On 2013-07-03 16:24, Giacomo Catenazzi wrote:
last year t-shirt ================= email from moray (19.6.2012)Attendee Size Order XS 10 S 30 M 65 L 80 (including ones for sponsors) XL 40 XXL 25
Similar information for DebConf11 is in https://wiki.debconf.org/wiki/DebConf11/T-Shirts#Decisions
When choosing numbers for printing, the main thing is to increase the numbers for the extreme sizes proportionally more than ones in the middle (since the true distribution of human sizes is somewhat wider than the distribution of our t-shirt sizes), and to increase the numbers for unpopular-in-penta sizes proportionally more than for the most popular, in case they turn out to be more popular when people see the real sizes of the shirts. In both cases the *absolute* number added to the penta requests may be lower than for the most popular sizes. Due to the smaller overall numbers for women's shirts, the proportional increase will be higher there, but again with fairly low absolute numbers added.
Last year, although I had done that, we ran out of the smallest sizes quite quickly (= I hadn't increased those enough). That was partly due to demand from them from additional female local volunteers, but also from male attendees who found that the S or M shirts were bigger than they expected and wanted to switch to XS or S.
Of course, this could all be more efficient if we determined in advance what shirts we will use and their true sizes, before people chose their type, but in practice what what each size name means in a particular year is something of a lottery.... Fortunately attendees are happy to take any spare shirts (within reason), and we have also had success with selling surplus ones towards the end of the conference.
-- Moray