The Mets thought they had something special when they drafted Steven Matz back in 2009. But because of injuries, he didn’t make it onto a mound for his professional debut until 2012. His number were good in both 2012 and 2013 but he didn’t pitch a ton of innings in either season, plus he was a little old for his level both years. While Mets fans were dreaming about what could be, the prospect world was definitely more bearish on his chances.
But things started to change the following season. Baseball Prospectus’ Jeff Moore filed the following scouting report on Matz after seeing him pitch for St. Lucie on May 1, 2014:
Matz’s overall package looks like he’s setting himself up well for a number of years in the middle of a major league rotation. Injuries set him back, but now that he’s healthy he’s developing well and should move quickly through the minors. He has the potential for two above-average pitches and a third average offering. His fastball velocity should allow him to miss bats and his command of it doesn’t have too far to go to be truly effective. His curveball will be effective against lefties, and he’s comfortable throwing it to right-handed hitters to keep them honest. If the change-up continues to develop, he’ll have three offerings he can throw any hitter at any time. His mechanics have been refined from previous reports to the point where they really help his stuff play up. He repeats his delivery well and there is little effort in the motion.
As he continues to gain experience and remain healthy, Matz has a chance to become a number two starter while likely setting in as a mid-rotation guy.
That’s a pretty nice report on a guy who didn’t even crack Baseball America’s Top 100 prospect list coming into the season. Furthermore, Moore graded Matz’ pitches, giving current and future grades:
Fastball – 50/60
Curveball – 50/55
Changeup – 40/50
The change was thought to be the weak sister, both currently and in the future. In his notes about the pitch, Moore called it “inconsistent.”
In 2014, Matz split time between Hi-A and Double-A and pitched very well at both stops. In 71 IP at Binghamton, he had a 2.28 ERA, a 1.127 WHIP and a 4.93 K/BB ratio. It was enough to get him onto everyone’s prospect lists, as BA and BP both ranked him 33 and mlb.com put him at 66 heading into 2015.
As good as Matz was in Double-A, he was even better in Triple-A, despite the hitter-friendly environments of both the league and his home park. In 90.1 IP for Las Vegas, Matz posted a 2.19 ERA, which earned him a promotion to the majors. He pitched well right from the start but after two starts with the Mets, he went three months before making his next appearance for them, due to a lat injury.
Matz ended up making six starts in the majors during the regular season and three more in the playoffs. Everyone knows about his sparkling 2.27 ERA with the Mets, even if most don’t realize it came with an unsustainable 91.4 strand rate. While that figure may be more luck than skill, you still have to be making good pitches to pull it off. So, how did Matz do it?
According to FanGraphs’ Pitch Types, it was mostly with his fastball, which he threw 68.4% of the time. And it’s hard to blame a lefty with an average fastball of 94.3 mph throwing it that often. His primary offspeed pitch was his curve, which he threw nearly 20% of the time and Matz threw his change nearly 10%. It’s in line with what you would expect from the scouting report Moore filed back in 2014.
But the surprise was how well the results were when Matz threw his change. On a per 100 pitches scale, Matz had a 1.76 rating with his change, which is nearly identical to what Jake Arrieta did last year. Of course, Arrietta did it over 229 innings, making his results so much more impressive. Among qualified pitchers, Arrieta had the seventh-best change in the league by this metric.
Imagine if Matz pitches a full season in 2016 and in addition to his outstanding fastball he also has a good curve and a change that ranks in the top 10 of MLB? However, much like with WPA, these Pitch Type Values are descriptive and not predictive. Just because Matz had great success with his changeup in 2015 doesn’t mean he’ll experience the same thing in 2016. Because he didn’t throw the pitch often, it’s possible it merely caught hitters off guard. With scouting reports mentioning the change, perhaps hitters will fare better against the pitch in 2016.
TexasLeaguers shows that Matz threw his change 66 times during the regular season. And while he didn’t get a ton of swings and misses (7.6%), he did throw it for strikes (63.6%) and hitters didn’t feast on the pitch, as they put it in play just 18.2% of the time.
Generally, pitchers use the change most often as a weapon when they don’t have the platoon advantage, as the pitch typically moves away from that batter. And Matz followed that pattern as with his 66 changes, only 14 were thrown to lefties, compared to 52 versus righties. And apparently, Matz threw different changes to righties and lefties. Versus righties, his change had a 1,931 spin rate and versus lefties that mark was less than half, as his spin rate was 845.
By contrast, Arrieta’s spin rate with his change was nearly identical versus righties (2,000) when he had the platoon advantage and lefties (2,011) when he didn’t. Among qualified LHP, Cole Hamels had the best changeup in 2015. And like Arrieta, there was no meaningful difference in his spin rate when he threw his changes to either lefties (1,944) or righties (1,988).
So, it would seem there was room for improvement with Matz with his change, as top pitchers can throw the pitch with as much spin against a hitter regardless of which side of the plate from which he hits.
So when you’re watching Matz in 2016, pay attention to his change and see how effective the pitch is overall. Is he throwing it for strikes? Is it catching hitters off guard and producing weak contact? Does his change spin and move as much against lefties as it does righties?
Chances are, if you answer yes to all three of those questions above that Matz again is having great success with his changeup. And if that’s the case, Matz may be the fourth pitcher in the Mets’ rotation but by no means will he be a mid-rotation starter.
“Because he didn’t throw the pitch often, it’s possible it merely caught hitters off guard.”
How often you throw a pitch is directly related to how effective it is. That’s the key element wrt game theory in pitching. You basically throw each pitch in the exact frequency such that all of your pitches are equally effective given the exact same situation.
His (and other pitchers) spin rate versus RH and LH batters means nothing in an of itself. Clearly he is throwing a different type of change up for whatever reason. What is most important in terms of being able to throw the change up against same side batters is the horizontal movement. The more arm side movement you have, the less you want to throw it to same-side batters.That has to do with arm and wrist angles, as well as finger pressure (basically the spin axis on the ball) and to some extent the spin rate. The less the spin rate, the less the horizontal movement. So in general, you might want to throw the change up with less spin to same side batters to cut down on movement, which is exactly what he is doing. Why other pitchers don’t do that, I don’t know. Maybe their spin axes are more parallel to the ground (less horizontal movement) so they don’t need to change the spin rate.
“His curveball will be effective against lefties, and he’s comfortable throwing it to right-handed hitters to keep them honest.”
Typical scout nonsense. Curve balls typically have one of the smallest platoon splits of all the pitches, contrary to popular belief. Pitchers typically throw curve balls effectively to both same and opposite side hitters. Throwing it to righties (for a LHP) to “keep them honest” is nonsense. In any case, there is no such thing as, “throwing a (presumably bad) pitch to keep a hitter honest.” Again, every pitch that you throw in a given situation should have exactly the same effectiveness. The frequency with which you throw it (along with its intrinsic quality of course) establishes the effectiveness. An intrinsically “bad” pitch becomes good if it is thrown infrequently enough, however if it is still worse, when it is thrown rarely (say 10% of the time in that exact situation), than another pitch (or collection of pitches) thrown 90% or more of the time, then you must throw that bad pitch even less often. If it gets to the point where even thrown 1% of the time, it is worse that another pitch(es) thrown 99% of the time, then you scrap that “bad” pitch.
Editor’s Note – Please do not capitalize words in your post, as that is a violation of our Comment Policy.
I’m honored to have a baseball celebrity read and comment at the blog.
I disagree with your game theory point. Actually, I don’t disagree with the point – I disagree that’s how pitches are called in reality 100% of the time.
Thank you.
Right, there is nothing to agree or disagree with. That is game theory wrt pitch selection. That is also the optimal manner in which pitches are or should be selected in practice. Obviously no one can make those kinds of precise decisions in real time optimally, but that certainly is or should be the goal.
I am not entirely sure what you are disagreeing with, to be honest.
Here is an example of exactly how game theory comes into play on the mound and why all pitches must have the same value in any given situation.
Let’s say that in a given situation a pitcher and catcher are contemplating what pitch to throw, and that this particular pitcher has 3 pitches in his repertoire that he can throw. Obviously each pitch creates a “delta” (before and after the pitch) in each team’s win expectancy. That is defined as the value of the pitch.
Even though pitchers and catchers don’t think they are doing so, in any given situation, they are actually creating an optimal distribution frequency for all their potential pitches, in this case, three, and then flipping an X sided coin (where the chances of landing on each of 3 sides corresponds to that fixed frequency distribution) in their head in order to “decide” what pitch to throw.
If they approached it in any other (non-random) fashion, that would be a huge advantage for the batter. All he would have to do is put himself in the place of the pitcher/catcher and he would be able to “guess” what pitch was coming with a certain degree of accuracy, that degree being better than what it would be had the pitcher/catcher been flipping that 3-side coin (randomizing their pitch choices).
That would be like a person playing Rock Paper Scissors and trying to “figure out” what to throw rather than simply throwing a 3-sided die where each side was equally likely to come up. They would get killed in that game if they did anything but randomize their throws, against a reasonably smart opponent of course.
So, the pitcher and catcher have that frequency distribution in mind, say fastball 60%, curve 30% and change 10%. Then they flip that 3-side coin where fastball comes up 60% of the time, etc.
If those frequencies are correct, then they get the maximum value out of that pitch, whichever one is “chosen.”As well, the values of any pitch chosen must be exactly the same. The reason for the latter claim is this: If any one pitch has a better (for the pitching team of course) value than any other, then, by definition it should be thrown more often. Why would you throw pitch A 10% of the time with a value of plus 10 and pitch B or C 90% of the time with value plus 5?
Pitch A is clearly a better pitch (by definition) in that situation (to that batter in that count, outs, score, inning, etc.). So you should throw pitch A at least 11% of the time, rather than 10%. That is a given.
Now what happens if you throw pitch A more than 10% of the time? The value goes down. The value of a pitch in any situation is a precise function of its intrinsic quality and how often it is thrown. That value bottoms out when the batter knows it is coming 100% of the time, and maxes out when the batter rarely if ever expects it. So thrown at 11% of the time rather than 10%, the value might move from plus 10 to plus 9. Remember that it has to move. The other pitches are now thrown 89% of the time rather than 90% so their value increases a little. Let’s say plus 6 rather than plus 5.
So pitch A is still better at plus 9 than the others at plus 6, so the pitcher clearly and unequivocably must throw it more than 11% of the time.
This process continues, pitch A gets reduced in value as the frequency at which it is thrown increases while the value of pitches B and C (combined) increase as their frequency is decreased, such that eventually we reach an equilibrium point such that the value of all pitches is exactly the same. At that point we neither increase or decrease the frequencies of any of the pitches because that would create an overall value (the value of each pitch multiplied by its frequency thrown, added together) that is worse than at the equilibrium point where all values are equal. At that equilibrium point, the total value of all pitches, accounting for their frequencies thrown is at their maximum, for the pitching team.
Those frequencies vary depending on the precise game situation at that point in time (batter, count, base runners, outs, score, umpire, park, weather, etc.).
The only slight hitch is “future value.” For example, if by throwing a certain pitch a pitcher may increase or decrease his chances of eventual injury then he might throw a pitch more or less than optimally overall. I don’t think this is much of a factor. Another example is that perhaps a pitcher might throw a certain pitch more often early in a game or season than is optimal in order to get used to or work on that pitch such that it has increased value later in the game or in the season. Again, I think that is a small factor.
I hope that is understandable. Even though most baseball players do not think in this manner and most will not understand it, this is exactly what goes on during, before, and after a game. At least that is the goal and anything different is suboptimal.
Editor’s Note – Please do not capitalize words in your post, as that is a violation of our Comment Policy.
Well, it’s important to remember that the Mets are the team that had Rafael Montero throw the same exact pitch in the same exact location a dozen pitches in a row and then crucified him for it — even though he didn’t call the pitch. They are the ones who insisted that lefty relievers were the key to a good bullpen and kept running out stiffs and bums only because they threw with their left hand. The Mets are the team that thought it was necessary to bring in their closer, who was not in need of work, when they had a six-run lead.
Suboptimal is business as usual here.
Also the defending NL Champions!
Hat tip to Frank Viola.