Computer Chess Club Archives


Search

Terms

Messages

Subject: I'm seeing a "pattern" with your posts

Author: Russell Reagan

Date: 15:21:59 05/17/02

Go up one level in this thread


Well once again you've done a wonderful job of stating the obvious for all of
us, and once again you're not going to get a ton of help on understanding this
topic, because the large populatoin of chess programmers don't have any answers
for you. What you are asking here, and have been asking in your other posts, is
currently over the head of everyone but a handful of people in the world.

The "pattern" I'm beginning to see with you is that you think about some great
solution for chess programs but you haven't the slightest clue how difficult it
is to implement what you propose.

In another post you talked about how we need to make programs learn how to plan
better, and now you're talking about making chess programs recognize patterns
and use them as well as humans do.

Both of these are great ideas, and they've been around for decades. The problem
is that it's extremely complex to even know where to approach the problem from.

Before you throw out any more great ideas, please think about them yourself.
Think about how realistic it is. If you can possibly conceive of some method of
implementing one of your great ideas, then add that into your post and ask "what
do you think about this?". You'll get more responses, even if it's people
telling you that your idea won't work, and you'll at least get people thinking.
As far as I can tell you just declared some information in this thread's
original post without really wanting anything by way of response.

So next time you have a great idea of something like, "Let's just create a 100%
perfect evaluation function, and then we only have to search 1 ply!", use your
brain and think about how realistic that really is. You give off this attitude
like you're an expert on anything, and like you're the first guy to ever think
about something new in AI.

New ideas aren't the problem. Anyone can think of new ideas (even though your
ideas aren't new). The problem is implementing a working model of the idea. So
before you post some other vague topic like, "Let's just create the perfect
chess program! It's that simple!", remember that there are many people in this
world that are a great deal smarter and more clever than you are, and that your
ideas have probably been thought about before, and there is probably a reason
why no one has been able to "just make [insert whatever great idea here]".

Propose something yourself, or ask where you could find more resources on this,
or something like that. Don't just say, "Let's make a program that plays
perfectly! You guys do all of the work and let me know when you're done!". That
is the attitude that you give off, as if you throw out an idea and you want
everyone to go test it for you and come back and let you know how it went. I
don't think anyone ever achieved much success in any new frontier without a lot
of hard work. Ask one of the guys who has a great commercial chess program how
much work they've put into it. You'll either find that they've put in decades of
hard work into it, or they've put in at least 5-6 years of extremely hard work,
almost like a full time job. If you put in hard work, people will see that and
be more willing to help you out if they know that you're a good hard working guy
who will give them some results back in return.

Just my advice if you plan on continuing to post here.

Russell



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.