Skip to main content
9 events
when toggle format what by license comment
Dec 27, 2020 at 22:32 comment added flame @GabrielC.Drummond-Cole Ah, I misunderstood your question. The format you gave is fine, since I too suspect the answer won't be nice.
Dec 27, 2020 at 21:48 comment added Gabriel C. Drummond-Cole @flame The output of the algorithm for finite red-blue hackenbush positions is a dyadic rational number, but those positions have at most one canonical option for each side; the values of green-red-blue hackenbush are, I expect, generally not expressible in terms of common game values. I've given a complete (albeit recursive) description of the canonical form of a red-blue-green stalk. If that is not in a satisfying format, what would you want the output to look like?
Dec 27, 2020 at 19:37 comment added flame @GabrielC.Drummond-Cole Similar to an algorithm for evaluating red-blue Hackenbush positions. Though I suppose one doesn't exist yet (and it may be impossible to come up with a nice algorithm in general), according to my research.
Dec 27, 2020 at 18:22 comment added Gabriel C. Drummond-Cole @flame What would you want the output of such an evaluation to look like?
Dec 27, 2020 at 17:43 comment added flame Thanks for your replies! Is there a known way to evaluate red-blue-green stalks, like there is for red-blue stalks and green stalks?
Dec 27, 2020 at 13:40 comment added Gro-Tsen The following question (concerning forests, not just stalks) is at least somewhat relevant: mathoverflow.net/q/267112/17064
Dec 27, 2020 at 8:54 comment added Gabriel C. Drummond-Cole Assuming finite stalks, the moves that survive in canonical form are all green moves (for both sides) and the highest blue (respectively red) edge if and only if there is no green edge further up the stalk from it. I'd be surprised if there were a nice description of general stalks in terms of common elementary game values.
Dec 27, 2020 at 6:40 review First posts
Dec 27, 2020 at 7:48
Dec 27, 2020 at 6:37 history asked flame CC BY-SA 4.0