Ingo Althöfer wrote:
Stefan Reisz is the author of the website
http://www.reisz.de/gohome.htm
There he claims to have a solution for 6x6-Go
with Japanese rules.
This is not a "solution" in a mathematical sense because
- it is not specified which Japanese rules are used
- during the scoring, the rules are applied without showing exactly that
- during the scoring, the number of studied hypothetical-sequences is
zero instead of huge or infinite
- in every move-sequence, the game is not ended by successive passes
properly
- MANY other possible moves are missing (and my manual study of some
simplistic (but arcane) positions under some particular Japanese
rulesets has convinced me that there are more unexpected but correct
plays or passes than one fears)
Too often the word solution is abused. "preliminary study" is more
appropriate.
The largest board for that I could solve Go under Japanese 2003 Rules
manually was 1x1. Already 1x2 was too tough: While it is still possible
to denote all hypothetical-sequences, listing all
hypothetical-strategies is clearly no fun. Possible if one spends
several days or weeks. But if somebody or a program claims to have
solved under some Japanese ruleset, I am more than sceptical and want to
see mathematical proofs. Although I have done preliminary studies of how
to formulate and prove useful propositions, this is work for months. It
doesn't matter whether proving scoring propositions is done manually or
by algorithm. Only those board sizes that allow killing all are simpler
because all you have to do is to prove just that. There are exceptional
tiny board sizes that allow other types of elegant proofs, but they
won't help much for bigger boards.
Solving(!) Go under whichever Japanese ruleset is for the rules experts
rather than for computer go.
Does someone here know of other (documented) attempts
to solve 6x6 Go?
Didn't Erik van der Werf do it under his rules?
--
robert jasiek
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/