2025 PhD in Statistics Chances

[AManWithAPan]

Hi. I'll be applying to PhD programs for Fall 2025 and I wanted to check my chances on these reach schools. I already have two acceptances from decent universities in Europe for 2025 so I'm not bothering with "safeties" anymore, but want to try my hand at these programs as I want to focus on optimal transport, and these schools have faculty that directly work on it, not just use it as a tool in their research. For reference, I am an international student from South America studying in the US.

Undergrad - Applied Mathematics from a well known school in Los Angeles

GPA: 3.6/4.0 (I suffered an accident that destroyed my senior year GPA, which I will briefly mention in my SOP)

Honors: Dean's List, International Dean's Scholarship, Prestigious Grant from outside the university, 2021 Math Department's Award

Masters - Applied Mathematics from the same school

GPA: 3.9/4.0 in Master's, and 4.0/4.0 in 3 other grad math courses outside my masters at another university.

Grad Courses: Real Analysis (B+), Probability II (A), Mathematical Statistics II (A), Numerical Methods (A), Statistical Learning Theory (A-), Optimization (A), Data Science (A), Machine Learning at the Engineering Department (A), ODEs (A-), Statistical Consulting (A) and I have also taken 3 other courses at another T25 university with A's in all: Real Analysis II, Linear Models, Bayesian Statistics.

Teaching Assistant for 2 years for Calculus 1 and 2.

GRE: 169 Quant

Research

- One first author research conference presentation with a PhD student.

- Published master thesis supervised by a well known professor in probability.

- 2 single author submitted papers on the geometry and properties of the Sinkhorn algorithm for optimal transport.

- 1 single author paper on using OT to analyze training dynamics of neural networks still in the works, but will hopefully submit before October.

- One summer project on optimal transport supervised by a Professor at a T10 school which will hopefully end up in a publication.

Letters of Rec: I think 2 of them are strong, as they are from Professors that have both taught me and supervised my research. The third is weaker, from a Professor that saw me go from doing not too well in one statistics class to doing very well in the next one.

Schools I'm Thinking About: Cornell, EPFL, ETH Zurich, Kings College London, Yale, UCLA, UMichigan, Columbia, NYU, Penn, Harvard, Warwick, MIT (in no particular order, I would be more than happy to go to any of them). Cornell, EPFL, ETH Zurich and Warwick all have more than one faculty member focusing on optimal transport as far as I can tell. 

Are any of these schools within reach? Or should I just save my energy and go to one of the two universities that have already accepted me?

Edited June 10 by AManWithAPan

[John Watson]

Your academic background impresses me. You have high GPAs from both your undergrad and master's, and you have completed a great deal of coursework in Statistics and Mathematics. I can see that you're indeed a straight-A student. Also, your research experience is abnormally strong, which is not every-day in the field of Maths and Stats. Very good indeed!

I think Warwick, KCL, and perhaps even ETH will beg you for going to them. I don't know what your undergrad and master's school is, but, if it is UCLA, then your chances are so big that you deserve to go to the best school anywhere. If you go to England, I'm not sure about Oxbridge, but I think you can at least go to Imperial. You can also go to ETH. For the United States, I'm not sure about the Ivy League, I think Harvard and MIT are still "reach," but anything else, like Cornell, Columbia, NYU, and UMich, is quite match and safe.

You may just focus on the schools which have professors working on your area of interest, and you apply to them. At your level, it is more important to find a supervisor who suits you, rather than overthinking about schools as a whole. Don't worry, you will end up going to a decent place. Good luck!

[AManWithAPan]

Thank you for that response, but isn't it overly optimistic? For example, I don't think ETH Zurich will beg anyone. I am particularly worried about my low undergrad GPA.

Edit: It's not UCLA.

Edited June 17 by AManWithAPan

[John Watson]

If it's not UCLA, then I may think my estimation is indeed too optimistic. I personally know someone who is doing a PhD at ETH Z and their academic background is not as good as yours. They are in a different discipline, but still it's within Natural Science.

I totally understand that you're worried about your undergrad GPA, but it's partially due to the fact that you encountered an accident. You can explain it in your personal statement for it to be included into the consideration. If your major GPA is still high and it's just some random general requirement courses that make your overall GPA low, then it's not a problem. You have listed a lot of your major courses within your discipline, and you have a good grade in almost every one of them. Rest assured that your future school and supervisor will be more interested in how well you can do in your concentration, rather than focusing on some random, irrelevant courses that you happened to not do well in.

One last word is that, if your school is USC, then I still insist that you can have a good chance to go to a decent school. However, if it's not USC, then disregard what I have said in my previous post.

[John Watson]

I'm sorry, USC or Pomona I mean, Pomona is technically not in Los Angeles, but still a good school

[Sstat]

I think top10 would be tough due to relatively low undergrad gpa and B in real analysis. However, attempting a few schools seems reasonable bet. I think you could get accepted at t20 large state schools

Edited June 17 by Sstat

[bayessays]

If you've legitimately contributed to all the research you've mentioned and have letters from faculty at top schools saying this, I think your results could be quite good (including the top 10).  However, I strongly disagree that any of the schools you've mentioned are close to being "safe" (except perhaps UCLA).  But you seem to be aware of that fact, and I think you have a profile that makes these places worth applying to.

I wouldn't let a B or a slightly low undergrad GPA worry you too much. Your math background is strong and nobody is going to doubt that you can handle the work

[AManWithAPan]

I completely agree that no school is safe, and thank you for the feedback! @bayessays do you think retaking analysis at another school and getting an A would strengthen my application?

[bayessays]

1 hour ago, AManWithAPan said:

I completely agree that no school is safe, and thank you for the feedback! @bayessays do you think retaking analysis at another school and getting an A would strengthen my application?

Absolutely not - first off, a B+ is not that bad, and you took another analysis class and got an A.  Also, these are graduate real analysis classes which are well beyond what you'd need for a stats PhD.  And finally, you have a lot of math and a great GRE and lots of research, so I don't think anyone will have doubts over your ability to handle the math.  It's all going to come down to your research, letters, and random chance.  Your profile is great.

Your research field of interest is quite niche - I will say that if a school only has 1 faculty member working on that topic, it might be a little off-putting to say that you only want to research one very unique thing, which most statisticians have never heard of (what happens if that faculty member gets a job at another school?) Just something to think about when you write your statement of purpose. And when you are evaluating your choices at the end of the day, I'd try to figure out where it seems like you'll definitely have the option to work with the faculty you want to.