Dan Luu

because it's irrelevant to the main point.

I actually had two paragraphs in my draft refuting this reasoning the same way I have a couple paragraphs refuting the walk tweet that you posted then deleted (that deletion is the other reason the responses aren't linear). I may put that back, but I don't know what to do about..

...the general problem of there being an unbounded number of "corrections" to refute. By volume, this post is roughly half defense against nitpicking but this was clearly not enough and yet also way too much.

Yeah sorry about the walk tweet, I think I posted it and then realized that you said somewhere that the actual analysis looked at what happened when different people were at bat.

FYI, I think I would have nitpicked less if you had prefixed the examples with "these are simplified from analysis The Book, analysis there takes into account many more caveats, please look at their model if you want more detail"

I thought at first that you were presenting mostly-original analysis and didn't realize that the actual stats you cite were (it sounds like) taken from a much more detailed model than the one you present

Anyway, it's a totally legitimate move to write a completely caveat-free analysis as an example and then be like "please take your nitpicks to the original source, I promise it addresses them" + maybe this would have saved you time defending from nitpicks?

Oh no, is the appendix new or did I miss it the first time around? If I missed it, that's really embarrassing since it basically addresses the thing about variance :\