2) Both are reasonable statistical methods. The sum of expected is equal to the expected value of the sum. It’s fairly arbitrary what metric to use.

3) However, I suppose I’d prefer the Euclidean norm. (i.e., distance norm, square root of the sum of the squares.)

]]>1. Just because their four-seamers are “average” in terms of motion, velocity, usage, etc., doesn’t mean they can’t throw a great one when they need to. This is just a couple hundred pitches have average out over the last year and a half.

2. With Greinke especially, this isn’t his best pitch, or his strikeout pitch. Greinke (and other guys) throw 5-6 pitches and vary speeds within many of them. Even if Greinke’s four-seamer was truly average, it wouldn’t matter: his other pitches make up for it.

3. While this does consider “movement”, it does not consider placement; i.e., the pitcher is not really being credited for command or control here. That’s another thing that Greinke or someone like Oswalt is excellent at.

You probably knew all these things. But even considering all of the above, it’s still surprising to see some of these guys here. In the case of Lincecum, he still has movement in his four-seamer — at least relative to the guys listed here — but he doesn’t use it as much and his velocity has become pretty pedestrian.

]]>I went back to my spreadsheet and averaged them. It didn’t make any different in the standings, of course, because I was just dividing the totals by five…

Now, if I was really ambitious, I would have decided upon arbitrary weights to assign each stat according to which of them I think most defines a “live-with-able” heater.

]]>I am surprised to see Zack Greinke on this list and I am saddened to see Lincecum.

]]>Looks like you added, not averaged. I highly doubt either way is a statistically rigorous method, but that’s ok because this is notgraphs. Does anybody know the correct way to do this?

Fun article.

]]>