August 9, 2015
The Next Collin McHugh?
A few weeks ago we introduced some work that took an in-depth look at pitch spin featuring the insights of the terrific baseball physicist Alan Nathan. We explored the concept of gyrospin versus useful spin, a topic that Nathan had written about previously here at BP. This post then, is the payoff. We promised to make an attempt at finding "the next Collin McHugh;" below is part one of said attempt.
Here's the truth, though. We can't find the next Collin McHugh by looking at spin rates alone. Somewhere along the line this story got twisted.
There's this idea that the Astros looked at some data, saw that McHugh spun a good curveball, and scooped him up out from underneath the Rockies' noses, and that upon landing in Houston McHugh turned into a good MLB starter while changing almost nothing about his approach.
We gloss over the fact that Houston's pitching coach, Brett Strom, tweaked McHugh's approach in a significant way. We ignore that McHugh all but abandoned his sinker and now uses his four-seam fastball sparingly: He has MLB's lowest percentage of fastballs thrown, non–R.A. Dickey division.
Finding the next McHugh isn't as simple as finding a guy whose breaking ball has a ton of spin, but who hasn't had the good results that should accompany it. It's about finding a guy with a great breaking ball who, with some tweaks to his approach or repertoire, has breakout potential. It requires a convergence of analytics, scouting, and coaching. It requires an open mind and a vivid imagination.
Can Statcast get us there? It certainly helps.
Dozens of factors impact a given pitch's effectiveness, such as total movement, when the movement occurs, velocity, pitch mix, pitch usage and approach, and command. All play a role in determining how effective a pitch is.
What we're doing here is looking at pitch spin as a leading indicator for effective movement, something that's arguably a logical leap in the first place. It's important that we establish how strongly correlated the various movement statistics we'll discuss in this post are to overall pitch value. We'll use the linear weights value of each pitch per 100 offerings to determine pitch value.
All the data points we'll discuss—peak spin, useful spin, useful spin ratio, etc.—range from 3 percent to 7 percent in terms of correlation with pitch value. This is not very strong, but it does help provide some direction in terms of identifying potentially valuable pitchers, especially when used in conjunction with other relevant data points like the overall value of the pitch, pitch-usage rates, and so on.
Below is a table of data, sourced from both Statcast and PITCHf/x, highlighting some of the best—in terms of movement—sliders in the game.
The names above are the top 10 pitchers in ratio of useful spin to total spin. A quick refresher might be in order, so we'll use Corey Kluber as our test subject here. Kluber's slider (if it really is a slider) spins at nearly 2,500 rpm on average. More than 80 percent of that spin, about 2,050 rpm, is useful in generating movement. This spin number is derived from PITCHf/x data and an algorithm identified by Alan Nathan in laboratory tests. So far this season, Kluber's top slider peaked at more than 3,300 rpm, which equates to roughly 23 revolutions en route to home plate if we assume 0.42 seconds of flight time.
Masterson leads the pack in terms of ratio, though his slider doesn't spin nearly as much as Kluber's. There are some familiar names on this list, but it's largely populated by relievers, some of whom I have written about in this space before. One name, though, presents a massive opportunity.
Tyler Lyons is a 27-year-old starting pitcher in the Cardinals' organization. He's a 6-foot-4 lefty who has made 18 starts and 11 relief appearances over three major-league seasons. He throws one of the best sliders in baseball in terms of spin, with an average total spin rate of 2,060 rpm. More than 65 percent of that spin is useful, and his best pitch peaked at more than 2,600 rpm. That said, Lyons' slider actually has a negative linear-weights value.
Lyons has bigger problems. Over his career he's posted a 4.72 ERA with a slightly better 3.88 FIP. In 2015 Lyons has posted a 6.79 DRA, but his cFIP- of 99 suggests that there's a league-average pitcher in there somewhere. Perhaps with some adjustments Lyons can not only reach the middling potential that cFIP suggests exists, but exceed it. There's some work to do however.
I present to you the (probably) first ever GIF of Tyler Lyons pitching:
That pitch came in at 90 mph and ended up about waist high and over the middle. It's a good thing the batter didn't swing.
Lyons is an unremarkable pitcher outside of his slider spin. He throws five pitches, but leans heavily on three of them. His four-seam fastball sits at 91 mph on average and makes up 30 percent of his offerings. His sinker is about half a mile slower, and is thrown nearly as often. Lyons' slider is his third-most-used pitch at just over 26 percent, and sits at about 82 mph. He also throws a changeup (11 percent) that averages 84 mph and a curveball (4 percent) that has considerably more drop than his slider. Lyons' pitch chart looks like so:
His fastball and sinker are very similar pitches. The sinker has up to five inches more drop, and another five inches more arm-side run compared to his fastball, but both pitches average just over 90 mph. His sinker generates groundballs on more than half the balls put in play against it, while his four-seamer isn't even close to being the groundball pitch that his sinker is. This makes sense considering where he throws those pitches:
Lyons clearly works up in the zone with his four-seamer and down in the zone with his sinker. This makes intuitive sense given the movement profiles of those two pitches. The perils of pitching up in the zone however, are well documented. That said, opposing hitters actually square up Lyons' sinker more often than his four-seam fastball. Over his career, opposing hitters have a .036 better average and a .026 advantage in ISO against the sinker.
Lyons, much like McHugh, would likely benefit from ditching the sinker in favor of moving the four-seam fastball around the zone more.
This brings about another problem however. Lyons really only uses his sinker against righties, with 50 percent of his first pitches to opposite-hand batters being a sinker. Increased changeup usage could take some of those pitches away, especially since his change closely resembles the sinker already in terms of movement. He could also throw more sliders early in the count, using it as a tool to get ahead.
Lyons throws about one third sliders to left-handed hitters when they're ahead or even in the count. Against righties it's more like one tenth. Throwing more sliders, especially for strikes, would allow Lyons to use his fastball more strategically. There's also the potential for improvement by tweaking and optimizing the changeup, especially as a weapon against righties.
Over his career right-handed hitters have posted a TAv more than 100 points higher against Lyons than lefties have. That means that enhancing the changeup to go along with his slider, four-seamer, and show-me curve could be all that's standing between Lyons and a slot in the middle of a big-league rotation. An acquiring team could encourage him to work with a veteran pitcher on the staff to learn a new changeup grip that might prove to be more effective or easier to command.
The final missing piece here is that Lyons has been beat up because he's not working up or down particularly effectively. Take a look at his zone profile for ISO against:
The vast majority of the damage against him has come from that band across the middle of the zone. Quite frankly, that's something that can be fixed relatively easily, assuming Lyons can command his fastball decently. Working up and down in the zone while trying to avoid those hot areas should be the plan of attack for Lyons every time he takes the mound.
Lyons is a decent pitcher who has a seemingly good slider with a host of average accompanying offerings. As a result, he relies too heavily on the slider, something that potentially hurts the pitch's value and keeps him from reaching his potential. There's no indication that the Cardinals are going to designate Lyons for assignment anytime soon, but he could likely be had on the relative cheap given the Cardinals' pitching depth. If the cost of acquiring him matches his Quad-A/swing-man title, there's real opportunity for positive value.
Thanks to Alan Nathan for his help in deciphering, manipulating, and understanding the Statcast and PITCHf/x data, and for his ongoing support of this analysis.