[kafkaesq]

For ex parent's observation is probably spot on if you're comparing MIT/CMU/Stanford/Berkeley to an arbitrarily chosen regional state campus.

I'll readily concur that students from top-tier schools are on average better than students from Podunk schools.

But (based on a wide sampling of data points) I don't see anything like a 1:100 ratio of "better", by any metric.

[throwawayjava]

> But I don't see anything like a 1:100 ratio of "better", by any metric... So again, while it's not like school "tier" doesn't matter - this 1:100 ratio is just silly.

I think you misunderstood my point. Parent's observation may be accurate because of a combination of quality disparity and disparities in raw numbers.

The elite CS programs are relatively large compared to many Podunk States. So it's not just that there's a modest 1:10 or 1:20 or 1:5 or whatever quality ratio, but also that Podunk State graduates 20 or 30 students a year while the elite schools are churning out hundreds. So there's a 1:10 quality ratio but still a 1:100 yield ratio. Or whatever.

Again, inaccurate for the large state flagships, but very believable for the regional comprehensives.

(This isn't idle speculation. My observation is that larger schools tend to get more attention from recruiters, even setting aside quality, and I think this numbers game has something to do with it.)

[kafkaesq]

Or whatever.

Exactly - it's the "whatever" part I'm having trouble working with.

I'm seeing the basic point about population samples you're making. But even so, both logically and numbers-wise, the original commenter's argument was just extremely handwavy. Which is, how to put this nicely... strange, coming from someone dissing an entire class of people (Podunk U graduates) as not just lacking rigor... but 99% likely to be, in their words, "completely incompetent".