R.A. Dickey on “Letting the Butterfly Loose”
If being a subject of more study, research and incredulity than perhaps any other pitcher in baseball is a burden to R.A. Dickey, the down-to-earth New York Met doesn’t show it. Perhaps it’s because he’s studied, researched and been baffled by the knuckleball more than anyone in the game these days. Maybe he welcomes the help. Given his status as Lead Knuckleball Researcher, it made sense to ask the all-star about some findings about his work and his unique pitch.
Early on, Josh Smolow posted the seminal Dickey breakdown on AmazinAvenue. He found that the 37-year-old had two knucklers — one slow and one fast (with the line of demarcation around 76 mph) — and that he varied speeds depending on the count. Later, Smolow discovered a third super-slow-mo knuckler that moved even more than the others. It came in around 60 mph.
So, R.A., how many knucklers do you throw? “It’s all one knuckleball,” he said. “I won’t change the grip. I’ll subtract speed, or change elevation or location — but it’s all one pitch.” But he also saw it the other way. “It’s like having 15 different pitches,” Dickey added. “The knuckleball can be a sinker, it can be a breaking ball, it can be a changeup.”
This year has been a little different for Dickey. His swinging strike and strikeout rates are at a career high, which led to his first all-star-game nomination. My research found that he’s throwing the 70-74 mph knuckleball less often this year. “Lately I’ve sped it up a little bit,” he said, and added that he’s been “getting a lot of good reactions from that with the hitters — swings and misses.” But what happened to the super-slow-mo knuckler? “Well, I threw three yesterday against the Dodgers. I’ve been throwing it, but you can’t overexpose that one. The hitters up here are so good.”
I don’t have a good a slow one as [Tim] Wakefield or [Charlie] Hough or [Joe] Niekro. I might have a better hard one than they did, but my slow one is not nearly as good. I use mine as a changeup, like I struck out [James] Loney yesterday on a 68 mph knuckler. Which is slow. – R.A. Dickey on his super-slow-mo knuckler.
Physics professor Alan Nathan found that movement and velocity are negatively correlated for the knuckleball.That is, the more velocity you put on it, the less movement you get. Dickey agreed, kind of: “The severity of movement is different,” he said. The hard knuckler might move less on the way to the plate, but it then could take a hard turn at some point. He showed me with an imaginary ball how the slow knuckler dances and flutters in a wider-ranging pattern, while the fast, “angry” knuckler darts and dives more erratically.
Nathan also recently had a more controversial finding about the knuckleball. His last piece showed that, on the single-pitch level, knuckleballs have similar trajectories to other pitches. In other words, knuckleballs move with a “smooth tunnel” that isn’t very different from regular pitches. Instead, the knuckleball’s brilliance is the randomness from pitch to pitch — the unpredictability of each pitch compared to the last one. “I’d be curious to know his sample size,” Dickey said. “My hypothesis would be different. When I look back in slow-mo on video, I’ve seen ones with a circular motion, a corkscrew, and I’ve seen that with my own eyes.” He said he’s seen all sorts of movement from the pitch. Dickey also nodded at the suggestion of a zig-zag movement, which is where the pitcher and the research diverged.
A great piece from Dan Brooks echoed Smolow’s original findings, which showed that Dickey was using his high-movement knucklers in pitcher-friendly counts. On the other hand, he used his lower-movement knucklers in hitter-friendly counts. Pitching is still about “throwing strikes,” Dickey said. “You have to find that foundational one that you can work off of. For me it’s about 76-77 mph; that’s the one I can throw for strikes the most.”
Is it possible that R.A. Dickey has separated himself from other knuckleballers who came before by throwing the hardest knuckleball? That might not be a far-fetched conclusion. A collection of knuckleballers told exactly that to Rob Neyer recently, and they argued that the speed with which Dickey threw his pitch gave him better control. The general idea about knuckleball pitchers is that they don’t have great control.
But that’s not really the truth of the matter, at least not according to the numbers. From this list of knuckleballers, I created a custom list on FanGraphs. Run the numbers and you’ll see that knuckleball pitchers have walked 8.53% of all the batters they’ve ever faced. That rate fluctuates around 8% throughout baseball, and the figure has been fairly stable. Dickey’s walk rate has been hovering around a 6% the past three seasons, and though that’s really good, it’s no better when seen in the context of all knuckleball pitchers.
What does stand out in the context of his peers is his strikeout rate. The group of 80-plus knuckleballers had a 11.68% strikeout rate. Dickey’s 27% strikeout rate so far this year looks gargantuan in context. But fewer than 10% of the league struck out back in the 1920s, while batters are striking out more than they ever — at a 19.6% rate. Index the knuckleballers’ strikeout rates to the league average and you’ll see that they’ve shown a collective strikeout rate that was 9% less than league average. The fact that Dickey currently has a strikeout rate that is more than 37% better than the league average may actually be his most impressive feat.
So what does Dickey think about his additional strikeouts this year? “Just ’cause I’m a knuckleballer doesn’t change that philosophy that hitting is timing and pitching is upsetting that timing,” he said. The fast knuckler may have less movement, he admits, but that also means that “hitters have much less time to evaluate what it’s going to do and make a decision.”
And he does this all controlling about 30% of what happens with ball, by his estimation. Seventy percent of the beauty of the knuckleball comes from its interaction with the environment, or “letting the butterfly loose.”
The prospect of having a pitcher who can accumulate innings with whiffs and great control at a reduced price is one who should make Mets’ general manager Sandy Alderson smile. How many innings does Dickey think he could pitch? He doesn’t think that way, but he does admit that, even if his pre- and post-start routines are about the same now as they were when he was drafted as a conventional pitcher, his soreness is not. He feels like he’s “operating at about 70-75% capacity” and usually “feels strong” at the end of the season.
The knuckleball is a fascinating pitch. Lately, it had fallen on some hard times. In the late ’30s, about 16 pitchers were using the pitch regularly. In 2007, only Tim Wakefield was floating knucklers. With so few active knuckleballers in the majors, baseball fans have still produced reams of research about the unique pitch. And thankfully, there’s still R.A. Dickey, who is ready and willing to talk about the pitch at any time.
I like the suggestion, and I wish it got some actual testing, of Alan Shapiro, a neuroscientist talked to by the Washington Post. He suggested (link: http://www.washingtonpost.com/2012/06/06/gJQAqPfcJV_graphic.html) that an optical illusion – similar to that on curveballs – is responsible for batters thinking the pitch flutters despite it having a normal trajectory.
Love the comments on that WaPo graphic along the lines of “the fluttering isn’t an optical illusion, I’ve seen it with my own eyes!”
Yes. Yes you have. Unfortunately, you have also entirely missed the meaning of the phrase “optical illusion”.
I *think* the fact that he talks about having 15 pitches means that he might see that one both ways. Likening it to a sinker or changeup suggests that it’s like those other pitches.
Yup, ask a neuroscientist about baseball. That’s not at all a semi-educated journalist being lazy and agenda-driven.
I’m so glad it’s been proven to be an ‘optical illusion,’ and that snarky comments can be left by people like ABW who aren’t encumbered by actual baseball knowledge.
http://www.cbssports.com/mlb/blog/eye-on-baseball/21435826/amazing-ra-dickeys-slo-mo-knuckleball
Dickey is a beast.
Enjoyed the insightful article. The appearance of fluttering (whether illusory or real) to batters and the shape of a quantum particle cloud are both as real as they need to be. R.A.’s strikeout rate isn’t illusory, just ask the guys standing at the dish. A correction to his own historical mean is what most prognosticators expect, but it sounds like Dickey’s self-awareness and effort to improve his art could make this year’s performance sustainable.
What amazes me most with Dickey is how he can mix speeds to such a degree that hitters can sit waiting his 70 MPH “changeup” and they still can’t hit it.
As a Mets fan, it’s been fun and amazing following his sudden fame since he signed with the Mets. During spring training of 2010, he gave up tons of HRs and stunk really bad. Most Mets fans including myself were rolling our eyes and snickering at Omar’s latest acquisition. He was the first one cut that spring, and many hoped he would never come back from his trip to the minors.
Fast forward to spring training 2012 where Dickey has a real rough time in one of his early outings. In his post game interview with reporters, he tells them the reason for his lackluster results was that he was just doing research and testing the hitters … gauging their reactions. LOL. With any other pitchers you’d dismiss that as a lame excuse for poor pitching. But as we’ve come to know now, that was all pure Dickey and actually rang true.
Dickey’s evolution has been a process of his own quirky genius and hard work, and it’s been a treat to watch. Hopefully the front office offers him a generous extension this winter.
I missed seeing this back in the spring:
http://www.tribecafilm.com/filmguide/knuckleball_-film42758.html
Amazing timing of the showing of the documentary and Dickey’s 2012 season so far.
Maybe the reason the “fluttering” hasn’t been scientifically documented is the small sample sizes they’re using in their experiments.By RA’s own admission, he has thrown 10′s of thousands of the pitch in his search for mastery. Anyone with more than a passing knowledge of statistics and analysis, knows that sample size and an accurate model of existing conditions are essential to the results of any pitch analysis. Just a small mistake in the force modelling calculations, can lead to massive errors in the results obtained. While I’m not dismissing the works cited, I would really like to see the raw data and methods used to analyze it, before taking it for granted that the conclusions are Gospel.It may be that the flight of the ball was close to standard, because they only used a minimum number of observed pitches thrown, or or over/under estimated the forces at work on the ball.
Liz: If you look at the article that Eno linked to (http://www.baseballprospectus.com/article.php?articleid=15919), you can learn more about how I did the analysis about the “fluttering”. If you have any questions after reading the article, I’d be happy to try to answer them for you.
Nice Work!
Yes, the fast knuckler has less movement on it. However, the advantage is that it’s much easier to locate since it leaves his hand more straightened versus the slower knuckleballs which will begin to flutter almost right as it’s released.
Another advantage of the fast knuckler is that most of it’s movement comes much later than the usual knuckleball. Dickey’s hard KB’s are coming out straight then all of a sudden have these short and violent condensed movements right as the ball is hitting the plate, much like you see with more traditional pitches like sliders and curves.
It just must be so hard for batters to face Dickey. Not only because his knuckleball ranges from 68 to 81MPH, but because the knuckleball imitates all other pitches. Some knuckles have a long sweeping break like a 12-6 curve, others breaking like a slider, some look straight and then all of a sudden drop off of the table like a change up, and others have a hard sinking motion to them which induces a lot of grounders or foul tips.
The scary part is that Dickey’s knuckleball is only going to get better during the next two summer months. The more hot and humid it is, the more the knuckle dances.
Ryan: regarding your very last sentence, my expectation is that there will be *less* knuckleball movement at higher temperatures and higher humidity. Under such conditions, the air density will be less and all aerodynamic effects, such as movement, will be reduced.
I agree wtih Alan here.
I’ve watched a lot of Tim Wakefield and he always had trouble in the heat.
Dickey’s one bad start in June was the first really hot and humid day of the summer.
Just to be clear: did Dickey saying that he’s Less sore now after he pitches than when he was younger?
“He doesn’t think that way, but he does admit that, even if his pre- and post-start routines are about the same now as they were when he was drafted as a conventional pitcher, his soreness is not. He feels like he’s “operating at about 70-75% capacity” and usually “feels strong” at the end of the season.”
Yes, less sore now then when he was younger, ewhen he was a conventional pitcher.
I’ve caught four knuckleballers, and a perfectly thrown, zero rotation knuckleball invariably flutters until it breaks. One with a slight rotation will not, but will eventually break in the direction of that rotation.
I only caught a really hard throwing knuckler for one inning, but that is a beast of a pitch. It is my recollection that Joe Niekro threw his harder than Dickey’s. I’m sure the hard knuckler is a pitch that has it’s own problems, and Dickey probably deserves more credit for mastering it than he is willing to give himself. And if he is throwing it at 75%, that is truly a miracle, and perhaps a big part of his sucess.