Re: [gnugo-devel] strategy

[Martin Girard]

On Sep 12, 2004, at 19:48, Arend Bayer wrote:

> > > What people see as a lack of strategy is, im my opinion, mostly just 
> > > a
> > > consequence of GNU Go's unprecise (to say the least) assessment of 
> > > weak
> > > groups.
> >
> > WRONG.
> >
> > It is a wrong assesment of the value of influence and other intangible
> > factors.
> > It is inability to make moves consistent with each other.
> > It is inability to detect distant threats.
> > It is inability to read through the opponent's moves' meaning and 
> > foil the
> > opponent's plans.
> >
> > Again, I encourage those who haven't already to read the following 
> > book:
> > "The Thirty-Six Strategems Applied to Go" by Ma Xiaochun
> > Also, there is a book about the flow of stones that is somewhat 
> > relevant here:
> > "The Direction of Play" by Takeo Kajiwara
> > Finally, the games of Kato Masao are good examples of applying 
> > stragegy to
> > killing groups in the center. He wrote a book on the topic (out of 
> > print,
> > unfortunately): "Kato's Attack and Kill"
> >
> > If after reading the above you still can't figure out why strategy is 
> > of
> > utmost importance, then you should quit playing Go.
>
> My claim is that a computer should not be taught this explicitly. That
> a computer program will automatically not show the weaknesses you
> describe above, once it
> - has a good estimation of weakness of groups
> - much better life and death reading
> - global look-ahead.

Again, this is only part of the answer. You fail to see the whole 
picture when you claim all it requires is a simple patch. Even with 
this, there is no way GNU Go could go beyond 1d, even by reading 50 
moves ahead. It's mathematically impossible.

> You are trying to teach GNU Go in the same way thay one teaches a human
> kyu player.

Indeed.

> Your arguments are very similar to arguments about chess programs from 
> 20
> years ago, that they will never be able to play chess well, since they
> have no idea of strategy. When in fact they just don't understand
> strategy the way humans understand it.

Chess is a very dumb game when it comes to strategy. If you study 
Chess, the first thing you will realize is that simplification is 
central to every game; somehow, you need a way to get the better of the 
unavoidable massive slaughter that will take place. That is why most 
professional games are won through pawns being promoted to queens. One 
does not need to look very far ahead to solve such problems -- eight to 
ten moves are usually enough and advantage is relatively easy to 
determine. Chess is a game about mass slaughter and hardly involves any 
other strategic element.

I have read Samuel Reshevsky's "The Art of Positional Play", which is 
about the only strategy Chess book I've seen. Read it, then compare 
with the books I suggested in my previous reply. It's like night and 
day. Similarly, one shouldn't compare Chess to Go; it amounts to 
blasphemy.


> And btw, even among human players, let's say in games between two 2ds, 
> I
> think very many strategic mistakes result from misreadings or 
> misjudgements
> of group safeties/shape problems.

Computer programs are still doing much worse overall.

If I may add, humans and computer Go players don't make the same kind 
of mistakes. Usually, a human player will fail to accurately assert 
tangible value, to consider unusual moves, or will lack vigilance 
through distraction; a computer program never would. On the other hand, 
abstract reasoning allows a human player to read far deeper than the 
average computer program, as well as to read his opponent and adapt his 
play.

--Martin

> Arend
>
>
>
> _______________________________________________
> gnugo-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/gnugo-devel