American League

National League

Team Audit | Player Cards | Depth Chart
Return to Top

Acquired RHP Jose De Leon from Los Angeles Dodgers in exchange for INF-R Logan Forsythe. [1/23]

In so many ways, Jose De Leon is the classic Rays pitching acquisition. Firstly, he’s a changeup guy. Only the Angels threw changeups more often than the Rays last season. If you look at the last three years, or the last five, or if you go all the way back to 2008 (when we first had full PITCHf/x data), no team has been more reliant on the change.

De Leon’s is a doozy too, and perhaps in precisely the way the Rays most covet. This week BP is rolling out research by Jeff Long, Jonathan Judge, and Harry Pavlidis that centers on "tunneling"—essentially, how pitchers use sequencing, deception, location, and the repetition of their delivery to take advantage of the batter’s fundamental, visual disadvantage. Later this week, you’ll be able to read about the way the Rays have specifically sought to take advantage of this kind of information. For now, let’s look at De Leon’s fastball-changeup sequence.

Last season 219 pitchers had 25 or more instances in which they threw a four-seam fastball followed by a changeup. De Leon did it 41 times in his very short time in the majors. Our new data on tunneling shows us where the pitcher released the ball on each pitch and how far that release point was from the same on the previous pitch. The data also tells us where the ball was when it reached what we’re calling the tunnel point (“decision point” might also be an enlightening, if slightly misleading term for it) and how far that location was from the location of the same pitch at the same point during its journey to the plate. Then we have data on how far each pitch breaks in the final 24 feet (or so), after it passes out of what we’re calling the tunnel; data on the spread between the location of consecutive pitches as they cross home plate; and differences in flight time for the ball by the time it gets to the tunnel point.

Obviously, a consistent release point is good, especially when it comes to disguising a changeup after a batter has just seen the heat. A pitch that is very close to occupying the same air space as the preceding one when it passes the tunnel point is also good. A large amount of late break is good. Big speed differentials can be good, but it depends on the pitch pairing, and on the amount of movement (especially late movement) the pitches each have on them. Large differentials in location at the plate mean you’re changing the hitter’s eye level from one pitch to another, so usually you’d want a big number there in a fastball-curveball sequence, but not so much if you’re going fastball-changeup.

Of those 219 pitchers on our leaderboard, De Leon was just below the 95th percentile when it came to minimizing variation in his release point and at the tunnel point, and in a ratio stat (Break:Tunnel, which is the ratio of the break after the tunnel point to the variation at the tunnel point) our stats crew came up with to help show who gets the most out of hitters being unable to identify a pitch early. He was comfortably in the top quartile in flight time differential, raw post-tunnel break, and was among the bottom 15 percent in Plate Differential.

What I’m telling you is this: De Leon has one of the toughest changeups in baseball for opposing hitters to pick up, especially when he throws it on the back of his fastball. He doubled up after throwing a fastball 88 times last year, so hitters could keep in mind that he was twice as likely to do that as he was to throw the changeup right after a heater. Still, when he did throw the change, they struggled to distinguish it from that fastball, which is how they ended up swinging at just under 59 percent of them, a figure in the 86th percentile of the league when it comes to swing rates on changeups. That’s despite the fact that De Leon generally threw the change off the plate, or below the knees.

Unfortunately, in its current form and given his 2016 usage level for the offering, De Leon’s change isn’t a bat-misser in the majors. It had average movement and a large velocity difference from the fastball, a combination that doesn’t leave hitters completely fooled or the ball completely out of potential hitting zones. He also didn’t throw it as often as a hurler might be well served to throw a pitch with such deceptive potential: only about 25 percent of the time. Batters were able to fight off the pitch, knowing that the fastball (or one of De Leon’s two pretty ordinary breaking balls) was probably just around the corner.

This is the kind of insight tunneling data offers, and might represent the untapped potential the Rays see in their new pitching phenom. De Leon could tweak his changeup a bit, trying to throw it a bit closer to fastball speed. He could try tinkering with his grip in order to get more downward movement on the pitch, especially after the tunnel point. Since his slider is nothing special, he could try shortening it up into more of a cutter, a pitch he could throw out of that same fastball-changeup arm slot and that would look the same up to the tunnel point, adding another layer of uncertainty and difficulty for opposing hitters.

Absent any of that, De Leon is a pitcher with one distinct and valuable skill: that repetition of release point and early pitch movement. His delivery is pretty clean, and he’s so consistent with it that his fastball explodes on hitters and his changeup stays hidden longer than most. The list of things he’s missing, however, is just as long. He doesn’t have electric movement. He doesn’t have even an average breaking ball, at least not right now. He had a below-average spin rate on his fastball last season. Let’s not forget, he’s also had injury issues in the past and has yet to throw even 120 innings in any season.

To finish a cruel laundry list, we must point out that it would seem De Leon lacks command too, or at least that he lacked it last September. Earlier this week we published work from our stats mavens on the value of Called Strikes Above Average (CSAA) for pitchers as an expression of their ability to hit spots and help their catchers earn strikes, in addition to staying out of the middle of the plate. Of the 596 pitchers who threw at least 10 big-league innings last season, De Leon had a CSAA (at -0.88 percent) in the bottom quartile.

We have to splash all of this criticism with the following cold water:

  1. De Leon threw 17 innings over four rocky MLB starts in 2016. The sample isn’t small; it’s minuscule.
  2. In the minors, De Leon has racked up dizzyingly high strikeout rates, powered largely by his fastball-changeup combination. That just didn’t pan out in the big leagues, in a brief stint.

It’s perfectly possible that the Rays are getting an immediate mid-rotation starter in this deal. The minor-league performances suggest as much. The big-league performance could easily be a hiccup. However, I have to issue a counter-caution, too: Don’t ignore good information just because the sample is small. We’ve wondered for years what makes a Quad-A pitcher or a Quad-A hitter. We might come to find that big-league hitters are not as easily fooled by changeups coming out of deceptive deliveries, with perfect release-point mimicry, if they don’t also have terrific depth or a narrow velocity gap from the fastball. Information like this—granular, quantitative information about the way a pitcher does his work—can become useful even when there seems to be very little of it. For that matter, with CSAA, where each and every taken pitch is a datum, there really isn’t even a shortfall of sheer information on hand.

It’s an interesting path the Rays are carving late in this winter. Their signing of Wilson Ramos suggested an interest in competing in the relatively near term (unless he was signed solely as an arbitrage play, which is never out of the question where this organization is concerned), but nothing else they have done recently points in that direction. Rather, with this trade and the one that cashed in Drew Smyly for Mallex Smith and others, the Tampa Bay front office seems to be signaling that they’ll accept their fate as an also-ran in 2017 in order to continue building on the strong farm system they have cobbled together. In that light, trading two years of control over Forsythe (whom they can replace in the short term with Nick Franklin and Tim Beckham, and in the long term with either Daniel Robertson or Willy Adames) for six years of De Leon makes perfect sense, even if he’s an imperfect trade target.

Team Audit | Player Cards | Depth Chart
Return to Top

Acquired INF-R Logan Forsythe from Tampa Bay Rays in exchange for RHP Jose De Leon. [1/23]

The Dodgers are smart. They know Brian Dozier is a better player than Logan Forsythe, and by a decent margin. That’s why they spent most of their winter in a staring contest with the Twins, offering De Leon for Dozier but refusing to go further. They were working with two top executives who have never been truly top executives before, and who might have felt a certain pressure to accelerate a rebuild their fans already wish was over. Derek Falvey and Thad Levine wouldn’t budge and the Dodgers (who might rightly feel that they already have the second-best roster in the National League and who never got to the point of having zero good alternatives to Dozier) refused to pair De Leon with a piece they liked more. (It’s now fairly clear that there are at least a handful of prospects in their own system whom they like more than De Leon.)

When the Dozier door closed, the window in which the team needed to move decisively on either Forsythe or Detroit’s Ian Kinsler opened. It seems the Tigers’ November plans to cut payroll have been abandoned, so this was the easy choice. In De Leon, Andrew Friedman sends his old team a player he might have been able to use well this coming season—as the Dodgers figure to wade through another year of brilliant but injury-plagued starting pitching—but a player on whom he seems not to be terribly high. In Forsythe, he gets a player who has spent the last two seasons discovering just the amount of gap power he needed in order to become a valuable second baseman.

Forsythe is a shade below average as both a defensive second baseman and a baserunner, which isn’t great for a team that may pencil him into their leadoff spot and into an infield already somewhat short on range. He’s become an above-average hitter by raising selective aggressiveness to an art form. Of the 133 batters who saw at least 2,000 pitches last season, Forsythe was the 13th-least likely to swing at any pitch and the 11th-least likely to swing at pitches in the strike zone. That leads to walks, but also to good counts.

Forsythe is an extremely count-sensitive hitter. He likes to be ahead of the pitcher, to feel as though he knows what’s coming and can gear up to drive it to a certain part of the park. Only eight batters swung at fewer first pitches than did Forsythe last season, but when he does swing at the first pitch it’s because he feels he has a read on the opposing hurler. He’s a career .325/.346/.583 hitter on the first pitch. If the count begins 1-0, he’s a .279/.390/.463 lifetime hitter. If it goes 0-1, he drops to .229/.275/.325.

Because he doesn’t add much value (other than positional value) in other ways, Forsythe has to hit to help Los Angeles. It would be no surprise, given his fairly wide platoon splits (a trend he did notably violate in 2016), if the team now sought to bring back Chase Utley as a platoon option. They also have Enrique Hernandez, Jose Miguel Fernandez, Charlie Culberson, Austin Barnes, and Willie Calhoun hanging around. Forsythe has the ability to provide something the team has been missing: a right-handed bat worth slotting near the top of the lineup. If he should falter, though, there are insurance policies aplenty.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
Wow, that discussion of the Rays was the most insightful Transaction Analysis I have ever read! Awesome stuff, Matt!
The Dodgers have to keep Kike Hernandez just to face MadBum three or four times per year.
I feel like for the past few years I have heard over and over that a pitcher wants a ~10mph (or perhaps 10%) velocity difference to be effective, whereas it seemed like Matt was suggesting that DeLeon would have a better change up if he had less separation in between. Am I missing something, or has the general wisdom changed on the optimal fastball/changeup separation?
I believe Eno has talked about the sweet spot being between 8-10 mph, but that's in aggregate. I'm sure there are some guys who benefit from firmer velocities. I *think* the idea Matt was going for was that for a pitch that's already hard to pick up, perhaps the velo separation might give it away a bit sooner. I think the bigger idea Matt was going for was just tweaking it in some way to make it even more potent (perhaps Alex Cobb can teach him the split-change he taught Odorizzi).
What Craig said is generally right. To it, I'll add this.

The greater the gap in velocity between a fastball and a changeup (holding movement constant), the greater the whiff rate. The smaller the gap (holding movement constant), the greater the ground ball rate on balls in play. Felix Hernandez and Zack Greinke have had good success in recent seasons with fastballs and changeups just a few ticks apart in average velocity. The optimal velo drop is different for everyone, and velo drop may not be the most crucial determinant of a changeup's effectiveness, anyway.

In this specific case, I'm looking at the pretty tepid movement on De Leon's change and thinking that he might do a bit better (since it would seem that his best asset with the pitch is the difficulty of picking it up, anyway) with a version that had a smaller velo drop but might induce weaker contact than the current form. It's a theory.

Thanks for the comment; it's an evolving topic.
Yesterday's article kind of completely contradicts your theory.


First, Maddux was convinced no hitter could tell the speed of a pitch with any meaningful accuracy. To demonstrate, he pointed at a road a quarter-mile away and said it was impossible to tell if a car was going 55, 65 or 75 mph unless there was another car nearby to offer a point of reference.

“You just can’t do it,” he said. Sometimes hitters can pick up differences in spin. They can identify pitches if there are different releases points or if a curveball starts with an upward hump as it leaves the pitcher’s hand. But if a pitcher can change speeds, every hitter is helpless, limited by human vision.

“Except,” Maddux said, “for that [expletive] Tony Gwynn.”
My first rebuttal to that is that while Maddux was right about tunneling on a general level, I don't believe (and have never believed, going back several years to when I first heard that quote) that he was right about this specifically. Indeed, while we're still not ready to make definitive claims in this area, I suspect we'll eventually find that minimizing velocity differential at the tunnel point *is* a valuable (though not always necessary) skill.

My second rebuttal is that I don't think my theory of De Leon's struggles centers as much on the velocity gap alone as you might have taken it to. (Obviously, to whatever extent that's true, it's no one's fault but mine.) Rather, it's mostly about having a big velocity differential *without* big movement, and especially without big post-tunnel movement. Without exceptional movement, the pitch relies more on deception, and a big velocity differential is an impediment to that deception.