Hacker News new | ask | show | jobs
by otl1248 3618 days ago
This is a weird critique of R-Squared. You (should) learn very early on that comparing R-Squared values of different models is a no-no.
1 comments
achompas 3618 days ago
And you're claiming it's weird because people learn this early on? Maybe that's where we disagree -- it's not as common to learn about the issues with R^2 as you might think.
link
otl1248 3618 days ago
Maybe the fact that this is in the (reasonably brief) Wikipedia entry on R2 is some evidence?

https://en.wikipedia.org/wiki/Coefficient_of_determination#I...

Also, I find that the OP is actually more confusing on this topic than Wikipedia.

Not sure how else I can support that this is a basic fact about this metric. I'm not in the mood to find quotes in intro textbooks, etc.

link