Skip to content
Little GolemMake your next move.

Go complexity

Public forum discussion.

Go

#51 · gamesorry

9 Oct 2014

@wanderer_c Thanks, yeah the first term is harder than I imagined, especially in case #6 which is a little counter-intuitive ;-)

#52 · richyfourtytwo

19 Oct 2014

Meanwhile it has occurred to me that the entire thread has ignored superko. If that rule is taken into account the mathematical position of the game in the game tree is not alone defined by the positions of the stones, but also by the positions that lead up to the current positions, as these restrict future legal moves. In that sense I think all numbers we found for 3 or more stones are wrong! :-)

#53 · William Fraser

19 Oct 2014

@richyfourtytwo

Wow, what a thought!

However, I think we can safely define the problem either way and this way is definitely more interesting (allowing passing is also interesting, but I would ignore it -- after all, there are historical games where someone was forfeited for passing, so it could be considered "illegal").

#54 · hyperpape

19 Oct 2014

I believe "position" is used to designate an arrangement of stones, without reference to how it was achieved. E.g: positional superko vs. situational superko. See senseis: http://senseis.xmp.net/?Position, (though the page seems to treat "position" as including some illegal "positions", which I wouldn't do. Not sure how people who study the game's combinatorial properties talk about that.

#55 · fhourplay

19 Oct 2014

> though the page seems to treat “position” as including some illegal “positions”, which I wouldn’t do

There are 3^361 positions which can be either legal or illegal. It would be cumbersome to talk

about legal positions versus illegal arrangements, so we equate the notions of arrangement and

position. Every legal position can result from a legal game, that can include passes.

Last page