joyeux anniversaires

Ms. Duncan and yours truly recently celebrated our birthdays (interspersed by the birthday of mon pere). Ms. Duncan celebrated this honorable occasion at the romantic Pigalle, with pork chops. She profited magnificently from the celebration making out with an MIT sweatshirt, an iPhone and best of all, cette blog*.

In addition, Ms. Duncan and three others in her program with very close birthdays feted (at some point I'm going to have to find out how to insert french accents in these posts) the occasions at a class barbecue. Curiously last week's statistics class included an exercise in calculating the probability of the likelihood of two (or more) members of the same class having an identical birthday - its surprisingly high**.

Yours truly celebrated in a more nerdy way by going to a Shakespeare play - Much Ado About Nothing - during which performance i was pulled onstage so that the drunk character could have someone to whom he could ramble on and ultimately implicate himself. Even better, Ms. Duncan prepared 2 peanut butter pies.

In other non-LGO related activities - there was much biking and some introductory sailing on the Charles River. So far, Boston has been an enjoyable place to celebrate birthdays.

*There's actually a double entendre here - the french "blague" meaning joke - a suitable epigram for this entire exercise

**If I recall from my undergraduate statistics class, the correct way to go about doing this is to calculate the probability that no one shares a birthday (which would be (365*364*...*365-(n-2)*365-(n-1))/(365^n) where n=# of students in the class) and then take that probability and subtract it from 1 - thus for a class of 20, the probability of at least 2 students sharing a birthday is something like 0.41 - relatively close to 50/50.

An interesting point was raised by Ms. Duncan's statistics professor - if the probability surprises us, perhaps we need to reassess our assumptions and realign them in a way that allows us to accept the probability without being surprised. A similar curious point was raised in The Metaphysical Club's discussion of the Howland Will case - a case in which a will forgery was revealed by statistical analysis of handwriting - in 1868. Apparently, the revelation that our actions are somewhat governed by statistical laws or that statistical proof might be superior to human testimony was quite shocking to the 19th century American public. Is it any less shocking now?