The Best Pitches of 2011: Curveball

Today marks part four of our week-long Best Pitches of 2011 series. You can find the fabulous entries on the fastball, slider and changeup here. I am tackling your other favorite breaking ball, the curve, and Ben Duronio and Eno Sarris will close out the series tomorrow.

Whether you call it Uncle Charlie, yellow hammer, the hammer, yakker, Public Enemy No. 1 or just simply a curveball, it’s one of the more charming pitches in baseball. Many try to throw it, but few master it. It is most definitely not the breaking ball of choice — over the past two seasons, Major League pitchers have thrown sliders five to seven percent more frequently than they have the curve.

If you go back to Carson’s article from Monday, you can see what the criteria we have been using to evaluate the pitches, but I added a category of my own — called strikes. Take a look at this chart:

Pitch Avg Velo Avg SwStrk% Avg Called Strk%
Fastball (FA) 91.7 6.31% 19.78%
2-seam Fastball (FT) 90.8 5.35% 19.58%
Sinker (SI) 90.3 5.15% 19.49%
Cut Fastball (FC) 88.4 8.40% 16.53%
Curve (CU) 76.7 10.43% 19.39%
Slider (SL) 83.5 14.31% 15.79%
Changeup (CH) 82.6 13.37% 11.28%

Though curveballs have the noticeable disadvantage of having the slowest average velocity, they garner as many, if not more than, called strikes as do the various iterations of the fastball. There are simply few things better in life than watching a hitter’s knees buckle as a curveball snaps harmlessly into the zone for a called strike, so I wanted to account for that in my evaluations. And with that, let’s get to our top three, which are presented in no particular order.

Note: The average movement for a curveball in 2011 was 3.1 X-move and -6.0 Y-move.

***

Roy Halladay

PITCHf/x: 17.1%, 77.0 mph, 5.9 X-move, -1.7 Y-move
Results: 18.27% SwStrk, 68.70% Strk, 25.38% CalledStrike, 14.1 wCU, 2.38 wCU/C
Comments: You didn’t honestly think that we were going to get through a list of the best pitches in the game and not include Halladay, did you? Though 25 pitchers (min. 100 IP) flashed their curve more frequently than did the Doc, he still wound up second in the game in wCU and was tops among qualified pitchers in wCU/C. In addition, his 2.38 wCU/C was his best mark individually in the past five years. Halladay’s swinging-strike percentage was tops among starters, and seventh-best among any pitcher who threw at least 100 curveballs. He was no slouch when it came to called strikes either, as his percentage was nearly one standard deviation better than the mean. Halladay doesn’t generate a lot of vertical movement, so the pitch has more of a bottom-falling-out feel than a typical curve. But the result is the same as nearly all of Halladay’s offerings — domination.

Video
In what had to be a bittersweet game for Halladay, he allowed just one run, but one run was all that was necessary, as the Cardinals eliminated the Phillies from the playoffs on October 8th. You can see Halladay’s curve in action at 0:24 on Albert Pujols, 0:43 against David Freese, 1:17 against Nick Punto, and at 1:35 against Matt Holliday.

***

Craig Kimbrel

PITCHf/x: 30.3%, 86.8 mph, 4.8 X-move, -6.7 Y-move
Results: 23.99% SwStrk, 64.65% Strk, 20.45% CalledStrike, 10.9 wCU, 2.75 wCU/C
Comments: As Chris Cwik promised on Tuesday, it is here we find Kimbrel. Whether you want to call his breaking ball a curveball or a slider is less important than realizing just how dominant it is. For the record, the confusion lies in the velocity Kimbrel generates with the pitch. He pumps it with slider speed, but his grip is that of a spike curveball, and he has labeled it a curve in the past. So while Pitch f/x classifies it as a slider, we’re calling it a curve.

As both Lucas Apostoleris and Harry Pavlidis have noted this winter, Kimbrel is quite nasty. And while Kimbrel is a reliever, because he is a two-pitch reliever who threw a ton of innings last season, he actually threw more curveballs — 396 to be exact — than many starters. Of all pitchers who threw at least 100 curves, his 23.99% swinging-strike percentage was the best by nearly one and a half percent (David Hernandez was next best at 22.59%). It is so devastating because of the velocity — it’s easily the fastest hammer in baseball. And as Pavlidis explains in his post — and as most batters would surely attest — it’s really an unparalleled pitch in the game right now.

Video
This is against the Mets on September 8th. Jose Reyes leads off the ninth inning with a single, but Kimbrel then sets the side down in order. See him strike out Ruben Tejeda (:06) for the first out and Lucas Duda (:23) looking to end it.

***

Doug Fister

PITCHf/x: 14.9%, 75.2 mph, 4.0 X-move, -9.4 Y-move
Results: 9.96% SwStrk, 64.62% Strk, 33.69% CalledStrike, 4.4 wCU, 0.93 wCU/C
Comments: Just six pitchers had a called-strike percentage more than two standard deviations above the mean. Five of the six threw between 51 and 241 total curves, but Fister threw 472, nearly twice as many as the next-highest total, and still maintained a ridiculous called-strike percentage. The next highest percentage among pitchers who threw at least 400 curves was Shaun Marcum, who at 27.93% ranked 28th. Fister didn’t generate an above average number of swinging strikes, but with that many called strikes, he didn’t need to. His curve really moves — his -9.4 Y-move score was sixth-best among qualified pitchers last season. He also scored well in wCU and wCU/C, and overall was one of the game’s best pitchers last season, as his 5.6 WAR tied for eighth-best among pitchers. I could see the argument for Fister’s teammate Justin Verlander, and wouldn’t begrudge you that call, but since we recognized him in the fastball post, I figured we’d spread the love around.

Video
In his final start of the regular season, on September 26th, Fister pitched a gem against the Indians, and his curve made both Asdrubal Cabrera (:44) and Ezequiel Carrera (1:04) look downright silly.

***

Other Curveballs of note:

Wandy Rodriguez threw 221 more curveballs than any other pitcher, and with good reason — he has a nice yellow hammer. But Rodriguez’s value was down a little overall last year, and he didn’t grade out particularly well in either swinging- or called-strike percentage. Furthermore, his overall strike percentage of 51.72 was more than one standard deviation worse than league average. As a result, he will have to rest outside of the top three for this season.

Livan Hernandez has little left in his bag of tricks, but chief among said tricks is his curveball. Because he twirls it up there at an ungodly slow 66.9 mph — slowest in the Majors last year — the pitch has time to generate some spoon-bending movement, and as a result Hernandez’s 2.32 wCU/C was second only to Halladay.




Print This Post



Paul Swydan is the managing editor of The Hardball Times, a writer and editor for FanGraphs and a writer for the Boston Globe. Follow him on Twitter @Swydan.


19 Responses to “The Best Pitches of 2011: Curveball”

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

    Just pure filth from Kimbrel. Can’t imagine trying to hit that.

    Vote -1 Vote +1

  2. joe 1 says:

    does this mean i gotta draft fister and kimbrel in all my fantasy squads?

    Vote -1 Vote +1

  3. Patrick says:

    No Kershaw?

    Vote -1 Vote +1

    • Bip says:

      Kershaw so far hasn’t been able to command it well enough for it to be totally effective. As his command has gotten better, he’s moved on to the slider so much so that he threw the curve less than 6% of the time last year. I think these posts are focusing on players for whom the specified pitch is a critical part of their repertoire.

      Vote -1 Vote +1

  4. kenshin kawakami says:

    LOVE these articles!

    Vote -1 Vote +1

  5. I think the best curveball I ever saw was Nolan Ryans, 90 – 93 MPH. Most assumed he was nicking the ball to get it to break at that velocity. Maybe that’s what Robin Ventura’s real beef was, but given the outcome, who’s going to ask to see the ball again.

    Vote -1 Vote +1

  6. Bip says:

    As I said above, I wouldn’t argue that Kershaw should be included on this list due to decreased reliance on the curveball, but I still have to post this video. Knee-bucklers at 0:22 and 1:01 plus a swinging strikeout at 0:50 are great examples of one of the sexiest, if not most effectively used, curveballs in the game.

    http://mlb.mlb.com/video/play.jsp?content_id=19465469&c_id=mlb

    Vin Scully cracking up after 1:01 is just gravy.

    Vote -1 Vote +1

    • Albert Lyu says:

      That was extremely fun to watch. It’s pretty crazy to think that Kershaw isn’t relying on his curveball as much anymore mainly because his slider just also happens to be one of the best pitches of its kind in baseball. That and he hits mid-90s regularly.

      Vote -1 Vote +1

  7. LTG says:

    I have a question about how we ought to classify pitches. For the purposes of comparing the effectiveness of pitcher’s pitches we ought to classify on the basis of the category the batters would use to recognize a pitch successfully. (For example, Kimbrel’s pitch discussed here would be a slider.) But for the purposes of comparing how the pitcher achieves (or fails to achieve) his effectiveness, we ought to classify on the basis of the mechanics of the pitch. (For example, Kimbrel’s pitch would be a curve.) Which of these two approaches is more appropriate for the general statistical measure that w/Pitch represents?

    Vote -1 Vote +1

    • Albert Lyu says:

      To me, if we had perfect information, how the pitcher grips the ball is what the pitch classification should be… that’s just my opinion though. If I try to throw a curveball, maybe I can get a little bit of forward spin on the ball, but it’s not going to look like a curveball in PITCHf/x. I’ll still call that a curveball. Even if it’s at 50 mph.

      Vote -1 Vote +1

      • Kris says:

        Eh, I just go based on the outcome rather than the means. If it’s got the spin-rate and angle of a curve, it’s a curve.

        With that said, I’m sure this has been done before, but since you guys are going through all the data for these pitches…

        How abooooot we get some correlation coefficients (min. 100 pitches) for wPitch/c with x-movement, y-movement, velocity for each of the pitches. Because silly stuff like that is fun and takes you literally 20 seconds if you have the data in front of you. Pretty please!

        Vote -1 Vote +1

  8. Doc probably also has one of the best cutters?

    Vote -1 Vote +1

  9. Phantom Stranger says:

    Whatever Kimbrel throws, it isn’t a curveball. It moves much more like a slider because of his arm angle and general arm velocity.

    Vote -1 Vote +1

    • Albert Lyu says:

      Craig Kimbrel and Mark Bowman both call it a curveball. Yes, it looks like a slider if you just look at the velo, but it breaks downward much harder than most sliders at that velocity, leading me to believe that it’s a curveball. Harry Pavlidis and Lucas Apostoleris classify it as a curveball too at Brooks Baseball.

      Vote -1 Vote +1

    • Albert Lyu says:

      The spike grip is also evidence, I would say, that it’s thrown like a curve.

      Vote -1 Vote +1

    • AA says:

      Francisco Rodriguez says he doesn’t throw a slider, which most people have though was his out pitch for years. He throws 2 curves.

      Vote -1 Vote +1

    • Big Gerard says:

      Yeah, that clearly is not a curveball thrown by Kimbrel. You can call it what you like but if but if it is a curve it is a hanger! You notice nobody swung through it – it may as well have been a change. Curve implies vertical movement.

      Vote -1 Vote +1

  10. jerbear1985 says:

    Geinke is worth mentioning too, as he throws 2 different curves (one in the 55-66 mph range, one in the mid-70’s) and both are pretty filthy.

    Vote -1 Vote +1

  11. jerbear1985 says:

    I mean Greinke. R button fail.

    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>