| |
BGonline.org Forums
How well would a bot do in a human tournament?
Posted By: Chuck Bower In Response To: How well would a bot do in a human tournament? (phil simborg)
Date: Sunday, 19 September 2010, at 12:22 a.m.
Your original post implies that the human has an advantage in being able to exploit weaknesses. This only applies when playing poorer players anyway, as good players who play equal or better players do no exploiting at all. And it also assumes that players who do change their checker and cube decisions to "exploit" are, in fact, actually increasing their equity when they make those shifts...and that subject, in itself, is highly debatable and those strategies can backfire, and when they do, it results in a far greater loss of equity than might have been gained.
Agree. Add to that the following: given that the bots are technically superior to humans (or is someone going to debate that?) then the human exploitation ploys have to make up for this deficiency.
No one has mentioned the fatigue factor. If we're really talking about a standard tournament, this should be the most significant bot advantage of all, although it could already be present in the bots measured technical advantage.
BTW, I don't think Timothy meant anyone would be betting that P(bot win)> 0.5. Presumably he meant P(bot win) > P(individual human[i] win) for all values of i. Or maybe Expectation in dollars greater for the bot than any other individual.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.