[ QUOTE ]
[ QUOTE ]
Negreanu is almost certain to have the advantage in rock paper scissors.
[/ QUOTE ]
if there were real reason to believe Negreanu would have an advantage in RPS, then Sklansky would almost certainly just randomize, making it a push by construction. This is pretty much the only one we can be sure of here.
[/ QUOTE ]
Quick story. Friend of mine whipped up a very simple RPS program once. All it did was make markoff charts of your last 1, 2, or 3 throws, depending on the version of the program, and pick randomly based on the liklihood of your next throw extrapolating from historic data.
Example (1 deep): You threw rock last time. You have thrown 100 rocks to date, folloew by 25 S, 60 P, and 15 R. 25% the program would pick R, 60% S, 15%P.
For two deep it had nine look up tables, 3 deep it had 27.
At any rate, the thing was beating people like 50% of the throws, even just one deep. (In stead of 33% W, 33% L, 33% T that would be expected).
Humans are not very good at randomizing.
btw,
here is a RPS thing on line that is supposed to be pretty good, not sure what its learning algorithm is, though.