Do Cutters Kill Fastball Velocity?
“Typically what we see is the more you throw that cutter, you can become dependent on it and you start to overuse it and typically what happens to guys that overuse the cutter is their fastball velocity drops. That has been consistent over the years.” — Orioles Director of Pitching Development Rick Peterson
Last week, Orioles Vice President Dan Duquette staked out a stance for his team: “First of all, the cut fastball, we don’t like it as a pitch, okay?” Focusing first on his contention that no frontline pitchers have succeeded with the pitch, and then on his opinion that the pitch didn’t lead to good results, and finally on his assertion about the developmental impact the pitch has on young arms, we found last week that his stance was defensible if unevenly defended.
But hidden within what Duquette said, and what Rick Peterson said afterwards, is an idea that should be testable. Both Duquette and Peterson made reference to the fact that young pitchers that use the cutter too often adversely effect their fastball velocity.
Peterson was actually very upfront about the relationship he sees between young cutter usage and velocity drops: “a cutter thrown 40 percent of the time for a young power pitcher can become a crutch, then your velocity drops,” he said in his followup to Duquette’s on-field talk with Steve Melewski.
Let’s see what we can find. Using the filters on our leaderboards, I put together a list of qualified starters that used the cutter more than 20% of the time in any given season since 2002. There are precious few veterans that throw the pitch 40% of the time, and it seems folly to make any assumptions based solely on Andy Sonnanstine, Dan Haren, Roy Halladay, Miguel Batista and Jesse Litsch. Since we have two classification methods on our site, I repeated the filter with the 20% cutter threshold for both BIS and PITCHf/x classifications.
The result was a list of 40 pitchers that became our sample. Pitchers that appeared on the list multiple times included Andy Pettitte, Chad Billingsley, Dan Haren, Doug Davis, John Danks, Jon Lester, Mark Buehrle, Mike Leake and Roy Halladay among others. Jeff Zimmerman was so kind as to run an aging curve on the sample:
Well, it certainly doesn’t look like cutters, as a monolithic group, lead to fastball velocity drop. But there are a few caveats buried beneath what look like damning results.
Not one of these pitchers threw in their age 21 season. And if you scan the names above, you’ll see some late-bloomers on the list. There might be something to the fact, as Peterson says, that “the cutter is a pitch that typically is thrown later on in your career, often after you’ve been in the big leagues several years.” Unfortunately, this isn’t something that’s easily test-able. Our big league stats on this go back to 2002, and minor league PITCHf/x stats are not readily available. It passes the sniff test, but then again, it still doesn’t explain why these pitchers didn’t see the velocity loss Peterson predicts.
There’s another twist that comes if you take a closer look at Peterson’s comments: “What happens is you start to get off to the side of the baseball (with your grip) and then you’re no longer consistently behind the baseball.” There’s a belief in baseball that there are two pitches commonly called the cutter. One, called the ‘grip’ cutter, or the ‘cut fastball’ focuses mostly on grip: you hold the ball like a slider, but you throw it like a fastball. You might point to Mariano Rivera, as Melewski did, for a prototype version of the pitch.
The other, called the ‘load’ cutter, or the ‘baby slider,’ has a release to it that is not unlike a slider. Typically, this pitch comes in slower than a regular fastball and slower than the ‘grip’ cutter. Sometimes it has more depth to it than a ‘grip ‘ cutter. And, since Duquette poo-pooed mention of Rivera’s cutter as a fastball, and Peterson emphasized the fact that pitchers get off to the side of the ball, it’s completely believable that they were talking about the baby slider when they denigrated the cutter.
If we were to cut our cutter pitchers into the two groups, it wouldn’t provide us a robust enough sample to say much. As it is, we were limited to 40 pitchers. But, if we were to set two and a half miles per hour difference between the ‘cutter’ and the fastball as a cutoff between the two ‘cut’ pitches — about halfway between a fastball and the five-plus mph slower that a slider travels — we could get a subset of baby slider pitchers. Adam Wainwright, Andy Pettitte, Barry Zito, Chris Carpenter, Cliff Lee, Cole Hamels, Dan Haren, Esteban Loaiza, Gavin Floyd, James Shields, John Danks, Jon Lester, Jon Niese, Josh Beckett, Josh Tomlin, Kyle Davies, Kyle Kendrick, Mark Buehrle, Miguel Batista, Nick Blackburn, Tim Hudson and Tim Stauffer could be your baby slider list.
Since it’s just over half the list, it’s doubtful that removing them would do anything but decimate the sample. Focusing on them as a mini-sample doesn’t produce much, either. There’s a whiff of the late-bloomer about the group, but Wainwright, Haren and Lester were highly touted prospects that did well from day one.
As for velocity loss, it’s unclear. Zito has lost 3.3 mph off his fastball since 2002. He’s 34 now and ‘should’ have lost three mph over that time frame according to Bill Petti and Jeff Zimmerman’s research. Loaiza lost 6.3 mph off his fastball from ages 30 to 36, when he should have lost 3.8 mph, but his career ended with shoulder issues. It’s fair to ask what the role of injury was. Tim Hudson has lost 2.6 mph off his fastball… since 2002… when he was 27. No other ‘baby slider’ pitcher sticks out in terms of velocity loss.
There’s little evidence of accelerated velocity loss among the veterans that use either form of the cutter. There is still a chance that the pitch isn’t good for the development of a young pitcher — certainly the Orioles are not alone in that opinion — and that the veterans that have turned to the pitch are better prepared, and therefore avoid velocity loss.
But then the question becomes when a prospect is ready for the pitch. The Royals ban it before Double-A. The Orioles seem to want to avoid the pitch until the major leagues. Other teams have similar philosophies. Maybe there’s an age at which a pitcher can turn to the pitch without consequence as a fully-formed adult. Maybe we haven’t seen the pitchers that turned to the cutter too young because they lost too much velocity and never made it to the big leagues.
Or maybe the pitch doesn’t actually lead to velocity loss.
Unless the issue isn’t velocity loss, but instead when you’re in the minors it’s building and maintaining max velocity and consistent velocity, and the cutter adversely affects that development? You sort of address that in the last paragraph, but it seems like, at least from a developmental perspective it’s still a bit nebulous.
Which is to say it could also be a counterfactual, and kind of fruitless to explore–”Well, Pitcher A could have developed more velocity, if only he hadn’t thrown a cutter” isn’t really testable.
Hmm, probably. But couldn’t you still compare outcome to scouting expectation?
Did Al Leiter not make the cut because he threw the pitch too often and too early?
Al Leiter is in the 40-man sample. I didn’t think he threw a baby slider.
As a suffering Mets fan I thought he did due to his troubling lack of velocity toward the end of his stint. He was not young so maybe it was just natural age related decline but I not so fondly remember him burying righties with a nasty cutter when he was in his prime, but then getting hammered toward the end.
Dwight Gooden’s velocity decrease after the 1985 season was largely blamed on him starting to throw a cutter beginning in 1986 after being told to pitch to contact more by Mel Stottlemyre. We don’t have any velocity data, so that’s all third-hand info. I wonder if Gooden’s fallen star still has repercussions in 2012. Cocaine didn’t take that fastball away, and maybe it more about 276 IP at age 20 than learning to throw a cut fastball.
There’s a glaring problem with this analysis: Duquette is talking about development of young, minor league prospects. ES is analyzing major leaguers. That’s apples and oranges.
Perhaps the players who make the majors are the survivors whose fastballs didn’t suffer from pitching the cutter. Perhaps it’s just a group that ages better. Perhaps DD is just nuts. Either way, I don’t think this sample is an appropriate group to use to test DD’s statement.
Well I got to that eventually but maybe you didn’t make it that far. I’d be surprised, though, if cutter usage felled all sorts of prospects but was fine for guys just a few years older.
I’ve heard sentiments like Petersen since Avery and Abbott tailed off dramatically after strong career starts. But Pettitte threw a lot of cutters his whole career with no noticeable impact to velocity.
I actually had some text in about Pettitte but cut it: He lost two-to-three mph LESS than he should have according to aging curves.
PED obviously….. Yankees fountain of youth serum. Say what you will PEDs may not make you a better player but it certainly can enhance your natural talent and for aging players, stall declines right?
I would say that the REAL reason not to use cutters is that it may be too “easy” to get minor league hitters out with cutters, so it’s better for the development of the pitcher to work on other pitchers…
And Duquette and Peterson do say things like this. I do not think the stance is unreasonable. I was more interested in the particulars.
An interesting study, Eno — thanks. The difficulty in assessing this question is the attrition problem, which you mention in your second-last paragraph, and which O’s Fan expands on. It might also apply to “veterans” who take up the pitch after reaching the majors: If Messrs. Peterson and Duquette are right — and they can be right on average without this effect occurring for ALL cutter devotees — then some will start to lose effectiveness and therefore pitch less or filter out of MLB altogether. The “qualified starters” in the sample will be the survivors, i.e., those clever enough not to fall victim to the effect. Perhaps your sampling method considered this and corrected for it. If so, my apologies for not reading carefully enough; if not, then the finding in this sample is that there appear to be pitchers who can, indeed, avoid the “cutter effect,” rather than that such an effect does not exist.
Here’s how I attempted to deal with the sampling issue: it’s not only qualified starters for every year, it’s anyone who ever put up one year that qualified and had more than 20% cutter usage. And I stand by that sampling — a pitcher needs to have at least one year of pro work for us to know they are a pro. Sonnanstine is on this list, as an example. So I’m not just looking at qualified pitchers every year, I’m looking at anyone who put up ONE qualified season with 20+% cutter usage. If you’re saying they fell apart in the minors because they used the cutter too much, well, that’s an untestable statement, as I said.
Thanks for the clarification, Eno. Amazing that since ’02, only 40 SPs have logged a season with > 20% cutters. Anyway, isn’t there nevertheless some attrition/survivor bias in the sample? E.g., suppose I qualify for the sample in ’02 at age 27, but the “cutter effect” emerges in ’03 and I start to lose speed and effectiveness, and by ’04 I really stink and am out of the league at age 29. As a result, my declining velocity is only apparent in the Atlantic League and does not pull your blue line down as much as it might if a ML team foolishly kept me around. Unless I’m missing something (and, again, apologies if a more careful reading would have covered it) this is much the same survivorship issue that afflicted early studies of peak performance and aging. I’m not expert on all that lit, but perhaps those researchers devised a work-around.
No, no, you are correct in one point about survivor bias. In your example, though, you’d show up in my sample, and then your velocity would drop in subsequent years, and I used all years by any pitcher in the sample, so I would have seen it and it would have been included in the aging curve. But there is a possible survivor bias if you think a high-A pitcher adopts the cutter, begins to see velocity drop before he hits the major leagues, and never makes it. Or makes it at the reduced velocity and pitches sparingly. He wouldn’t show up here.
I think the biggest issue with this analysis is that your sample only included “qualified” starters. If indeed cutter usage does lead to velocity declines, what we’d see if a higher attrition rate in cutter pitchers, and as a result fewer qualified pitchers. This analysis is basically taking the successful subsection of pitchers which, if the hypothesis you’re testing were to be true, is probably not representative.
It did not only include qualified starters. it only included pitchers that had one qualified year. If they never managed a qualified year, how do we know they are a major league starter.
When I hear about a player who has revived his career and become a totally different pitcher, it seems to me like half the time the change can be explained by the integration of a cutter into his repertoire. Maybe the Orioles don’t want their players to learn a cutter so that if they begin to fizzle out, they can learn one and basically have a second career.
I’m only half joking, by the way. Given the reality that everyone loses velocity and the longer you spend in the league, the more the league figures you out, I wonder if there’s an actual potential benefit to avoiding certain pitches early in one’s career so that he can acquire new pitches as his velocity decreases and the league learns all his tricks. They say it’s a game of adjustments.
It’s worth pointing out that pitch f/x struggled to differentiate types of fastballs in earlier years, certainly before 2009. This may have lowered the average velocity of many cutter-throwing pitchers, giving them a lower absolute, higher relative start point on the left side of the graph. As time progressed and classification algorithms improved, the separation of a ‘true’ fastball or 4-seamer may have offset the inevitable decline in velocity associated with increased age. This, in turn, may obscure or exaggerate ‘real’ differences in the change in ‘fb’ velocity over time between groups of pitchers, if there is any. It may be of use to compare league-wide occurrences of the pitcher, as deemed by the classification methods.
By no means does this provide evidence to support the handling of Bundy.
This is a great point. I tried to avoid it on the cutter side by taking any 20+% season in any year, but on the fastball side, it could change the velocity numbers, and you can see it if you scroll through the players of the era. In discussing the baby slider group, however, I used ‘FBv,’ which comes from PITCHf/x, and I believe it’s a little more stable — it includes two seamers and four seamers, even today I believe, and therefore hasn’t had as much put in or taken out.
Glad to see another article following up on the subject and trying to actually look at the effects of each type of pitch. Obviously we can’t draw any meaningful conclusions because of the issues that have already been articulated, but it’s an interesting exercise nonetheless.
One thing I would point out is that not all “grip” cutters are thrown with a slider grip. Maybe some guys do, but Rivera for one uses a 4 seam grip that is just barely modified in the direction of the slider grip. I think in general the grip is more of a hybrid in between the two, though that’s tough to know for sure without asking the pitchers or getting closeup pictures of the grip they use.
I believe sliders eventually kill fastball velocity and many sliders are now being classified/called “cutters” now a days. Curves are being called sliders, frquently, too.
Neftali Feliz is one example. He would go 97 fastball and 81 “slider” (or so he called it). Then he came up with a “cutter” in spring of ’11 at 88. That ain’t no cutter!
You can add Lincecum to that potential list.
Enos,
two questions come to mind.
1. How much is the cut fastball a next best refuge from the slider, a proximate cause of arm damage.
2. Is there an application of the law of diminishing returns to fastball velocity as it effects pitching performance.
Eno, first off, love your work… but this whole “cutter” issue seems overblown mostly because Duquette is a blowhard (total hearsay).
That said, I think pretty much every org wants it’s young pitchers throwing 90% fastballs and focusing on fastball command before and perhaps up to A-ball and then learning a breaking pitch and change-up, and finally exploring other pitches (splitter, cutter, knuckler, gyroball, etc) when they’ve mastered the basics or failed to develop in the traditional manner.
I think Duquette and Peterson’s poorly delivered point is that the cutter grip and success with it in the minors leads to optimizing one’s delivery for the cutter at a time when one should be optimizing his delivery for a 4-seam or 2-seam fastball. While the cutter should be thrown with a 4-seam/2-seam motion, that motion has presumably not yet stabilized. I don’t see any reason why this might not effect the “finished” velocity.
By this reasoning, presumably, everyone who throws the cutter either learned it in the low minors so that their fastball velocity was diminished (from what it could have been) by the time they reached the majors, or they learned it in the high minors or majors so that their fastball velocity would be unaffected.
The only way to disprove the point is to randomly select a group of young pitchers in Rookie ball perhaps, treat one as a control, one as a treatment, and teach one group the cutter and let them use it freely in the minors.
crap … s/it’s/its