I am a professor of mathematics at a small liberal arts college in the US. My department is fortunate enough to have a wonderful group of math majors, many of whom go on to graduate school each year. Consequently, each year I write quite a few letters of recommendation for students applying to (generally US based) PhD programs in mathematics.
I only agree to write letters of recommendation for students if I believe that I will be able to write them a strong letter. In the past I would never compare students (to either current or former students) in my letters, unless it was part of some sort of superlative statement of the form "X is the strongest student in their class / that we have had for several years / etc."
This year I decided to try to be more direct, thinking that doing so might be of value both to the students and to the admissions committees reading my letters. I therefore added a paragraph at the end of my letter explaining that approximately 25% of our students go to graduate school each year, that the top student or two tend to be admitted to schools such as (elite) University X, University Y, University Z, and that the median student in this 25% cohort tends to be admitted to a school whose math department is excellent but perhaps not quite as highly ranked as the previous schools. I then say that the student I am writing the letter for is, in my opinion, most similar to recent students that went on to study at University A and University B.
On the one hand, my hope is that this kind of directness will help students be admitted to departments where they will thrive. On the other hand, there are many potential downsides to such directness. It is possible that making these sort of explicit comparisons make my letter too determinative, that places at the level of Universities A and B will view this comparison negatively, etc.
I have never been on a graduate school admissions committee however, so I am very curious to hear from people who have. (Ideally at a US based math department.) How would you interpret the sort of explicit comparisons I described above? Is doing so a mistake on my part, or does it help admissions committees distinguish between the many different types of students that all have "positive" letters of recommendation?