RE: [computer-go] Pattern Matcher

["Mark Boon" <tesuji@xxxxxxxxx>]

> -----Original Message-----
> From: computer-go-bounces@xxxxxxxxxxxxxxxxx
> [mailto:computer-go-bounces@xxxxxxxxxxxxxxxxx]On Behalf Of Frank de Groot
> Sent: Monday, November 08, 2004 10:57
> To: computer-go
> Subject: Re: [computer-go] Pattern Matcher
>
>
> > > > Interesting. Can a whole game board be broken down into
> patterns and a
> > >valid evaluation function can be obatined based on these patterns?
> > >
> > >No, I am sorry. I made it all up.
> > >I guess the game is over for me now, I have beem exposed for
> the fraud I
> am
> > >:(
> >
> > I don't understand this response.  I think he was asking an honest
> question
> > but you must have interpreted it wrong (or maybe I did).
>
>
> I had a few issues with this question, and after the barrage of "your shit
> stinks" it was a little too much for me so instead of becoming impolite, I
> attempted humor.

Frank, I think you're exaggerating here, which seems to cause you to
overreact. Nobody said something remotely like "your shit stinks". Some, me
included, have questioned whether your program (or pattern-matcher, whatever
you want to call it) would have value in a Go playing program. If you read
carefully you'll see I never questioned whether or not your program is worth
$100,000 or any other amount. I've only expressed my opinion that it's less
valuable to a Go playing program than a good life-and-death module. This
because I think it's harder and more work to make.

I share your opinion that entering patterns by hand is ultimately not the
solution, and that automatic pattern generation has the future. However, I
do not share your opinion that a 'one approach fits all', generated from pro
games, is the answer. Instead I think smaller sets of specialised patterns,
used for specialised tasks, will work much better. And they will be
generated based on fundamental knowledge of the program, not based on a set
of pro games.

The reason for that is that I think move-selection is only as strong as it's
evaluation function that has to decide whether the candidate was any good.
Generating candidates by itself is not that hard, it's the ordering or
pruning that makes the difference. As long as the best move is in my list of
candidates. Maybe your approach will do the initial ordering a little better
than a more simple approach, but I think that chess has shown that the
initial ordering is not that important. The main cut-off comes from
alpha-beta, which relies on the evaluation, not the move order.

The only way your approach is going to be better is if your patterns include
Go knowledge such that you can reliably throw away the bad candidates and
not look at them at all. And do this without evaluation. Correct me if I'm
wrong, but I understood that this is exactly what you claim. But if that's
the case, then I don't see why you need evaluation and lookahead at all. You
pattern-matcher will always only suggest pro-quality moves. Didn't you say
you were able to do this reliably for the first 100 moves or more? Then may
I suggest a simple experiment? Play a game against another program, say GNU
Go. Let your pattern-matcher suggest the best move in each position and play
it against GNU Go. If you're right, then after 100 moves your program should
be miles ahead. If your Go level is not sufficient for such an experiment,
ask a strong Go player for help.

Now if your program indeed shows such good play during the first 100 moves,
then I have no problem to admit I was wrong. And in that case I think your
program is worth much more than $100,000. If the position is more or less
equal after 100 moves, then I think it doesn't mean much yet apart from that
your claim seems wildly exaggerated. If the position of your program is much
worse ater 100 moves, then I think you'll have to admit your program is
probably not worth more than any other plain pattern-matcher  when it comes
to a Go playing program. It could still have good value as a study tool, but
that's not what we're discussing here.

No need for all this animosity, just show us. If you say it takes another 6
months, no problem, we're not going anywhere in the meantime.

And please refrain from saying people here are just jealous. It's childish
and only makes us think you are the jealous one of the money that some of us
here seem to have made with their software.






_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/