Understanding the bias-variance tradeoff (scott.fortmann-roe.com)

Unfortunately, as you can see on the wiki article, accuracy and precision have a different meaning for binary classifiers.

I notice this sometimes too. I think another example is "burnout" and "learned helplessness" [1]. Whenever I see another article on HN trying to understand what burnout is, I feel like we're rediscovering something psychologists discovered in the 60's.

[1] https://en.wikipedia.org/wiki/Learned_helplessness

Funny, these terms are used in my field: https://en.wikipedia.org/wiki/Validity_(statistics) and https://en.wikipedia.org/wiki/Reliability_(psychometrics).

Terms aside, this is a very good article. I've experienced the trade-off between reliability and validity in devising measurements myself, but had not realized it is unavoidable.