)  Sorry, I should have been a little more specific, since
>>my professional life has had more than its share of researching
>>approaches that are later found to be bad....
>>Maybe I should have said that I hate being the *second* person
>>to invent the Edsel....
>This may actually be an excellent example, because unless I mis-remember, the
>problem with the Edsel wasn't that it was a bad car, it was that nobody wanted
>to buy it because it looked like, "a Cadillac sucking a lemon".

*NO*.  The Edsel was infamous because *everything* was different on the
car.  *everything*.  IE different starter hardware...  oddball this, oddball
that...  It's claim to fame was being "different".  It was so different it
"died"....  I don't remember all the oddball things this far back, but my
uncle owned one, and everytime he asked my dad to help him fix something on
it, I remember my dad asking "what idiot would put this here?"  Or "what
idiot decided to replace a perfectly good idea with this piece of junk?"

>There are lots of techniques that some people use and a some other people have
>decided don't work.  I think it's great to keep all of this old stuff in play,
>since the result is what I think could accurately be termed cross-pollenation.
>So by all means reinvent the wheel, see if it works, and if it does you can
>either sit on it and an enjoy an advantage because everyone else thinks this
>technique sucks and you know otherwise, or you can tell other people about it
>and they can enjoy the benefits too, or they can try it and find that it still
>sucks, for them, and maybe you both can gather some insights by figuring out
>This is not just blue-sky, it really happens.

I think retrying "the wheel" is good.  I think that spending several years to
develop a basic alpha/beta search is bad.. and that is likely what would happen
if someone starts off 100% cold...

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.