Yesterday’s Dancing Knuckleball
Yesterday, knuckleballer Charlie Haeger had a pretty amazing game. He struck out 12 batters while walking four in six innings. He also threw three wild pitches, two of which were on third strikes and got far enough from A.J. Ellis for the batter to reach.
Obviously, based on the 12 strikeouts, Ellis wasn’t the only one having a tough time with Haeger’s knuckler. By my count he threw 94 knuckleballs (along with 22 fastballs) with 43 swings and 13 whiffs. That works out to a 70% contact rate and a 13.8% swinging strike percentage. Both those rates are very good. That contact rate was about where Huston Street and Andrew Bailey were all of last year (not to say that Haeger will maintain such a rate, but just to put it in context). Haeger’s knuckleball was dancing like crazy.
I have talked about the knuckleball before, but here’s a quick refresher. While all other pitch types have consistent movement — fastballs rise and move towards the glove-side, curves drop and move away from the glove-side — and cluster out cleanly in horizontal movement versus vertical movement space, the knuckleball has no clear movement and instead forms a diffuse cloud when plotted in horizontal movement versus vertical movement space. Some move up-and-in, others up-and-away, others down-and-in, and others don’t move much at all. Not surprisingly, the success of a knuckleball is directly tied to this amount of movement. Those which move little are rarely whiffed and hit hard. Those that move far result in whiffs and weak contact. Tim Wakefield is successful because his knuckles have a large spread in movement.
With that in mind I wanted to see whether Haeger’s knuckleballs yesterday had more movement than they did in his previous appearances. Here are the fraction of Haeger’s knuckleballs whose movement fell in one of three categories (movement measured as the square root of the sum of the square horizontal movement and square of vertical movement). I included the value from yesterday, his career and the value for Wakefield’s knuckleballs.
Haeger Wakefield
Career Yesterday Career
0-5 inches 0.35 0.32 0.32
5-10 inches 0.53 0.50 0.48
10+ inches 0.12 0.18 0.20
Haeger’s knuckleball was moving a lot more than previously in his career, though still not as much as Wakefield has averaged through his career. Obviously this one value does not tell you everything, but I think it is a nice metric to show us that his knuckleballs were really moving yesterday — that is if the 12 strikeouts wasn’t enough.
Dave Allen's other baseball work can be found at Baseball Analysts.
Does wind current play a role, I wonder? From what I remember, knuckleball pitchers prefer to have a breeze blowing directly at them to help them generate movement on the pitch.
I believe wind is a major factor with the knuckleball, and not just in how you described it. With a headwind, there is more air rushing at the ball, which will magnify the Bernoulli effect (the physics equations that describe the uneven pressure on the ball created by rotation. That pressure causes the movement). This effect is probably similar for each different kind of pitch, creating more movement on breaking balls, etc., as well as the knuckler.
As Charlie Hough once explained to me, the knuckleball actually benefits as well from seemingly still days. Air currents are not continuous, you see, so the “jumping” motion of the knuckler is actually caused by the asymmetries of the wind force on the travel to the plate. These mini air-eddies cause that sharp and unpredictable movement, which is more beneficial to the knuckleball cmp to other pitches (a rotating ball, e.g. any other pitch, has greater angular momentum and is thus less affected by these micro-eddies).
So yes, wind current does matter, but not just macro-conditions. The micro conditions are just as important. However, it’s nearly impossible to measure those. I also expect diminishing returns on the headwind, as I suspect a faster headwind will make the microeddies move in a more consistent general vector; that will increase the movement of the knuckleball, but it will also make the movement less random (no darting in different directions like the true knuckleball; instead, it may resemble more of a knuckle curve).
If my memory and a quick Google serve me properly, most knucklers prefer to pitch indoors as there is no air flow to speak of allowing the pitch to move on its own and in a more consistent fashion.
Without looking things up (mostly because I don’t know how); Wakefield has typically performed above his own average at Tropicana field and the Blue Jays have specifically opened their dome when Wakefield is scheduled to pitch and the weather permits. Both anecdotal, but I am sure someone smarter than I could put some numbers behind his performance at Tropicana compared to his norms.
Maybe I am not as inept as it seems.
All stats are for Wakefield
At the Trop
505 PA / .651 OPS Against / .240 BABIP
At the Metrodome (RIP)
310 PA / .683 OPS Against / .258 BABIP
Against the Rays (home and away)
900 PA / .667 OPS / .251 BABIP
Against the Twins (home and away)
629 PA / .706 OPS / .280 BABIP
At Home (entire career)
6469 PA / .741 OPS / .278 BABIP
On the road (entire career)
6208 PA / .743 OPS / .271 BABIP
I’d say he does pitch better in doors.
For reference, Wakefield averages over 700 PA / season.
Dave, any thoughts on a measure of variation of movement? Could be both direction and magnitude, but I’m guessing direction would be more effective. If a knuckler always breaks the same way, a hitter could anticipate that. Maybe use the angle of the polar coordinates?
Speed of the pitch? Are they similar enough or this not to be a factor?
God, I love knuckleballers!
I wish more pitchers were good at it. I’m rooting for Haeger big-time.
I don’t have anything to offer from a stats perspective on this one, just a thanks for posting this.
Not to be a pain, but this article consists of the following misspellings of “knuckleball”: knuckelball, knucklball, knuckeball. All in the first two paragraphs plus a sentence. Just letting you know.
That data at the end is still pretty cool though. As a disgruntled partial Marlins fan who didn’t see the game, I figured they were just miserable at hitting slow pitches, like they always seem to be.
I love the knuckleball, but alas, I fear the best hope of the knuckleball surviving past the Wakefield-era would be Eddie Bonine. He isn’t even that impressive, however he does have a low-90’s fastball, and uses the knuckleball as his breaking pitch. It’s an interesting take on knuckleball, and I wish he went to it even more.
Using JUST the amount of movement at the end of the pitch can be misleading with a knuckleball — they are well known break TWICE on occasion: for example, breaking outward at first, then reversing to cross the middle of the plate… it would be interesting to look at the BABIP the knuckleballs with the least movement…
Surely if a knuckleball moves enough (had a number of ”breaks” in its flight) its most likely movement would most often be 0? If thats true should pitch movement not be a relatively poor measurement of the quality of such a pitch?
I love that a “Wild Pitch” can also be a strike. Only a Knuckleballer could do that.
There can be only two knuckleballers…the master and the apprentice. That is just how baseball is these days.
Why do so few pitchers, if any, throw a ‘regular’ fastball (~90) and a knuckleball?
my answer to that question would have to do with how knuckleball is perceived by coaches et al. the knuckleball tends to be used when pitchers don’t seem to have enough “stuff” to make it to the bigs; along with that is the opinion that the knuckleball is most effetive when it is the primary pitch. i’m not claiming those beliefs are universal but i believe they are common… and because of this most pitchers with a fastball up around 90 usually are considered to not be desperate enough to learn the knuckler. anyway, this is off the top of my head so (other readers) please correct accordingly! and, finally, i would like to join the chorus of love for the knuckler!
Knuckleball’s a tough pitch to control, it’s easier to learn the “conventional” baseball pitches.
There probably is some truth to those who try and develop it when their stuff if graded “pedestrian”.
I suppose many would apply that argument to Yoshida Eri as well, who recently signed with the Chico Outlaws.
http://www.youtube.com/watch?v=WGIIz9wwU8s
That trip down memory lane reminds me of when I used to live in Virginia Beach. White Trash and Dookie were on heavy rotation in my Walkman while I would skate to my friends house.