Re: [computer-go] Pattern matching - example play

[Vincent Diepeveen <diep@xxxxxxxxx>]

You can't compare backgammon to all the efforts that gets put in go & chess.

Backgammon is a statistical game with only a few players who play online.
The few that play online are usually rich guys who play for money against
each other. 

Further backgammon from software respect has not much commercial value, let
alone sportive challenges there are left there.

Did you ever play the game?

It works with dices. Dices matter. If you are unlucky, you can lose even
from an utmost beginner. 

Not if you play 100 games against him. But if you play for money you
usually play only a few.

Are you used to play for money?

Well i am.

So we cannot take backgammon serious at all.

there is a few professional players there who earn a lot of money playing
rich fellows and a very generous organized world championship.

all entry fees of such tournaments are very high. $250-$500 to enter a
single day tournament is no exception.

It's one big commercial show. 

Making software for it has been done already by several.

There is not a single sportive challenge for software in backgammon other
than making a better interface for it.

No wonder ANN works good in it.

In Netherlands we have a saying: "one eyed is king in the land of the blind".

If you can't make money from selling a backgammon engine, then organize a
backgammon tournament where i am allowed to join with a computer and make a
first price $30k.

Then we can talk and we will see whether your ANN enhanced backgammon
engine searching 3 ply can win from something a good engine programmer
produces.

There is zero challenges there.

See my point?

Claiming ANN works in a non game playing challenge, as backgammon is about
statistics, not about game playing as the goal is just too easy, and use
that as a proof that it works in game play is wrong.

At 12:17 4-12-2004 +0100, Heikki Levanto wrote:
>On Fri, Dec 03, 2004 at 10:27:58PM +0100, Vincent Diepeveen wrote:
>> 
>> >I would not completely write off higher-level planning and neural nets,
>> >and other fancy theories. Many of them have shown their values in
>> 
>> We can have lengthy discussions, but the majority of ANN top researchers
>> agree with me here that for game playing ANN is completely useless.
>
>Well, as far as I know, neural nets work quite well for backgammon. I
>admit it s a very different game, but shows that ANNS are not
>*completely* useless for *game* *playing*. 
>
>I agree with you, that if you just feed the board position into an ANN,
>you can at best train it to recognize similar games, and you will not
>anywhere. But there are other ways to use ANNs.
>
>For example, if you feed it the number of groups, the number of their
>liberites, and so on, it should be relatively easy for it to learn that
>a position where groups are connected is a better one than a situation
>with many small isolated groups. And if you feed it the results of some
>influence calculations, it should easily learn that the one that has
>most territory is often ahead. Both of these are easier to program in
>hand, I admit. But the same ANN could also learn to balance these
>separate considerations, and a few more, and come up witha  decent
>evaluation function.
>
>> Yet no one is stopping you believing in a black box that can solve the game
>> without much calculation. A formula X based upon a few artificial neurons.
>
>Of course I do not believe in a magic black box! But I believe that as
>things are now, it is not yet sure which technique(s) are the best
>suited for solving computer go. You advocate a deep global search, with
>some good arguments, I give you that. I am still not convinced that it
>is the only way, or even the best way to go. Sure it is worth studying,
>but being as problematic as it is, I believe progress can be made in
>other directions too. 
>
>
>Best regards
>
>  Heikki
>
>-- 
>Heikki Levanto   "In Murphy We Turst"     heikki (at) lsd (dot) dk
>
>_______________________________________________
>computer-go mailing list
>computer-go@xxxxxxxxxxxxxxxxx
>http://www.computer-go.org/mailman/listinfo/computer-go/
>
>
_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/