Josh Willingham hasn’t been very good this year. He’s on pace to put up his worst full-season WAR mark. I’m secretly glad about this because I’ve never really liked Josh Willingham as a player. (Well, I guess it isn’t a secret anymore.)

One of the reasons Willingham has been considered a good player is his selectivity at the plate. He’s certainly a patient hitter: this year, he’s swung at a lower percentage of pitches than any other qualified batter. But has he been selective?

One simple selectivity approximator is O-Swing%. The less pitches a hitter swings at outside the strike zone, the better he probably is at identifying where pitches will end up and which ones not to swing at. And sure enough, Willingham is third-best by this measure for 2013.

But O-Swing% isn’t the best way to measure selectivity. After all, an infinitely unselective batter could take every single pitch he saw and record a perfect 0.00000 O-Swing% — but a 0.00000 batting average as well. Last year, Carson Cistulli noticed this and constructed a leaderboard of hitters with the greatest discrepancies between their O-Swing and Z-Swing rates, in order to highlight batters who swing at lots of hittable pitches but lay off balls. And guess what? Josh Willingham proved to be baseball’s 8th-most selective hitter.

So there’s no doubt, then, that Josh Willingham is selective? Well, it depends how you define “selective”. First, we should notice that Willingham ranks in the bottom 20 of Z-Swing% this year — which means that he made it onto Cistulli’s leaderboard not because he swung at a lot of pitches in the zone, but just on the strength of his amazing ability to lay off pitches out of the strike zone. This is a useful skill — it leads to a high walk rate — but the low Z-Swing% is concerning.

There are two possibilities that I can make out: either Josh Willingham is only swinging at pitches he likes in the zone, but taking “bad” strikes, which would make him a true selective hitter; or he’s just indiscriminately taking too many pitches, kind of like that hypothetical 0.00000 O-Swing% guy.

Which is it?

Here are the pitches swung at by Josh Willingham in 2013, courtesy of Texas Leaguers:

If we just look at pitches in the strike zone, it looks to me like he’s avoiding pitches down and in, as well as pitches down and away. So there’s a pattern, then. That’s a good sign.

Now let’s take a look at his career swing rates, this time via Baseball Prospectus:

Looks similar, right? So Willingham must know what he’s doing. Throughout his career, he’s made sure to lay off not just balls, but also pitches in the bottom-left and bottom-right parts of the strike zone. Those must be the kinds of pitches he’s bad at hitting.

Here’s his BABIP chart:

Well, that changes things, doesn’t it? He seems to be pretty good at hitting those down-and-in pitches he selects against. Willingham appears to have a good eye, but it doesn’t look like he’s choosing his pitches wisely.

But maybe BABIP isn’t the right thing to look at. After all, it leaves home runs and whiffs out of the equation. So here’s the same chart, but using TAv this time (TAv is basically Baseball Prospectus’s version of wOBA; wOBA is better, though):

This seems to justify him swinging at those high, inside pitches, but it still doesn’t explain him laying off pitches down and in, or swinging at pitches up and away.

So I think we can tentatively conclude that Josh Willingham isn’t as good at selecting pitches to swing at as one might think. This analysis is incomplete, of course, because there are a whole lot of factors — pitch type, count, base-out states, etc. — which I haven’t taken into consideration. I also haven’t looked at other hitters’ data — maybe Willingham is actually good at this compared to other major league hitters.

Why not take a look at another hitter? Say, Marco Scutaro. Here’s his Swing% chart:

And TAv:

I would say Scutaro’s selectivity is a little better. He’s good at hitting pitches inside and he knows it, so he mostly swings at pitches inside. Now, he’s inexplicably not very good at hitting pitches right down the middle, but Scutaro probably figures he should swing at those anyway because, well, they’re easy to hit, in theory.

Marco Scutaro is more selective than Josh Willingham. Intuitively, that seems right, as it would explain (to some degree) Scutaro’s league-best SwStr%.

Conclusion

I guess the moral of the story is don’t trust O-Swing% and Z-Swing% as selectivity indicators. And PITCHf/x data is fun to look at.