autumnleaves wrote: ↑Wed Nov 13, 2024 2:20 pm
whatsgoingon wrote: ↑Sat Nov 09, 2024 10:42 pm
autumnleaves wrote: ↑Sat Nov 09, 2024 8:06 pm
Undergrad Institution: Top2 UK; did an integrated Master's degree (graduating next sumer)
Major(s): Mathematics
GPA: First Class honours (85%)
Math GPA: (Only took maths subjects, so 85% here too)
Type of Student: International Male
GRE Revised General Test:
Q: 170 (92%)
V: 170 (99%)
W: 6.0 (99%)
GRE Subject Test in Mathematics: Not Taken
Program Applying: Mainly Stats, a few (applied) Maths, one CS
Research Experience: 2 at my school, no publications but led to one conference presentation in applied probability.
Awards/Honors/Recognitions: Only minor awards that other top3% students get too
Pertinent Activities or Jobs: Tutored at high school level for a few years
Math Courses Taken: Functional Analysis, Stochastic Analysis, PDE theory, Probabilistic Combinatorics, Machine Learning, Bayesian stats, etc. About 30 courses total, and 8 are graduate level. Good grades in all, except PDE theory which was 65% (about a B)
Any Miscellaneous Points that Might Help: 1 of the reccomenders I have is very famous, one of the best in the field I am focusing on.
Any Other Info That Shows Up On Your App and Might Matter:
Applying to Where: (Color use here is welcome)
Current University - Stats (Accepted unofficially by potential supervisor, awaiting funding... would prefer to go to US)
Stanford - Stats
UC Berkeley - Stats
Other Top2 UK School - Stats
MIT - (Applied) Maths
Harvard - Stats
UMichigan - CS
Caltech - Applied Maths
Princeton - ORFE
Cornell - Stats
Yale - Stats
Carnegie Mellon - Stats
UPenn Wharton - Stats
Columbia - Stats
I'd appreciate any comments, apparently these places are very hard to get into. I've browsed the threads from last years, and I see people with publications getting rejected from most of the places I apply to. Are publications expected or a significant advantage? It certainly isn't the case in the UK for pre-PhD students to have publications.
If I were you, I’d apply to more schools of a slightly lower caliber. Coming from Ox/Cam/Warwick/Imperial is impressive, and your grades reflect your ability, but there’s no guarantee of getting into top schools. As an international student, you're grouped into the international pool (unless you have an American passport). While PhD programs technically have tuition (which is more expensive for international students), the department is covering it when you’re hired and provide a stipend for research or teaching. Hence, sponsoring international students for PhDs is more costly for universities due to tuition and visa requirements, making the application process more competitive for international students. This partly explains why, in Pure Math PhD programs, GRE subject test expectations are higher for internationals than for domestic students. So, it’s you competing against strong stats and applied math students from around the world, not the US applicants. Think about that in the grand scheme of things, there's only a finite amount of places these top schools take in and there are a lot of people applying. Statistically and probabilistically, it is not in your favor -- you want to maximize your chances of getting in somewhere so you don't sit around waiting a year to reapply (speaking from experience).
On another note, I’ve heard from several professors that funding is tighter this year (likely due to economic factors), so applicants may need to lower their expectations.
I’m not saying you won’t get into any of the schools you listed—you might even get into all of them. But if your goal is a PhD, I’d suggest adding more schools where you have a stronger chance of standing out. I’m in pure math, so my perspective may not be entirely accurate for statistics, but don’t hesitate to apply to state universities or schools you might not have considered. There are strong faculty at many institutions, including Duke, UNC Chapel Hill, Rutgers, Texas A&M, UT Austin, NYU, Northeastern, Northwestern, Boston College, Boston University, UCSB, UCSD, Penn State, etc. I've also heard UWash Seattle is very good for stats. Obviously I don't know your specific area, so you have to see where there will be faculty in your research interests, but in general broaden your goals. Have a list broken down into Dream/High Reach/Reach/Safety.
I’m sharing this as someone who did undergrad in the UK and then came to the US.
Thanks for your reply and suggestions. My main interest is in stochastic process and applications of these to Bayesian stats. Duke is indeed a good place for my interest and I should add this to my list. For safety schools, I think I'd prefer to apply to UK based schools (Warwick/Bristol/Bath/Durham) for this; personally it's hard for me to justify the difficulties of going abroad in uncertain times otherwise, and I dont have to worry about extra fees (I'm already paying a lot!). If I can't get in a place I want to be in, I also have a data science job lined up so I can work for a year, build up experience and reapply later, so I'm not too worried right now.
Actually, the best looking backgrounds I saw that faced rejections in the past few years were by domestic students. Good GPA from decent schools and good GRE if they took it, actual publications in difficult abstract fields like algebraic geometry, but they were still getting rejected a lot from the top places. i.e. in the 2024 thread you can check users 'calcifer', 'KhanComplex', ' HurewiczFan2024', etc (there's way more if you go back). I wonder why that is - is pure math just insanely competetive in the US? I've known people in top5 UK universities who just do a good degree, maybe a summer or two of research experience (no publications), and get into Oxford/Cambridge PhDs for stats. But the people in these threads getting rejected from Princeton, Berkeley, etc., for maths, I don't know how much better they could really make their profiles. Maybe strong connections with the faculty in those universities in addition to publications and good grades is required to get an admit.
It’s reasonable to prioritize safety schools nearby.
The top institutions in the US are undeniably more competitive than those in other parts of the world. This is largely due to higher funding, a much larger and often better-prepared pool of applicants, and the global draw of these institutions. The best students from around the world compete to gain admission. In contrast, gaining admission in the UK or EU, as you said, generally requires being a strong student, finding a supervisor willing to take you, and securing funding—a process that tends to be more straightforward.
Luck also plays a significant role. Consider this: every year, top math undergraduates from institutions like MIT, Princeton, Harvard, Stanford, and Yale apply to peer institutions. These students have often benefited from exceptional research supervision and mentorship, as well as connections with faculty who are either prominent mathematicians, collaborators with professors at other institutions, or former students or advisors of those professors. As a result, these schools often "trade" their best students, admitting some from less prestigious institutions and a number of international students (some schools take more internationals than average, some also give opportunities to students who don't have a perfect record like Berkeley). If you examine graduate student rosters at these schools, you’ll find a noticeable skew towards students from MIT/Harvard/etc.
This isn’t to say that these students are inherently better mathematicians than those from outside this elite group. However, familiarity with recommenders can have a significant influence on admissions decisions. For instance, a professor personally vouching for an applicant—whether through an email or a phone call—can make a considerable difference. A student might have participated in an REU program at another university, built a strong relationship with a faculty member there, and later applied to that institution for a PhD while notifying the faculty member of their application. This personal connection is invaluable at such a competitive level.
Many applicants to these schools are exceptionally well-qualified. They typically have perfect grades, robust research experience (even if unpublished), stellar GRE scores, and outstanding letters of recommendation. Beyond these qualifications, there is often little to distinguish candidates, apart from their personal statements. Ultimately, admission can hinge on factors as unpredictable as luck or external considerations, such as the school’s recent admissions trends. For example, if a program admitted an abundance of geometry/topology students the previous year, it might prioritize applicants in other areas during the current cycle.
This is simply how the system works in a capitalist society like the US. Consider Germany, for example: many of its universities have extremely strong researchers and faculty who are experts in their fields. Yet, you may not recognize the names of these institutions because universities in the EU operate under a more "public" system, where pursuing a career in mathematics generally ensures a good standard of living. In the US, however, the prestige and concentration of faculty at certain schools create a “dream” tier of institutions. This dynamic is further amplified by the association of these schools with wealth, exclusivity, and access to resources.
Ironically, many universities with lesser prestige also have outstanding faculty—often graduates of these top-tier institutions. In the US, there is a general consensus that if you complete a PhD at a certain level, you will, on average, become faculty at a university one tier below. This hierarchy, while imperfect, reflects the broader structure of academic opportunities in the country.
That's why it's often suggested that no matter how good of an applicant you are and whatever university you are coming from, apply broadly and at all levels -- nothing is guaranteed. Don't forget that the applicants you see posting on this website are small proportion of the whole pool, MOST people don't post here but just browse so you're not seeing the full picture.
. Passed my school's algebra qual. 
The number of profiles posted is on steady decline on this forum since 2022, yet the number of application is a.e. increasing (according to admission stats posted by each school, if any). The viewing was of course at all times high for Fall 2023 cycle, but the view and post number dropped so steep.