Questions about Harvard Math Grad School Admissions

[ForMyThunder]

I'm aspiring to Math Grad School at Harvard, and I had a couple of questions.

Is it absolutely necessary to have a 4.0 overall GPA to get in or would they allow maybe one or two B+'s?

Which school has a better math grad school, UC Berkeley or Harvard?

 

[hadsed]

The trend seems to be 3.8+ or so. From what I've seen, it looks like it's one of those things that is sort of expected for people applying to the program. Things like a perfect math GRE score and (if you're doing physics) probably in the 90th percentile or higher on the PGRE, 3.8+ GPA and of course some kind of research experience. It seems like they put a big emphasis on how well your recommendations say you are and who they're written by. Research experience also holds about as much weight, if not more. If you've published a paper in a good math journal, that would probably guarantee you admission to some really good schools (it's hard to say that anything would make you a shoo-in to Harvard or any school specifically since they have tons of really good applicants every year).

It's sort of like if you were a high school student asking if you could get into Caltech with a 3.5 GPA. I mean, I'm sure it's possible, albeit unlikely, but that student must have done something extraordinary to get in (like winning math competitions at a high level, international science fair, research, patents, etc.). Still, Caltech is a place where the extraordinary students go, and though they don't care about how well you can analyze literature, they expect you to be smart or hard-working enough to blow through it with an A regardless. I think schools like Harvard are like that too, with GRE replacing SAT/ACT scores + GPA. Those things won't get you in by themselves, but if you don't have them then it will hurt your application.

That's what I've read about and information gathered from people who are involved or have been through the process. Plus, it doesn't matter if Harvard is 'better' than Berkeley at math because those rankings are based on peer opinions. Even if they weren't, you're still not applying to the whole school, you'd be applying to work with a specific group.

 

[G01]

It's not all based on GPA. There are many people with 4.0's who are rejected from Harvard's grad programs every day. Same thing with Berkley.

Having a stellar GPA and GREs is important, but won't get you into Harvard by themselves. Have you done undergraduate research of any kind? Any REU's? What sets you apart from everyone else applying?

Honestly, the answer is no, a 4.0 is not necessary, but having one doesn't assure you'll get in either. Even for the best and brightest, getting into Harvard, Berkley, Caltech, etc. is a bit of a crap shoot.Good luck to you, but make sure you have some good safety schools on your list.

 

[deluks917]

Do you have any idea what you are planning to study? Harvard is not a huge department and they aren't strong at everything. If you are thinking of doing Algebraic geometry you should consider a different list of schools then if you want to do functional analysis.

edit: sorry I thought you said applying not aspiring. Either way its worth noting that the "best school" varies dramatically by subject area.

 

[Anonymous217]

I'm sure at least a couple graduate courses with As in them are expected as well considering that almost all the top applicants have taken quite a few.

 

[zpconn]

From what I know, what is expected classes- and grades-wise is (1) you have a lot of advanced coursework, probably the equivalent of at least one or two years of graduate work at a much less highly ranked program; and (2) your grades in these classes are all As and A-s. I doubt they're going to care about an A versus an A-. Anything below an A- will raise flags.

If that's all you have, though, you'll have a really, really good chance of getting rejected. What will really get you in will be absolutely stellar recommendation letters. You want your letters to say you're the best student your school has had in a decade. You want to be best friends with the teacher who writes your letter. Research publications help, but again even places like Harvard will understand that you could have all the potential in the world but just weren't lucky enough to get something publishable as an undergraduate.

So I'd really say it comes down to recommendation letters, provided you have very strong coursework, grades, and test scores to back them up. Research is a bonus. Math competitions are also a bonus (and it must be kept in mind that you will be competing with Putnam Fellows).

 

[deluks917]

The worst difficulty in publishing is that even if you finished your article by the time you start junior year it could still not be published before you apply to grad school. The review process can take a lot of time.

 

[mathwonk]

have you read this?

http://www.math.harvard.edu/pamphlets/gradsch.htmlyou might also try emailing a professor or two at harvard, such as joseph harris, with your questions. anonymous answers here are not worth much.

 

[QuantumP7]

I'm hoping for a math PhD at Harvard, also. What I'm going to do, is try to get a high score on both the Putnam exam and the math subject GRE. I have research experience (in physics, but not math). What sets me apart from others is that I'm severely hearing impaired. I have dedicated hours each day to studying math, because I adore it so much. I study best on my own.

I've been to Harvard's Center for Astrophysics to work on a presentation and tour the observatory. I love the place, so I definitely want to go there.

Just wanted you to know that you aren't alone.

 

[twofish-quant]

Let me break the bad news to you. You probably won't get into Harvard Graduate school. Lots of applicants. Very few places. Even if you had a "perfect" admissions profile, there is a good chance that you won't get in.

It's a good idea to apply to 6 to 8 grad schools. One or two of them should be "shoot the moon" schools. You probably won't get into Harvard, but it doesn't hurt to apply, and you might get lucky.

However, most of your time should be spent trying to figure out what of the remaining schools you should be applying to.

 

[Phyisab****]

I'm pretty sure more people have posted on this forum in the last year about a math PhD at Harvard than are accepted each year. The odds are not in your favor.

 

[ForMyThunder]

Are there any statistics? What's the acceptance rate?

 

[G01]

Are there any statistics? What's the acceptance rate?

A ball park estimate of the acceptance rate is easy to work out:

A tier 1 research university like Harvard will generally get at least 500 applicants a year to anyone of it's programs.

From PhDs.org, Harvard Math's incoming class last year consisted of 12 people. Assume they accepted twice as many people as the actually wanted to have in their first year class. (Admissions people are very good at playing this game.) So that puts us at 24 students accepted to the program.

So we have 500 people who think they have what it takes to study at Harvard (and a lot of them probably do). 24 people are actually accepted.

That's a 4.8% acceptance rate!

The Moral

Assuming you have the numbers that make you a potential Harvard applicant, your chance of getting in is still only around 5%!

Even if you assume the school accepts 3X as many students as they actually want, that only pushed your chances up to: 7.2%

The point is that a "Harvard or Bust" plan is not a good idea. Many, many qualified people will still be disappointed comes time for acceptance letters to be sent out.

The Bright Side

There are many, MANY more great places in the world to study math. They may not have the big names, but they are great schools with great programs. You can go to many schools and find great faculty members, learn a lot of math, physics, etc. and experience a lot of personal growth. And that is what it is about in the end, love of the discipline and personal growth. Trust me, Harvard and Berkley do not have a monopoly on that.

 

[QuantumP7]

There are always naysayers. The fact is, people DO get accepted into Harvard's math PhD program. The chances are not negative, and are not 0. If it's meant to be, it will be.

 

[deRham]

G01 - I wanted to venture and point out that while your message is very good, admissions folk from top tier school also mentions how they receive many applications that are severely lacking for the level a top tier grad program might expect. So the numbers of five and seven are too generic in some ways, which is why time and again, people will keep asking how they can develop a good shot. So one could be more optimistic than your post might first suggest.


However, one can put the opposite spin...twelve is tiny, so really for most applicants, there is no hope at all. Not everyone has an equal shot.

 

[G01]

G01 - I wanted to venture and point out that while your message is very good, admissions folk from top tier school also mentions how they receive many applications that are severely lacking for the level a top tier grad program might expect. So the numbers of five and seven are too generic in some ways, which is why time and again, people will keep asking how they can develop a good shot. So one could be more optimistic than your post might first suggest.However, one can put the opposite spin...twelve is tiny, so really for most applicants, there is no hope at all. Not everyone has an equal shot.

Well, I guess we could assume that half of the applicants (~250) are woefully inadaquate and stand no chance, which is an overestimate I think, especially for Harvard.

Then, as an upper limit on ones chances we have about a 10% acceptance rate.

Either way, the point is that even if you are legitimately qualified, you chances of acceptance are arguably not really much higher than 10-15%. And they are never going to be in your favor for a school like Harvard or Berkley.

My argument is that focusing your hopes on getting into one specific program is not a good idea. One should develop a list with a range of schools at which they would be happy. Have some reach schools, but also some respectable schools at which the chances of acceptance are a bit better.

There are always naysayers. The fact is, people DO get accepted into Harvard's math PhD program. The chances are not negative, and are not 0. If it's meant to be, it will be.

I am not saying he shouldn't apply. By all means he should. I'm only trying to point out that one should not have one school where they must attend, and then a list of safeties where you'll attend if you have too. Your safeties should be programs at which you will be happy and content. This is why one should spend a non trivial amount of time thinking of other programs that are good fits, not just one program.

Perhaps I misread the OP and he is not following this line of thinking. All I am saying is that aspiring to one program in particular, and a top tier program at that, is not necessarily a good idea. Perhaps the OP will get in and good for him if he does.

 

[twofish-quant]

There are always naysayers. The fact is, people DO get accepted into Harvard's math PhD program. The chances are not negative, and are not 0. If it's meant to be, it will be.

But you have to deal with the fact that it's almost certainly not meant to be.

One thing that I find fascinating is how people that are supposed to be the worlds more numerate and intelligent people suddenly start thinking like lottery ticket buyers when it comes to important personal life decisions.

 

[twofish-quant]

Either way, the point is that even if you are legitimately qualified, you chances of acceptance are arguably not really much higher than 10-15%. And they are never going to be in your favor for a school like Harvard or Berkley.

Whereas there are a lot of decent math programs where the acceptance rate is more like 30%, which means that if you have a good application, the odds really are in your favor. If you apply to eight schools, then the odds that you will get in somewhere are quite good.

Your safeties should be programs at which you will be happy and content. This is why one should spend a non trivial amount of time thinking of other programs that are good fits, not just one program.

And there's also the issue of time allocation. The stuff that you can do that gets you into Harvard is more or less the same standard graduate school stuff, so there is no point in thinking too much about that. Get your grades up, get good recommendations, get good research, yadda, yadda, yadda.

Also, you really have to ask yourself why you really want to get into Harvard. My guess based on my own experience is that you've spend your life getting lots of attention and praise for being the smartest kid in the school, and you figure that the rest of your life is going to be like that.

What's going to be a shock is what happens when the spotlights go off, and you suddenly find yourself in a situation in which you are not the smartest person in the room, but the dumbest. It's going to happen to you eventually, and a lot of whether you end up being productive or whether you have a nervous breakdown and leave science and math completely is what happens once you find that the "get awards for being the smartest person" strategy doesn't work any more.

 

[rasmhop]

But you have to deal with the fact that it's almost certainly not meant to be.

One thing that I find fascinating is how people that are supposed to be the worlds more numerate and intelligent people suddenly start thinking like lottery ticket buyers when it comes to important personal life decisions.

If you win the lottery the first 10 times you buy a ticket, you will be surprised when you lose on the 11th ticket.

"Smart" people with a shot at going to grad school at Harvard have probably gotten into a lot of "top 5%s" throughout their academic career (top 5% GPA, top 5% undergrad, top 5% in science contest, etc.). It is not rational but even smart people are human and fall victim to the usual logical/emotional fallacies.

I do not really have any useful information besides what others have mentioned. If you've got good research, recommendations, grades, etc., then you likely have a shot, but almost no matter who you are that shot does not exceed something like 3-10%* so do not put all your eggs in one basket. There are plenty of good grad schools, and while Harvard may be one of the most recognizable brand names I'm sure you can find at least 10 places that are just as good for you. By all means apply, but do not count on getting in and do not discount the idea that some other school may actually be better for you. Also if you are rejected do not take it personally.


* Percentages are made up based on my personal perception of Harvard admission and G01's comments.

 

[Stan Marsh]

Just for your information:
I know several physics students in my univ go to harvard.
They are usually: top 1 in GPA, with 3 or more publications. They are so perfect as I see.
Of course, they are international students and thus may have a smaller chance compared to us students.

I thinkd harvard is the most selective school. Last year MIT physics admit around 7 students in my school, but harvard admit only one.

 

[QuantumP7]

But you have to deal with the fact that it's almost certainly not meant to be.

One thing that I find fascinating is how people that are supposed to be the worlds more numerate and intelligent people suddenly start thinking like lottery ticket buyers when it comes to important personal life decisions.

Well, thankfully, it's not really up to random people on a message board what is meant to be and what isn't.

As far as the lottery ticket thing? Well, if you don't ever buy a ticket, you DEFINITELY won't win. And if you don't apply to Harvard, you DEFINITELY won't get in. Sometimes, you have to take lottery-type chances to get what you want. *shrugs*

 

[mathwonk]

3 years after writing my PhD I spent a year at Harvard as an NSF postdoc, one of 10 awarded in the US that year. Even then I was not as advanced as the top grad students at Harvard. I was more or less in the middle of the group of grad students there in terms of knowledge. Indeed while I was there I wrote a joint paper with one of those grad students, which was accepted to Inventiones.

In some ways it was an intimidating place, but nonetheless it was a wonderful experience. Still I am very happy I was not there as a grad student, but waited until I had more knowledge.

If your advisor thinks you belong at Harvard, fine. But I can think of much more pleasant experiences than being at Harvard and realizing you do not belong there. In general one should always go where one will thrive, not where one hopes to derive the most prestige. If you go to the place most suited to your current development, you will improve and then maybe go to Harvard, or some other top place, after showing what you can do.

If Harvard is a reasonable choice for you, quantumP7 reminds me that when I considered applying for the NSF postdoc I also assumed I would not get it, and someone said exactly what she said: "if you do not apply, then you definitely will not get it." Still, you might go there and visit with the current grad students to get a feel for the environment there. And read the information in my previous post. But you cannot possibly make an informed decision just from reading advice here.

 

[QuantumP7]

Thanks, mathwonk!

Yeah, there are other schools I definitely would be happy to attend, also. However, Harvard remains my first choice. But we'll all see how everything goes.

P.S.

If Harvard is a reasonable choice for you, quantumP7 reminds me that when I considered applying for the NSF postdoc I also assumed I would not get it, and someone said exactly what he said: "if you do not apply, then you definitely will not get it."

I'm a girl.

 

[mathwonk]

thanks...fixed.

 

[mathwonk]

One more remark, I did not get the postdoc from just a paper application, it definitely helped that I visited Harvard and met people there so they were able to do a face to face evaluation. It is very hard to evaluate candidates accurately just from test scores, and the best way is to talk to them. This may or may not be feasible, but it can help them size you up as well as help you size them up as potential teachers and advisors. Very helpful people there whom I have met include Joseph Harris, and Barry Mazur, and Curt McMullen. Talking to one or more of them cannot hurt.

 

[G01]

Well, thankfully, it's not really up to random people on a message board what is meant to be and what isn't.

As far as the lottery ticket thing? Well, if you don't ever buy a ticket, you DEFINITELY won't win. And if you don't apply to Harvard, you DEFINITELY won't get in. Sometimes, you have to take lottery-type chances to get what you want. *shrugs*

Once again, no one is saying the OP should not apply to Harvard, only that he should have a good backup plan thought out, should he not get in. I applied to Harvard myself when I was applying for grad school, half on a lark! (I didn't get in, but ended up a great program where I'm happy.) I'm still happy I applied to Harvard, though. Otherwise, I would never have known.

As for me pointing out the low percentages: The OP asked for Harvard admission percentages, so I gave him the best estimate I could. It's not my fault the numbers are low!

My advice can be summed up in one line as follows:

Apply to Harvard, but have a few reasonable backup programs as well. Good luck!

What is wrong with this advice? It is sane, realistic without being cynical, and comes from experience.

 

[deRham]

I think what I meant is when there are twelve people projected, those winning the Putnam (or being of that level) and having a ridiculously good command of advanced mathematics at a young age and lots of good ideas are a nontrivial part of the competition.

So I agree that it is bad to fixate on a single school, but my reason is that chances can be much greater than ten percent for some and much lower for others...and by default for nearly everyone, it is probably much lower. If one is not a freak of nature but extremely strong, it can make sense to try to boost the chances, which are still low.

It is always good to have a range of programs that one would like attending.

 

[QuantumP7]

Once again, no one is saying the OP should not apply to Harvard, only that he should have a good backup plan thought out, should he not get in. I applied to Harvard myself when I was applying for grad school, half on a lark! (I didn't get in, but ended up a great program where I'm happy.) I'm still happy I applied to Harvard, though. Otherwise, I would never have known.

As for me pointing out the low percentages: The OP asked for Harvard admission percentages, so I gave him the best estimate I could. It's not my fault the numbers are low!

My advice can be summed up in one line as follows:

Apply to Harvard, but have a few reasonable backup programs as well. Good luck!

What is wrong with this advice? It is sane, realistic without being cynical, and comes from experience.

There's absolutely nothing wrong with that advice. That post was directed to two-fish quant. I definitely have other programs that I'd love to attend (as I said in my last post).

 

[G01]

There's absolutely nothing wrong with that advice. That post was directed to two-fish quant. I definitely have other programs that I'd love to attend (as I said in my last post).

Oh sorry. I guess I just wanted attention!