sswatson

Thanks for the tip; I was going to take what they said at face value. Incidentally, do you know when people typically hear back about the NSF fellowships?

I thought of the NSF as my best chance as well, but actually MIT told me that the NSF doesn't provide "a significant financial incentive" for them, because it costs them something like $300K to put a PhD through grad school (and I calculate that the NSF only represents something like $90K for them). They did, however, tell me I should notify them if I got one.

I received a Gates scholarship to do Part III at Cambridge for next year, but my wife got into Teach for America and cannot get a deferral. TFA has graciously agreed to place her wherever she needs to be, though, so if I get a deferral to one of the places I applied this year, she can begin her two-year commitment with them while I go off to England. I have gotten into Duke and Penn (only the latter is offering a deferral), and I am waitlisted at MIT. Her TFA deadline is March 23, and I can be almost certain that if I were to get into MIT off the waitlist, it would be after March 23 (their initial estimate was "at least mid-March," but today I found out that they have not been able to make any offers to waitlisted candidates). So my choices are: (1) Go with the sure thing at Penn for 2010, or (2) Decline Penn, ask that my wife be placed in Boston, and roll the dice on getting into MIT. Hopefully that would be this year around April 15, when (presumably) they will get their last wave of decision notifications. Otherwise, I would retake the GRE and apply to as many different programs at Harvard and MIT as possible. (Note: presently my GRE is 880/97%). The MIT admissions committee has been extraordinarily responsive to my inquiries, and for that reason I think I have a non-negligible chance of getting in if enough spots open up. Also, they have already told me that they will give me a deferral if I gain admission, and if I do have to reapply I think that being at Cambridge will strengthen my application quite a lot. My question is whether anyone has an insight that would help me make this decision, and in particular I'd like to know whether my various assumptions are reasonable.