Re: About brute force and knowledge...

[Robert Jasiek <jasiek@xxxxxxxxxxxxxxx>]

[...]
> The hypothesis is that P may
> look unreasonable or ununderstandable even to a highest-ranking dan
> player.

I would state the same.

[...]
>>
 "simple moves
 win" or something like that. [...] the reason can be that the more
 complex variations you initiate, the more probable it is that either
 player will plumber, and thus the game becomes more random for human
 players.
<<

Agree, although I sometimes initiate complex situations in the hope
to be less confused than my opponent...

>>
 Thus, it is possible that the reason why go programs should incorprate
go
 knowledge, understanding of shape for example, is not that it would
lead
 to perfect or better play, but that it would reduce the state-space to
 search (by brute force if you wish) to a space considered by human
 players.
<<

This is an important aspect.

> The point is, it can be that humans play go after all in a very simplified
> way. Horrible positions can perhaps be constructed where best players
> don't have a clue what to do, but they never arise in a game, because
> humans prefer to simplify things.

Complex kos, e.g. Only certain strong pros manage to use them well
as a means of confusion. Also some Japanese traditionalists argue
that they would not like to handle additional complexity given by
superko. As a rules theoretician I know that one can contrive
ko positions that professionals can hardly handle. E.g. Bill Spight
+ Co have created a demanding theorem on n-tuple-ko multiples. For
different n[i]-tuple kos everything is an open problem. Similar
things have been shown for 1-point endgames (see Mathematical Go 
Endgames).

> Thus, perhaps the complexity of go as a game-three searching problem is
> not so overwhelming as it looks, if we just take it for granted that
> humans usually play only moves of certain type, usually explainable by
> some heuristic concepts instead of an enumeration of 10^10 variations.

Oh, this a surprising approach. One might try to analyse the game tree 
itself (instead of the corresponding board positions, as I have
suggested
in my last article) as a physical space, set models of physics or 
probability for it, and operate tree algorithms accordingly. Not that I 
would know how to start, except maybe from endgame positions.
Nevertheless,
one could use the fact, that general game tree algorithms are a
fashionable
research area.

-- 
robert jasiek
http://www.snafu.de/~jasiek/rules.html