How accurate are csrankings.org Computer Science Rankings?

Seems to be a pretty complex site, not sure what the criteria of the rankings are. Opinions?

The ranking method is rather simple. It’s based on the number of faculty publications in the listed fields. Whether the counts are accurate or not is hard to determine just from the page (without exploring the links at the bottom).

A ranking like this might be helpful for finding suitable PhD programs. You might find it useful for choosing a college, to supplement other information, if you’re fairly sure you want to major in CS.

^Number of faculty publications isn’t necessarily going to be a meaningful metric for an undergraduate CS major. First of all, honestly the majority of undergrads don’t get involved in research at the undergrad level. If you’re not interested in a career in research it makes more sense to focus on other things - like career services or student life or whatnot.

Even if you are interested in a research career, though, number of publications doesn’t speak to the professors’ ability to mentor, teach, and shepherd an undergrad. A professor who publishes 4 articles a year might actually be better at that than a professor who publishes 8.

FWIW, 8 of the top 10 departments according to csrankings.org are also in the USNWR CS graduate program top 10. The 2 rankings identify exactly the same set of top 5 departments.
csrankings.org is based on publication volume. USNWR is based on a peer opinion poll.
Two different rankings, using two different methods, identify many of the same top departments.

Now, it may be the case that the USNWR “peers” are heavily influenced by publications.
It also may be the case that some highly-published professors aren’t very good teachers or aren’t very attentive to undergraduates. On the other hand, perhaps the same schools that attract excellent CS researchers also tend to attract excellent CS teachers. Perhaps the same professors who are good at writing journal articles also tend to be good at explaining CS concepts in class. I haven’t seen evidence to prove (or disprove) a correlation between research and teaching quality. Maybe some other poster has.

I’d agree that a publication ranking by itself isn’t a very good basis for choosing a college. But absent any other kind of department ranking, you might find it helpful to build an initial list of strong departments for further exploration. Just understand that it may fail to expose some very good undergraduate programs (for example at liberal arts colleges).

@tk21769- Here is an interesting research paper (circa 1996).

One can click on the PDF icon in the upper left to get a copy of the paper.

http://journals.sagepub.com/doi/abs/10.3102/00346543066004507