Apologies for the ludicrous line-breaks there. Apparently my mail program inserts line breaks according to where the lines happen to wrap in the window you're editing the message in. How bizarre. Here's the exact same text with (I hope) more sensible line breaks. On Thursday 15 April 2010 22:09:00 Allan Wechsler wrote:
With pen and paper I've managed to produce the Nim values for the game with between 1 and 6 disks. The values I get are almost certainly a mere tease: 0, 1, 2, 0, 1, 2. I don't expect that pattern to continue.
It apparently doesn't. A maximally-stupid program in a slow language gives me the following, for n from 0 upwards: n | 0 1 2 3 4 5 6 7 8 9 A B C D --+---------------------------- | 0 0 1 2 0 1 2 3 1 0 3 2 4 1 The largest nim-value encountered in calculating that far is 6, which occurs e.g. for the position 0 1 2 36 4 5 7 8 9 A B (and not with any smaller number of discs). Anyone want to check and/or extend this? -- g