Hacker News Reader

> There really are good reasons for such a choice. The schools that are so famous for their top-end research are achieving that research success by having those elite professors doing research, and that comes at the expense of teaching. Going to a merely "most selective" school rather than "Ivy Plus" is an decent solution to that problem.

One of the points of going to good research universities is to immerse yourself in exactly that environment: research. Good students probably don't need good teachers to learn. The reason to go to a top research university is to get involved in exactly that research. Especially in computer science and engineering, this is part of the pipeline to the really interesting jobs in industry.

I completely missed this when I was in school, and I regret it.

> The schools that are so famous for their top-end research are achieving that research success by having those elite professors doing research, and that comes at the expense of teaching.

There's no guarantee that the less selective schools are any better at teaching. From what I've seen CMU and Stanford both had smaller class sizes and better teachers than RPI, but that was just from quick touring.

There are some schools with a strong focus on teaching (Harvey Mudd), but that seems to be unusual. I think in most cases if you get into a super selective school you should go if you can.

The Wall Street Journal had an article a while ago where they tried to find schools with great research as well as teaching.

https://www.wsj.com/articles/where-great-research-meets-grea...

Can't get behind the paywall now, but it was a lot in line with your arguments.

Most of the best professors I have had were researchers teaching they favorite subject. Most of the worst were "CS education" types who should have been good teachers.

How are you judging "best professor"?

I had professors at my school that taught their pet research topics. They all sucked at teaching, partly because their research areas were utterly irrelevant for undergrads who were still struggling to learn programming. RDF, anyone?

The best profs were the ones who had been working developers in industry. They found ways to make the classes engaging and fun, vs just lifelessly reading their notes out loud (often in impenetrable accents).

Best professor:

1. Actually know something about, and be interested in, what they are teaching.

2. Being able to explain the subject without making a complete hash of it.

3. Being able to explain the subject without leaving me comotose in my desk-thing.

Extra credit for including tips and tricks that don't get covered in textbooks and are only known to people involved in the field. Plus, background and amusing stories about how the goofy nomenclature and such got the way it is.

On the other hand, when and where I was an undergraduate, if you were "still struggling to learn programming" after the first couple of semesters of programming and data structures (taught by aforementioned "CS education" people), then you were pretty thoroughly hosed. (There were, however, 1-course hour classes in individual programming languages, if you couldn't pick up those yourself.)

I definitely agree on this.

Mathematical logic class taught by logician? Awesome

Automata class taught by expert in complexity theory? Fantastic

Rudin class taught by PDEs expert - amazing

The worst class I had was an introductory algorithms course taught by someone who specialized in teaching.

I toured it in ‘99. I don’t remember the exact stats but the avg HS GPA at the time was around 3.3, it’s SAT’s not very high, and admission % decidedly higher than everywhere else I applied.

At the time I went, I chose RPI above Dartmouth and Johns Hopkins. Times change, I guess.