Alex Lanciani:
> But probabilities are useful only to compute averages, not for making
> case-by-case decisions, which should be made on a subjective degree of
> confidence. Of course probabilities are a factor, even an important one, but
> by no means the only one. Wasn't 60% enough for him? 75%? 51%? Where do you
> draw the line?
At 50%+epsilon. As soon as you have even a tiny advantage, it's in your interests to bid as low as possible. (Unless other factors like having APs to burn, a net handicap, or a circumstantial need to win quickly.) At 50% it doesn't matter, and at 50%-epsilon, you ought to start bidding high...
Cheers,
Alex.
Powered by hypermail