The League Catches On To Tom Wilhelmsen's Curveball, Maybe

My favorite thing about Tom Wilhelmsen is Tom Wilhelmsen's story. Everyone loves Tom Wilhelmsen's story, and everyone should love Tom Wilhelmsen's story, because it's not all that often that a twentysomething bartender turns into a major league closer. Tom Wilhelmsen's story is so incredible that I'm contractually obligated to mention it every time I write about him. Or I would be, if I had a contract.

My second favorite thing about Tom Wilhelmsen is Tom Wilhelmsen's dance moves. On a scale of one to incredible, Tom Wilhelmsen's dance moves are a twelve. Behold:

Hypnotic.

My third favorite thing about Tom Wilhelmsen, unlike the first two, actually has to do with his pitching. I love Tom Wilhelmsen's curveball. If I could marry a pitch - no offense to King Felix's changeup, or Carter Capps' slider - it would be that curveball. The Bartender's curve is entertainment at its finest; it has that unusual combination of remarkable effectiveness and gut-busting hilarity that makes for unforgettable baseball cinema. I will always remember when Tom Wilhelmsen did this to Alexei Ramirez, and I will always remember when he did this to Asdrubal Cabrera. Tom Wilhelmsen's curve is maybe the best curve in the sport and definitely the most potent weapon in the Mariners' bullpen.

Which is why it makes me a little bit sad to write this post.

See, though Tom Wilhelmsen's results this year have been very good - he's got a 0.45 ERA and a 2.43 FIP - his peripherals are some cause for concern. The Bartender's strikeout rate has taken a serious hit so far in 2013, and the ERA is supported by an unsustainably high strand rate and an unsustainably low BABIP. Several of my fellow writers have commented on the seeming inevitability of Wilhelmsen's regression and expressed concern about his dropping K rate.

Still, using only K% and BB% to analyze a pitcher's performance is pretty darn lazy, so I decided to go investigating. My investigation started, as most baseball investigations do, on Wilhelmsen's Fangraphs Player Page. And it was there that I discovered that Wilhelmsen's swinging strike rate hasn't really dropped all that much. It doesn't explain the big drop in K%, for sure. He's also started throwing more of his changeups and two-seamers, but those are still good pitches (the changeup in particular generates a bunch of ground balls), so that's not really a good explanation either.

I was all set, then, to write an article about how you shouldn't be worried about Tom Wilhelmsen's strikeout rate. His peripheral peripherals, the core skill stats, say he's fine! Cheer up! Except, unfortunately, curiosity got the better of me and dragged me over to Wilhelmsen's Brooks Baseball player card. And so it was that I delved into the fourth level of the layer cake that is Tom Wilhelmsen's 2013 season, and discovered that it is not very tasty.

Wilhelmsen's game, as you probably know if you've watched him, is a pretty simple one. He has a plus four-seam fastball that sits at 96 MPH, touching 99, and that devastating wipeout curve. He uses the four-seamer to filch strikes early in the count (over 50% of Wilhelmsen's fastballs in 0-strike and 1-strike counts go for strikes not in play). Once he's ahead in the count, he either blows hitters away with a high fastball or humiliates them with a low curve, sometimes mixing in a two-seamer and a change just to keep hitters on their toes. His change/curve repertoire means he gets a lot of ground balls while allowing fairly few home runs, and the four-seamer and curve combine to get him all the strikeouts he needs.

Or, I suppose I should say, they have in the past.

This is a chart I assembled using data from Brooks Baseball's player card for Wilhelmsen. I'd like to draw your attention to the Called Strike column. Notice how Wilhelmsen's percentage of called strikes on two-strike curveballs has gone through the floor this year? That's bad.

And it's not Jesus Montero's fault, before you ask. Montero's pitch framing is dreadful, but Texas Leaguers thinks that Wilhelmsen has only lost four strikes on curveballs this year, with only one of those coming in a two-strike count. The problem isn't that Wilhelmsen isn't getting calls that he used to. It's not his curveball command, either - Wilhelmsen has actually been throwing more two-strike curveballs than he did last year. The problem is that the hitters facing him have been much better at swinging at curves inside the strike zone and laying off the ones outside.

So far this season, Wilhelmsen has exactly one called strike three on a curveball. It came on this pitch.

That pitch down in the bottom left hand corner is in the rulebook (rectangular) strike zone, but not the called (circular) one. It's the only curveball Wilhelmsen has thrown within the rulebook zone that hasn't been swung at.

This is kind of a bummer, because my third favorite thing about Wilhelmsen was his ridiculously good curveball, and so far this season it's been worse. At any rate, it hasn't been freezing hitters like it used to. The easy narrative to build from this is that the league has caught on to Wilhelmsen's curve, that after a few times seeing him most hitters know how much it breaks and will swing on anything near the strike zone in order to avoid the supreme embarrassment that is a called strike three on a meatball. Perhaps that's the correct narrative - I can't say for sure - but it behooves one to remember that we're dealing with a sample of about twenty-five pitches. Maybe Wilhelmsen's curve is getting exposed, and maybe he's just run into a bunch of hackers.

The thing is, it doesn't really matter. Even if the league has figured out when to swing at Wilhelmsen's curve, it sure hasn't figured out how to hit it. The pitch still generates weak contact and absurd whiffs. Its movement hasn't really changed; it's still the same curveball that humiliated Alexei Ramirez. It's just a little bit less effective now. Last year, called strikes on curveballs accounted for about 23% of Wilhelmsen's strikeouts. Even if he were to lose all of them, which won't happen, he'd still be an 8 K/9 reliever with lots of ground balls generated and no home runs allowed. That's still really good. So whether it's been a small-sample fluke, or the league adjusting... isn't important. Wilhelmsen is still a great reliever either way.

In summation:

Tom Wilhelmsen's current 0.45 ERA will not last. Eventually he will allow a home run, and hits will start falling in, and some of them will be with runners on base. He will not be this perfect forever. And the drop in K% isn't just luck - so far this season, hitters have been swinging at his curveball in two-strike counts instead of getting frozen. This is a shame, because it was really fun to see Wilhelmsen's curve freeze opposing hitters. But while it's tempting to declare that the league has acclimated to the pitch, this sudden bout of good swing decisions has occurred over a small sample of only about 25 pitches. And even if Wilhelmsen's called strike threes are gone forever, which I doubt, all it means is more weak contact in two-strike counts, which in turn means more ground balls. This just in: Tom Wilhelmsen is really good. You shouldn't be worried.

If the question is "Is Tom Wilhelmsen worse now?", the answer is "maybe very slightly, if you believe in a small sample size".

Or, in other words, "no, not really."