Tracking R.A. Dickey’s Knuckleball

You all know the R.A. Dickey story by now. Journeyman major leaguer reinvents himself as a knuckleball thrower in his 30s, then refines the pitch to become one of the better starting pitchers in baseball, culminating with his selection as the National League Cy Young Award winner last year. The knuckleball is always a fascinating pitch, and Dickey is a fascinating guy, so there has been no shortage of media attention focused his direction.

While I was not working here at FanGraphs last year, I could not resist taking a belated look at some of Dickey’s dominating knuckleballs from that 2012 season. I’ve selected three particularly impressive pitches from that campaign and used an effect known as StroMotion to help track their movement.

RA Stromotion

While traditional pitches create movement through spin induced by grip and arm action (exceptional curveballs have been measured at 2,500-3,000 revolutions per minute) creating high pressure zones that deflect the ball in the opposite direction, the key to the unique movement of the knuckleball is a near complete lack of spin. The ideal knuckleball makes just 1 to 1.5 rotations (~150 RPM) between release and home plate, causing the drag on the ball to shift significantly mid-flight as the leading seam of the ball rotates slightly, resulting in unpredictable and uniquely sudden movement.

While the pitches shown above are an extreme, an average knuckleball still varies its trajectory from pitch to pitch enough to throw off hitters effectively.  Below, I put together a composite of a more typical variety of knuckleballs for comparison.


You can see why opposing hitters had some problems. If there’s no way to predict where the ball is going to go, hitting it is quite difficult indeed.

However, with a trade to the American League in 2013, Dickey has not yet been able to repeat his success that he achieved in New York.  His walk rate in particular has risen from 2.08 per 9 innings to 3.65 BB/9, well above any mark he posted during his time in New York.  Along with a drop in his strikeout rate, Dickey simply hasn’t been the dominant front-line starter that Toronto was hoping for when they acquired him.

While the inherently inconsistent movement of the knuckleball makes the small sample size movement comparisons I’ve done on other pitchers impossible to take conclusions from, FanGraph’s own Eno Sarris looked at one possible explanation for his struggles earlier this season: the drop-off in use of Dickey’s “power” knuckleball, which he throws in the 78-83 mph range.

This image from that piece really demonstrates how many fewer hard knuckleballs Dickey has been throwing this year.


While a slower knuckleball should allow for more movement, the varied speed and possibly improved control of the increased velocity likely contributed significantly to Dickey’s 2013 season.  Last week, Eno published an interview with Dickey, where he noted that health issues were bothering him but he felt he was able to take the mound and compete even if he was at less than full strength.

Sarris: Why don’t you just take 15?

Dickey: Because I can still muster through six or seven innings. Even though I haven’t had my best knuckleball, I still have seven or eight quality starts. So I can still give something. Just a matter of not being as dominant as I was because I’m missing a weapon or two. It’s feeling better, though, this last week.

While most changes of speed in baseball are done through grip, Dickey’s varied knuckleball speeds are induced through arm speed.  When he talks about “missing a weapon”, it appears that this might be what he’s referring to. I’ve compared his pitching motions this season on several pitches in each speed range to illustrate how Dickey’s adds velocity.

RA compare

RA compareslow

It might be difficult to see, but there is in fact a difference in arm speed and a slightly more violent motion to the hard knuckler, and given his back and neck issues, it appears that he’s less willing to inflict that motion on his body this season. Thus, the switch to more of a traditional slower knuckleball, and while correlation is not causation, it seems likely that this change is contributing to his lessened performance in 2013.

While he may never repeat his Cy Young season, the Blue Jays probably haven’t yet seen a full strength version of R.A. Dickey yet. If he gets back to being able to throw his harder knuckleball in the second half of the year, he might yet give them a glimpse of the pitcher the Mets saw last year.

Print This Post

Drew Sheppard is a writer for FanGraphs, graphic artist and GIF enthusiast. If you have a topic you would like Drew to take a look at in the future using overlay GIFs, please let him know in the comments here or on Twitter @DShep25.

40 Responses to “Tracking R.A. Dickey’s Knuckleball”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. payroll says:

    Maybe he’s not “reaching” quite so well because of the back.

    Great gifs.

    Vote -1 Vote +1

  2. Kevin says:

    And who says knuckleballs are an optical illusion? Great work!

    Vote -1 Vote +1

    • paperlions says:

      The GIFs actually demonstrate the optical illusion as the balls don’t travel the path you expect them to.

      -8 Vote -1 Vote +1

      • williams .482 says:

        Well, yes, I don’t expect them to veer left or right at random.

        How is this an optical illusion?

        Vote -1 Vote +1

        • paperlions says:

          Because if you watch the ball and ignore the tracking, the ball has the illusion of moving in multiple direction, but it doesn’t.

          Vote -1 Vote +1

    • MrSelatcia says:

      Anybody that says a knuckleball is an illusion has never hit against a decent knuckleball pitcher. You have better luck blindfolded when they are on.

      Vote -1 Vote +1

  3. tehzachatak says:

    The StroMotion GIFs are hypnotizingly awesome. Phenomenal post, as always.

    +9 Vote -1 Vote +1

  4. Transient Gadfly says:

    That first gif is blowing my mind.

    +8 Vote -1 Vote +1

  5. Jaker says:

    Phenomenal post and something that many a Jays’ fan have been observing since the start of the season. He’s lost 2-3 MPH on his hard knuckler. He should take 15.

    Vote -1 Vote +1

  6. Wow! Excellent post, excellent graphics. This is great!
    I never noticed before that the catcher doesn’t give RA Dickey a target… presumably because nobody knows where the pitch is going.

    Vote -1 Vote +1

  7. diegosanchez says:

    oh. my. god.

    Vote -1 Vote +1

  8. dc says:

    I’ve never like the approach of not being 100% to prove you are tough or what not..

    granted the jays have run in to issues with pitching, remarkably, Wang and Rogers have provided some stability. With Morrow close and Happy beginning to throw, hopefully he will be able to step aside, and get healthy.

    because jays need him to be 100% if he’s going to get back in this.

    Vote -1 Vote +1

  9. monkeywrench says:

    Old ARPANET technician is sitting in an old folks home right now looking out the window watching ducks when a cold breeze washes in from an open door. At last he feels satisfied knowing that the reason he built the internet has finally come to pass. Happy sigh.

    +15 Vote -1 Vote +1

  10. Carl Weathers says:

    Holy awesome post man! I agree with others that he should take the 15 day DL stint – but wait for Morrow and/or Happ being back.

    Vote -1 Vote +1

  11. Peter2 says:

    Innovative gifs are revolutionizing fangraphs.

    Vote -1 Vote +1

  12. Radivel says:

    Dickey needs a break. Once the others get feeling better, he should take a couple starts off to hopefully close out the season well. Josh Johnson looked good last night, so that’s a start. One of these starts, the Jays are going to score more than 0 runs in one of his starts and he’ll get a win.

    Dickey, we know you’re a manly man, take a rest!

    Vote -1 Vote +1

  13. Skin Blues says:

    I’d like to see some effort by Alan Nathan to describe why a knuckleball appears to (but in reality, doesn’t) change trajectories mid-flight more than other pitches with lots of spin. He claims trajectories are equally smooth for all pitches. I’m inclined to believe that these are accuracy issues with Pitch F/x cameras. How could anybody watch that first GIF and think otherwise?

    Vote -1 Vote +1

    • Well, I would say that type of drastic direction change is very uncommon from the footage I looked through. There were limited pitches using the extreme high frame rate camera that allows a clear visual picture of the movement, but going through most of his 2012 strikeouts there were only a handful more replays on that level. Many more did appear to have the hesitation (if you will) in movement that you can see in some of the composite pitches though.

      I’d need to double check (or perhaps he will clear it up), but I do believe I’ve seen him acknowledge the existence of such pitches – the question being what is typical among the thousands thrown. I certainly couldn’t say on that myself.

      Vote -1 Vote +1

    • Synovia says:

      Its tough to get anything from that first GIF because the camera is moving. Most of what we’re seeing is from that.

      Vote -1 Vote +1

  14. tz says:

    In the second GIF, I love the phantom glove that shows up at the last minute to catch the slowest of the six knucklers.

    Vote -1 Vote +1

  15. Rockwall Tim says:

    Maybe some opponent needs to point out to an umpire that his right foot isn’t touching the pitching rubber…

    Vote -1 Vote +1

  16. JS says:

    Has anyone looked at his indoor/outdoor splits too? A nice breeze helps a knuckleballer by making the ball move more. Toronto being indoors, I would expect that he’d lose some of that extra movement possibly making him more hittable.

    Vote -1 Vote +1

    • flinflon says:

      Dickey said in an interview that the opposite is the case; he prefers pitching indoors because it’s better for knuckleballs. His explanation was that the fewer external environmental factors there were, such as wind, the more control he had over the pitch by making very slight adjustments on grip, etc.

      Vote -1 Vote +1

  17. LRG says:

    Those GIFS are awesome. Great job.

    Vote -1 Vote +1

  18. Oh, Beepy says:

    I like how I’m reading about one guy losing his skill (Dickey) while I watch another guy gain his post by post (DShep)


    Vote -1 Vote +1

  19. Alan Nathan says:

    I have done a fairly thorough analysis of one of the pitches in the first GIF, the one thrown to Will Grimes last year at Tropicana Field. See for the analysis. That particular pitch is one of the very few for which I have both tracking data and high-speed video that both clearly show multiple changes of direction (in this case, two). It is also a highly unusual example of a “gyro-knuckleball”, which you can easily see from the way the ball is spinning by looking at the motion of the MLB logo.

    One thing I discovered by looking very carefully at the Grimes pitch is that there is a missing frame, which occurs at about the time of the first break. That missing frame tricks the eye into thinking that the ball jumps. In fact, it does not. If the location of the MLB logo is plotted, it follows a smooth line and appears to jump only because of the missing frame. All by way of saying that one has to be very careful drawing conclusions about the smoothness of the trajectory by watching video.

    I have looked long and hard for other examples of multiple movements. In an earlier analysis I did (referred to upthread by Skin Blues), I found no examples in a sample of several hundred pitches, at least within the overall precision of the tracking data (less than 0.5″). Drew…I would love to get from you GIFs of the individual pitches, plus any identifying information that would help with finding the tracking data (date, inning, batter, etc.).

    +7 Vote -1 Vote +1

  20. Matt says:

    Very interesting article and cool .gifs on the knuckleball. One quick comment on the statement in paragrpagh 1 about a pitch’s rotation creating pressure zones that cause the “curve”. This theory has actually been disproven, as it is generally just an “optical illusion” that gives the appearance of pitch movement. Here’s a link to a video of the research:

    Or, if you prefer, the article describing the phenomenon:

    It’s a similar effect with the knuckleball, where an unconventional rotation of the ball tricks the eye into thinking the ball “darts” unexpectedly. Maybe a technicality, but just wanted to share some info.

    Vote -1 Vote +1

  21. Alan Nathan says:

    Responding to Matt: The illusion video is very interesting but it has little or nothing to do with the movement of a curveball (or any other spining pitch). We have plenty of PITCHf/x data showing the movement of all kinds of spinning pitches. They really do change direction, it is not an illusion. If you were to contact the authors of the illusion video/article (as I did several years ago when it first came out), they would concede that point.

    Vote -1 Vote +1

  22. channelclemente says:

    Great stuff!

    Vote -1 Vote +1

  23. ALEastbound says:


    I could watch those all day.

    Vote -1 Vote +1

  24. Shawn says:

    Drew Sheppard, these are amazing. I read every article you are a part of. I say Drew Sheppard for mayor of Toronto…. as long as he doesn’t have a crack cocaine problem….

    Vote -1 Vote +1

  25. Stond Jays Fan says:

    Amazing. I am mesmerized by those GIF’s.

    Vote -1 Vote +1

  26. Poinmonster says:

    Fantastic piece and gifs, especially the first one. That is the only way to really see a knuckleball. I do wish he would take a break or at least skip a start. For the sake of baseball, I want a full strength R.A.

    Vote -1 Vote +1

  27. Dave says:

    Subtle thing and hard to measure (that is, I don’t know where you can go back and check game time weather conditions for each of his starts in one website), but he’s always struggled in cold and wet weather and we have certainly had plenty of that this spring. As summer weather comes, he should have more success.

    Vote -1 Vote +1

  28. Jon L. says:

    I would love to see a Fangraphs author take a stab at figuring out the net effect of letting players play through known injuries after they are clearly diminishing performance. Like leaving pitchers in after they tire, it seems to always backfire, but maybe this is just selective attention…

    That said, let me be the 37th commenter to observe that these images are mind-boggling. And the movement on Dickey’s pitches is too.

    Vote -1 Vote +1

  29. curtis says:

    tragectory is out of this world

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *