Investigating Rosenthal's increased walk rate

Last year, Trevor Rosenthal was one of the best relievers in baseball, ranking 6th in fWAR. That was accomplished by striking out nearly 11 hitters per nine innings, making up for a more average 3.28 BB/9. This year, however, Rosenthal is so far walking much more hitters, to the tune of a 7.48 BB/9. In terms of per plate appearance, that's 18.4%. If Rosenthal's opposing hitters were a qualified hitter, they would rank 3rd in BB%, which is good for the hitters, very bad for Rosenthal.

It bears pointing out that an increase in K% makes up for a lot of the increase in BB%. Trevor's 32.7% K% is the highest of his career, and again, if Trevor's opposition was a qualified hitter, they collectively would be 7th in K%. In terms of qualified relievers, it's the 7th highest K% and the single highest BB%.  In the end, the rub is that Rosenthal's xFIP is up from 3.19 in 2015 to 3.40 thus far in 2016. Worse, but certainly not the dumpster fire you might think it is from the excessively high walk rate..

One drastic change for Rosenthal is that his swing rate is down, from 51.8% to 43.4%. That's whether his pitches are out of the zone (30% to 21.7%) or in-zone (77.9% to 64.1%). That's despite Rosenthal's Zone% increasing from 45.5% to 51.1%. Because hitters are laying off a lot both in-zone and out-of-zone pitches at a very high rate, this implies that hitters are just taking a more patient approach in general with Rosenthal, not that they're picking up something different that's allowing them to better discriminate balls and strikes.

As you probably know, Rosenthal is primarily a fastball/change-up pitcher. In his career against LHH, it's a 75/20 split. Against RHH he mixes the curve in more often, with 80% fastballs, 9% change-ups, and 6% curveballs (the remaining pitches are classified as sliders or unknown pitches). First, let's look at some stats for Rosey's fastball, in 2015 and 2016:

Rosey's fastball has lost a tick of velocity, which isn't that big of a deal because pitchers in general throw slower earlier in the year, and he still has the 4th highest four-seamer velocity in the league. It'll be something to watch going forward though. He's getting less Whiffs per Swing (W/S) from his fastball, but an increased ground-ball rate didn't make up for it as his fastball's Pitch Value per 100 pitches dropped from well above average to below average (For an explanation for how these are calculated, check here). Rosenthal's swing rate on his fastball has seen a 17% drop, and the pitches' O-swing% (Out of the zone Swing%) has experienced an 18% drop. So against Rosethal's fastball, hitters have been much more passive, but are not differentiating between balls and strikes better.

Let's look at some selected stats for Rosenthal's change-up, comparing this year to last:

In terms of vertical and horizontal movement, the pitch is about the same. The difference in velocity between the change-up and fastball has stayed the same (speed diff), as well as  the difference between his Vertical Release Point for the two pitches (VRP diff), so it remains at an optimal speed difference with the fastball and it's tough to differentiate between the two pitches coming out of the Rosey's hand. However, Rosenthal has managed way more swings and misses on the change, as well as ground-balls per ball in play (GB%). Hitters have swung at his change 23% less often on the year, but have decreased swings below the zone (B-Swing%) by 35%.

That's where I think the key is. Whereas hitters are judging the fastballs about the same, hitters are doing better at laying off the change-up when it's out of the zone and doing worse when they do swing, despite the similarities between this year's and last year's change. So if hitters are seeing the same change-up in terms of movement and fastball relationship, let's see if there's a difference in location. Here's a side-by-side of Rosenthal's change-up location in 2015, and then 2016:

Overall, Rosenthal pounded the zone much more often with his change-up last year. In 2016, the most often spot for Rosenthal to put his change-up has been about half-way in-between the bottom of the zone and the ground. That's probably ideal when ahead of the count, especially in an 0-2 or 1-2 count. But let's select for when he's been behind in the count:

Whereas Rosenthal was able to throw his change-up for a strike when behind in the count in 2015, he hasn't been doing as well at it in 2016. This year, he has two different concentration points, one up in the zone and one down where his overall change-up selection is concentrated. If hitters are picking up on this then they're much better able to just lay off the change-up and sit on the fastball. That may be the only way to be successful against the fastball, as Trevor currently ranks fourth in four-seam velocity out of 166 qualified relievers this year. When hitters do swing at the change-up, he's getting much more swings and misses, also probably a result of the fact that a larger portion of those pitches have been in the dirt or just above the ground. Overall though, those Swings and Misses aren't making up for less chases out of the zone, as the pitch value per 100 pitches on his change-up has dropped as well, from 3.47 runs above average per 100 pitches in 2015 to 1.91 in 2016.

Over his career, Rosenthal's change-up has been just fine when left in the zone. Visualize the strike zone as three vertical  thirds, a top third, middle third, and bottom third. In his career, here's how Trevor's change has performed in each of those thirds, plus all change-ups below the zone:

As you would expect, the change-up is much better when located down, but overall, in 102 balls in play against Rosenthal's change-up in the zone, hitters have managed just a .137 Slugging percentage on a 51% ground-ball rate. Rosenthal has never given up a homer on a change-up in the majors, whether in the zone or not. It just hasn't been a pitch that you have to worry about being hit hard. In his career, opponents have whiffed on the change-up in the zone 26% of the time, though when it's below the zone that jumps all the way up to 59%.

Opponents are laying off pitches at the highest rate they ever have with Rosenthal, perhaps thinking that's their best chance to get on base against the sometimes wild flame-thrower. While his fastball has been in the zone often, the change-up hasn't had the same consistency. Perhaps teams have picked up that he isn't throwing the change-up for a strike very often, and are sitting fastball. Maybe he's just throwing too many change-ups too low, and it's not so much a league wide adjustment. But then again hitters have also been very patient against the fastball, so it seems like hitters are certainly trying to work a walk, perhaps sitting on a fastball in a certain location as the only way to square him up.

The change-up is incredibly effective when he gets swings below the zone, but it looks like Rosey needs to get the change-up a little higher up more often, and force more swings against hitters. I don't want him to only throw the pitch in the zone of course, as that 59% Whiff/Swing below the zone is delicious. With just 61 change-ups thrown on the year, perhaps Trevor just hasn't developed the same feel he had for the pitch last year. I'd expect that to change sooner or later.