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

[Arend Bayer <arend.bayer@xxxxxx>]

On Mon, 29 Nov 2004 Compgo123@xxxxxxxxxxxxxxxxx wrote:

> In a message dated 11/29/04 2:30:55 AM Pacific Standard Time, 
> frank@xxxxxxxxxxxxxxxxx writes:
> 
> 
> > I wonder at what stage in the development of a Go program, and at what
> > strength, Go knowledge becomes more important than hacking ability.
> > 
> > I even wonder whether Go knowledge is very important at all.
> > 
> 
> You are so behind on this issue that you don't even know that you don't know. 

I really wish we could get back to discussions on this list without
statements like this. 

> >None of the Go programmers have ever given any arguments as to why you can't
> >extract Go knowledge from game records, and be better than "manual" Go
> >knowledge. I have given a working example for "shape" and theoretical
> 
> Just to set the record straight. Automatic extraction of Go knowlege from 
> game records has been discussed on this list for many years. Pattern harvest is 
> one of them. It's more difficult than many people think to come up a new idea.

Jusgt to set the record straight, here is a list of tasks in order of
difficulty, according to my opinion:

1. Discussing an idea on computer-go
2. Having an idea
3. Implementing it well
4. Pushiing it furhter than anyone else has done before in a good
implementation

I think Frank may have achieved no. 4 wrt harvesting patterns from
professional games (I don't know, since GoPP may also be very good at
this).

Arend

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