I hate it when talks have insufficient numbers of handouts. The probability that a handout will be wasted has to be about 98 to 99 percent before the marginal cost of production exceeds the expected marginal utility of producing it. (Unless it’s a really bad handout, in which case you should make a better one.) So Brian’s rule of thumb for calculating numbers of handouts to bring to talks is make a generous estimate for how many people will be at the talk, and double it.

But the problem is that in practice even making these estimates is a risky business. So I’ll appeal to something like Condorcet’s theorem and take some estimates from what other people think. The session I’ll be doing at the APA is in a pretty good timeslot (Friday 1-4) with some other very good speakers (George Bealer and Paul Churchland) and it’s been heavily advertised (at least here), but there is some strong competition (especially from Bach/Stanley/King). So I’m guessing 100 handouts should cover it easily. But I’m wooried that we could ‘catch fire’, and I’ll be left with audience members without handouts. Anyone out there with better estimating abilities than I want to make a suggestion about how many I should plan for?