3 ms·
The World of Competitive Rock Paper Scissors
- cheepin 12y agoThis is really interesting considering that if you just memorize a sufficiently long string of random Base 3 numbers you can't lose more than 50% of the time, but with any discernible strategy you use to gain an edge there is a counter strategy to remove it.
- pinko 12y agoNot necessarily -- even if you could reliably choose your throws pseudo-randomly, that wouldn't eliminate physical tells (like the elbow swing the article mentioned). The interesting thing about this (to me, as an amateur) is that it's a little more multi-dimensional than I would have assumed.
- baddox 12y agoThis bot can reliably beat me, until I start using a randomly-generated sequence. Presumably, even in the absence of physical signals, if you could compute this algorithm, you could beat a human who isn't also using such an algorithm. http://www.essentially.net/rsp/play.jsp http://www.essentially.net/rsp/play.jsp
- baddox 12y agoThe article makes some convincing cases why an experienced RPS player should be able to beat someone who has never considered the competitive aspects of the game: things like poor randomization and physical clues. But I'm not convinced that the skill ceiling is high enough for there to be an interesting competitive scene. I could be wrong, but it seems like tic tac toe. The slightly experienced player can beat the noob, but no one can reliably beat the slightly experienced player. Surely if you can eliminate your physical signals and memorize a randomly generated sequence, no one should be able to reliably beat you. Perhaps eliminating the physical signals is just extremely difficult?