Dallas Braden and the screwball

Dallas Braden
Dallas Braden throws a circle change-up to Willy Aybar in the ninth inning of his perfect game. (ZumaPress, via PicApp)

If you haven’t been hiding under a rock for the past week, you probably know that Oakland Athletics pitcher Dallas Braden threw a perfect game against the Tampa Bay Rays on Sunday, May 9. It was an outstanding achievement and a fascinating story, one that has been ably covered by many other writers.

I, however, was particularly fascinated by one question.

Braden was known as a screwball pitcher coming up through the minor leagues, but after shoulder surgery in 2006, he abandoned the screwball because of injury concerns and because he developed an effective circle change-up, as reported by Athletics Nation.

Nico: There have been reports over the years that you weren’t throwing your screwball anymore, that you had never thrown a screwball, that your screwball and change-up were the same pitch called different things, so this is my chance to try to clarify: What is the deal, past, presen, and future, with Dallas Braden and the idea of throwing a screwball?

Braden: In the past, it was absolutely a screwball. I did not throw a breaking ball—a traditional left-handed slider or curve ball—so I had to add another pitch that was going to create a little more depth, less velocity, and that’s where the screwball came into play. I had arm trouble and then it was eventually shut down in ’05 and needed arm surgery and coming back from surgery I just didn’t feel comfortable using the torque right away and so I stayed away from the screwball, and then coincidentally afterwards I pitched in Winter Ball and worked on a breaking ball in Winter Ball, got the invite to big league camp in ’06 and ’07.

(But) showcasing the screwball, the screwball really wasn’t there for me because I was babying it—I was scared to throw it, I didn’t want to get up on top of it—and Curt (Young) felt strongly enough and had enough confidence in my change-up to go ahead and just bag the screwball, let’s just focus on the slider. And you know what? Because I had such a good feel for the change-up, I could kind of tweak it, make it do different things, versus just trying to throw a traditional change-up, so we went that route and now I don’t throw the screwball. I have variations of (the change-up)—different counts to different hitters, I’ve got a real good command, a real good feel for it.

I have not seen any evidence in the PITCHf/x data that Dallas Braden has thrown any screwballs in the major leagues, at least not until his perfect game. His career roughly corresponds with the PITCHf/x data collection era (2007-2010). We don’t have data for every single pitch Braden has thrown, and I may have overlooked something, so it’s possible he’s thrown a screwball or two before, but it’s certainly not a pitch he has used in the majors on a regular or even semi-regular basis.

So I was very intrigued to read the following from Braden via Susan Slusser after his perfecto:

In the sixth, Kapler had the best at-bat of the day against Braden, who needed 12 pitches before getting him to hit a foul pop toward third.

“I was thinking maybe the knuckleball, the gyroball, the behind-the-back pitch, because I’d tried everything else,” Braden said. “I threw him a 64 mph screwball and he fouled it off. I threw him one more pitch and it was the correct location.”

Can we see any evidence of the screwball in the PITCHf/x data? In fact we can. While we’re looking for it, let’s also compare Braden’s circle change-up to two other off-speed pitches from left-handed pitchers: the screwball of Danny Ray Herrera and the circle change-up of Cole Hamels.

First, look at the difference in speed and in the spin they put on the ball. The following graphs, courtesy of Trip Somers and TexasLeaguers.com, show some characteristics of the trajectories of the pitches thrown by Herrera, Braden and Hamels. The different pitch types, as classified by MLB Advanced Media, are shown in different colors. I’ve added a red circle to each graph to indicate the screwballs and change-ups of interest.

Herrera change-ups speed vs. spin axis angle

Herrera’s screwball is thrown around 65 mph, and he actually has a separate change-up that he throws about 75 mph. He has quite a bit of topspin on the screwball due to pronating his wrist/forearm (twisting it the opposite way compared to the curveball and slider), which can be seen in the spin axis angle of around 50 degrees or so. (A spin axis of less than 90 degrees means the ball has topspin, like a curveball from a right-handed pitcher.)

Braden change-ups speed vs. spin axis angle

Here we see that Braden’s circle change is thrown around 70 mph. You can also see the one slower pitch there at 64 mph that he called a screwball. The spin axis on his circle change is around 130 degrees, whereas the “screwball” had a spin axis of 100 degrees.

Hamels change-ups speed vs. spin axis angle

Hamels’ circle change is thrown harder, around 80 mph. It has a spin axis similar to Braden’s, at around 130 degrees.

The effect of the spin can be seen more apparently if you look at the spin deflection of the pitches. This is how much the spin on the baseball moves the pitch up and down or left and right. These graphs are shown from the catcher’s viewpoint. Again Herrera is first, Braden second, Hamels third.

A Hardball Times Update
Goodbye for now.
Herrera change-ups spin deflection

The spin on Herrera’s screwball makes it drop about six inches on average. That’s partly due to his lower release angle but also due to pronating his wrist/forearm to get additional topspin on the ball.

Braden change-ups spin deflection
Hamels change-ups spin deflection

The spin on Braden’s and Hamels’ change-ups make them rise by about seven to eight inches. That’s why Braden’s change-up is not a screwball. Now, the one “screwball” that he threw rose only an inch or two due to spin. Is that a true screwball? I would say probably not in terms of how it moved. But it is more screwgie-ish than his other pitches, so I wouldn’t argue too much with him about the name.

What makes Braden’s circle change-up look like a screwball to some people when Hamels’ similar change-up doesn’t look that way is that Braden throws it 10 mph slower. That gives gravity a lot more time to act on the ball and makes it drop a lot more. This corresponds better to what the batter or other observer sees when he watches the pitch. He may not know how it was gripped or was spinning, but he can see it dropping.

Here are the previous graphs with the effect of gravity added in:

Herrera change-ups spin plus gravity

Between the slow speed and screwball spin, Herrera’s screwgies drop between three to four feet over the last 40 feet of their travel. That’s a drop similar to a Barry Zito curve.

Braden change-ups spin plus gravity

Braden’s circle changes drop about two feet on average over the last 40 feet of their trajectory, and the “screwball” dropped about three feet.

Hamels change-ups spin plus gravity

Hamels’ circle changes, because they are thrown harder, drop only about 15 inches over the last 40 feet of travel, even though they have the same spin as Braden’s circle changes.

Now back to the question of whether Braden threw a screwball to Gabe Kapler. That pitch is different than his circle change-ups. However, it’s not unheard of for a pitcher to vary his pitches a little bit. Without the quote from Braden, I’d be skeptical of that as a screwball based on the data (spin angle > 90 degrees, i.e., no topspin). With the quote from Braden, I certainly accept that he thew what he considers a screwball. And that counts for quite a lot. If I had to call that pitch something, I suppose I’d call it a screwball, too, but not without some reservation. It’s enough different than Herrera’s screwballs that I’m loathe to consider them as part of a single entity.

I have often wished that we had good data on more screwball pitchers than Herrera. Unfortunately Jim Mecir retired after 2005, before we had PITCHf/x data, and we haven’t had anyone but Herrera since. Does anyone know of any screwball pitchers currently in the minor leagues?

References & Resources
PITCHf/x data is courtesy of MLB Advanced Media and Sportvision. PITCHf/x graphs are courtesy of Trip Somers and TexasLeaguers.

Newest Most Voted
Inline Feedbacks
View all comments
Alan Marshall
13 years ago

Maybe we could get Dr. Mike Marshall to throw the equivalent of a hundred of innings of screwballs in some stadium to create the baseline data you want.  grin

13 years ago

It’s interesting that you chose Hamels to compare to Braden.  Even though you were comparing their changes, Hamel’s curve is relevant to this conversation as well.  He throws it so overhand, with such 12-6 movement, that some of them actually have screwball-type movement as they have “positive” horizontal movement (as labeled on the graph).  In other words, some of his curveballs actually move IN towards a left-handed batter.  I don’t know of many other pitchers who have such a true 12-6 curve.

Now, we wouldn’t call them screwballs because they were essentially thrown the same way as the other curve balls just with a “more overhand” motion.  Just as we wouldn’t call the one curve with positive vertical movement a slider just because he didn’t throw it as sharply or whatever.

Mike Fast
13 years ago

@Alan Marshall, I’d love that.  Piece of cake to arrange, I’m sure.

@ElBonte I agree.  I actually discussed pitch classification methods more in depth on another discussion board that served as the prompt for me to write this.

Quoting from what I wrote there,

what we look for in pitch classification is not just what a pitcher calls his own pitches but also how they compare to what other pitchers throw. E.g., Sabathia calls his breaking pitch a cutter, but it’s not a cutter by the standard of what any other pitcher in the majors throws. By that standard he throws a slider bordering on a curveball. What I look for in classifying pitches is (1) the objective facts about the trajectory of the pitch: speed, spin deflection, etc., (2) how that pitch compares to the other pitches thrown by the same pitcher, (3) how the pitches thrown by that pitcher compare to the pitches thrown by other major leaguers, (4) what the pitcher himself calls the pitch, as disclosed in interviews, (5) what the manager, coaches, catcher, teammates, or opponents of the pitcher call the pitch, and (6) pictures of the grip if I can find them.

Item (2) addresses your point it’s not just the absolute out-of-context facts about the speed and spin on the pitch that determine what it’s called. 

Zito’s curve used to be a true 12-6 curve like that, but he’s changed his delivery and it’s not any more.  There’s probably somebody else, but nobody comes to mind at the moment.  Wandy Rodriguez has a big curve, but it’s not quite as 12-6 as Hamels.  Same with Roy Oswalt.

Matt Lentzner
13 years ago

Interesting possible discovery of Braden’s screwball in the wild.

In the pitch continuum the screwball covers a lot of territory. In theory you could have screwball-sliders, screwball-slurves, screwball-sweeping curves, and screwball-12/6 curves. I think Dr. Marshall identified at least two variations on the screwball although I found his terminology to be very confusing so I’m not sure. He may have been considering a sinker as a type of screwball.

Going back to the types, I can’t see any reason to throw a screwball-slider or a screwball-12/6 curve since those pitches would be very similar to their conventional counterparts. So that leaves the screwball-slurve and screwball-sweeping curve as reasonable candidates. Herrera looks like he’s pitching more of a screwball-curve while Braden’s looks more like a screwball-slurve.

Mike Fast
13 years ago

That’s a good way to explain it, Matt. Thanks.

Alan Marshall
13 years ago

I’ve looked over the data you’ve presented and when back to gameday to watch the trajectory of the pitches.  Both visually and numerically, that 10th pitch to Kapler was unlike any other he threw that day.

My conjecture is that it was the circle change, but taking even more off it than normal.  “The pitch” was 64 mph, and had noticeably more vertical drop than his change.  Thrown at 64 mph instead of 69-71, that ball would drop like a rock as it closed in on the plate.

Alan Marshall
13 years ago

A disclaimer:  As a kid, my two favorite pitchers were Steve Rodgers and Mike Marshall.  I am not related to Mike.

13 years ago

Seems as though the screwball is such a very arbitrary and vague pitch distinction there isn’t really any way to definitely say ‘player x throws it’.  I’ve seen ‘changeups’ tail over a foot with little sink from pitchers; basically a reverse slider, but they are not called screwballs despite the obvious movement of one, seems as though the only ‘true screwballs’ are reverse-curveballs.