I'd like to raise a question. If you put the shift on someone, you don't want to pitch him so as to make it easier for him to hit outside the ####, do you? Doesn't that mean that it will make it easier for the hitter to guess pitch location, and selection even? And if you're pitching a pull power hitter to make him pull, aren't you, at least to some extent anyway, going to be pitching in his wheelhouse?

The real problems with playing The Shift are

1) That it leaves sections of the field undefended

You are always going to leave sections of the field undefended; it's physically impossible for seven players to cover the entire field. When you shift, you just leave different sections of the field undefended than you usually do, and you do it because (a) you consider it less likely that the hitter will be able to hit the ball into those sections of the field than a typical hitter will, and (b) even if the hitter does hit the ball into those sections of the field, you come out ahead because the hitter is less likely to hit the ball hard into those sections of the field, and thus is less likely to hit for extra bases (in principle).

2) That players are playing out of position, which very often requires them to do things that they don’t exactly know how to do.

While that's true, the flip side is that The Shift can push a hitter to do things differently than he normally does, and "get himself out", so to speak.

There are other things that would be interesting to study - for example, the impact of playing your infield at double-play depth with a runner on first vs playing the infield at normal depth, the impact of guarding the lines vs having your 3B and 1B positioned normally - but they are hard to study because there isn't a lot of variability between the approaches that teams take.

-- MWE

I don't know anything about Dewan's study so I'm not saying it's set up correctly. But I'm not sure I'm grokking James' point either. Maybe there are in fact no situations when it makes sense to play a shift but, if there is, clearly it's a "threshold" kinda thing. If Dewan is saying "it only makes sense to play a shift against guys who pull the ball X% of the time" and Ortiz doesn't pull the ball X% of the time, then it's perfectly OK for Dewan to point that out and adjust the study in some way to account for that (granted, excluding him sounds too extreme on the surface).

I only just read the article last night.

Emeigh's comment:
they are hard to study because there isn't a lot of variability between the approaches that teams take

plus some comments from the BJOnline article conceal a fact that should make us sit up and take notice. In cricket, the fielders get moved around a lot. In baseball, not so much. That has implications that I don't think defensive sabermetrics has quite accommodated. For one thing, speed might be a much more important variable than location for fielding batted balls.

My completely unscientific opinion is that the shift was underused in the past, and is overused now.

The idea that professional ball players "don't know what to do" when they're "out of position" makes me giggle. I don't believe that's the case. They know what to do, and they know how to do it. Field the ball, throw the ball.

#6 - It's more like there are different responsibilities for each player since they are not playing their usual position (who covers 2nd, who takes the cutoff, etc.). But yeah, it's probably overblown.

Is the shift used when there are baserunners? I don't recall.

Well, there is potential one at the plate.

Why, I oughta...