Jun 18

In Search of Uncle Charlie

Sometimes I wonder if the days of the curveball have passed.  It seems like a pitched rooted in nostalgia, as if the great ones can only be witnessed on grainy celluloid made famous by Ken Burns.  When I think of monster hooks I envision Sandy Koufax, Bert Blyleven, and Dwight Gooden.  Big leg kicks, electric arms, and an Uncle Charlie that buckles batters’ knees in pure fright.  Pedro Martinez was a curveball artist.  So too was Darryl Kile who threw a 12/6 break so nasty that I blew out my elbow trying to emulate it.

The curve still exists of course.  It’s not as though pitchers have forgotten how to throw it.  But when I think of pitchers who defy the laws of physics with the pitch, my mind doesn’t conjure up too many images.  I think back to years ago, and this fact brings me great shame.  So, in an effort to remedy my own ignorance, I’m going to spend a few words on which pitchers have the best curves going today.

What does “best curve” even mean, though?  Is it the pitch with the biggest, sharpest break; the most likely to induce a swing and miss; the pitch most likely to send batters scurrying for cover; or the most effective, saving the most runs?  For the purposes of this discussion, I opted to go with saving the most runs.  While thanks to PITCHf/x we can determine all of the above a pitch is only as useful as the end result.  Orlando “El Duque” Hernandez threw one hell of a deuce dubbed Eephus, but a looping, memorable curve that’s crushed for a home run doesn’t wipe runs off the board for style.

That being said, I focused in on the standardized Pitch Values as shown on FanGraphs.  While wCB is a cumulative value of runs saved with the use of the curve, wCB/C standardizes this figure per 100 pitches, allowing a more accurate measure for comparison since differing pitchers rely upon the curve more than others.  I neither made a distinction between starter and reliever nor filtered out those that didn’t have enough innings to qualify.  Just because a pitcher doesn’t meet the MLB standard for innings started doesn’t mean he can’t qualify for my distinguished list.

Here, then, are the top 10 pitchers in regards to wCB/C:

Player Name wCB/C wCB Thrown
Matt Albers 20.36 0.2 5
Evan Scribner 12.31 0.5 4
Mat Latos 8.63 0.6 9
Jeff Beliveau 6.2 0.1 2
Brandon Morrow 5.37 1.1
Junichi Tazawa 4.81 1.7 120
Joe Kelly 4.57 2.1 39
Lucas Luetge 4.51 0.2 4
Bruce Billings 4.34 0.1
Mike Dunn 4.1 0.2 7

Pitchers Ranked by wCB/C

Well, I think we can spot the obvious problem with simply ranking the pitchers by standardized Pitch Values.  Matt Albers has thrown only five curves, and according to wCB/C those might be the best curves ever.

Is there a way to reward quality, consistency, and volume?  You bet there is.  Using data collected from baseballsavant.com, I took the percentage used (how many times a pitcher throws his curve relative to his other pitches) and multiplied that value by wCB/C.  It’s rudimentary, akin to opening a Coke bottle with a rock, but it gets the job done in this case and allows me to try and determine who throws a curve well and often.

And here’s our updated top 10 list:

Player wCB/C wCB Thrown % Thrown Score
Vic Black 2.64 1.7 66 33.52 88.49
Felix Hernandez 1.99 4.2 213 13.80 27.45
Sonny Gray 2.23 8.5 380 26.64 59.42
A.J. Ramos 2.88 0.3 121 20.45 58.89
Jesse Hahn 1.94 1 50 30.12 58.43
Zach Duke 2.15 2 112 25.85 55.58
Corey Kluber 3.39 8 244 16.28 55.19
Kevin Quackenbush 2.93 1.3 46 18.18 53.27
Jered Weaver 2.78 7.1 289 18.98 52.76
Joba Chamberlain 2.58 2.4 97 19.91 51.37

wCB/C Adjusted for Percentage Used

Joe Kelly was actually second on my recalculated list, but since he’s only started three games this year due to injury, I replaced him in the chart with Felix Hernandez.  Was Hernandez 12th?  Nope.  Actually, he was all the way down at 32, but I just wanted to see how he compared to the others.  There’s a pretty fair distribution in this table.  There are starters (Gray, Weaver, Kluber, Hahn, Duke, and Hernandez) and relievers (Chamberlain, Ramos, Quackenbush, and Black).  Sure, that’s all well and good, but how exactly are these guys getting it done?  We can look to the event breakdowns for each pitcher to see what the end results are for their curves.

Player Hit % In Play % Ball % Called Strike % Foul % Whiff%
Vic Black 1.5 3.0 47.0 19.7 18.2 10.6
Felix Hernandez 2.8 9.9 31.9 28.6 13.1 13.6
Sonny Gray 5.3 11.8 33.7 18.2 16.6 14.5
A.J. Ramos 1.7 8.3 40.5 20.7 14.9 14.0
Jesse Hahn 4.0 10.0 32.0 24.0 16.0 14.0
Zach Duke 2.7 5.4 41.1 25.0 14.3 11.6
Corey Kluber 3.7 12.3 29.9 20.1 20.9 13.1
Kevin Quackenbush 2.2 8.7 34.8 30.4 6.5 17.4
Jered Weaver 4.5 15.6 34.3 21.5 13.1 11.1
Joba Chamberlain 1.0 9.3 44.3 23.7 9.3 12.4

Curveball by Events

If Pitch Events tells us anything it’s that a called strike on a curveball isn’t exactly the desired outcome.  Only Quakenbush touches 30% on called strikes, with Felix Hernandez at 28.6, and every pitcher is 30% or greater (okay, Kluber is 29.9) for Ball %.  This shows that the curve is a difficult pitch to control, but it also shows that a pitcher’s desired intent is to get the batter to chase.  Vic Black takes this to the extreme, throwing a ball nearly ½ the time, but batters only put the ball in play 3% of the time.

Player K% Avg OBP SLG ISO wRC+
Vic Black 76.9 .077 .077 .077 .000 -54
Felix Hernandez 17.6 .182 .206 .273 .091 44
Sonny Gray 38.1 .185 .221 .194 .009 28
A.J. Ramos 34.6 .095 .269 .095 .056 33
Jesse Hahn 50.0 .133 .188 .333 .200 58
Zach Duke 62.5 .125 .125 .125 .000 -26
Corey Kluber 54.8 .101 .130 .135 .034 -19
Kevin Quackenbush 28.6 .143 .143 .143 .000 -15
Jered Weaver 18.8 .178 .250 .233 .055 52
Joba Chamberlain 31.3 .067 .125 .200 .133 -3

Curveball by Outcome

Remember how batters only put the ball in play off of Black 3% of the time.  Look at that batting line against his curve:  .077/.077/.077.  Batters don’t really know what to do with the pitch, and so far it shows.  Also, take a look at the ISO on Gray and Kluber:  .009 and .034 respectively.  For two guys who’ve thrown a combined 624 curves to put up lines like that is ridiculous.  For a starter to throw his curve nearly 27% of the time as Gray does and batters hit .185/.221/.194 off of it is impressive.  Then there’s Kluber.  One reason he’s having such a great year is that batters have gone .101/.130/.135 off of his curve, and he’s the only starter with a negative wRC+ (-19).  I might have found my Jedi Master.

Just to expand upon this idea some more, but not to bury the article with images and charts, I’d just like to take a moment to see how Black and Kluber approach batters with the curve.

Vic Black_righties


Vic Black Curve vs Righties

Black tends to go low, outside corner more to righties than any other zone, perhaps hoping for them to chase.  He has come inside enough that you wouldn’t call it a complete avoidance of the inner half though.

Vic Black_lefties

Vic Black Curve vs Lefties

This is absolutely unlike how Black approaches righties with his curve.  Black keeps to the inner half, inside corner at the knees.

Corey Kluber_righties

Corey Kluber Curve vs Righties

Kluber avoids righties by going outside corner primarily, but it’s safe to say he enjoys baiting right-handers on the outer third.

Corey Kluber_lefties

Corey Kluber’s Curve vs Lefties

Kluber comes middle-in enough to lefties to not make trends too obvious, but from the chart, it appears that Kluber tries to get the lefties to chase, changing levels and zones on the outside fairly frequently.

While this didn’t necessarily answer the question which pitcher has the most dominant curve in baseball, from a purely aesthetic value, it does attempt to answer whose curve, reliever or starter, saves the most runs and is hardest to produce offense against.  The best part is that now I can appreciate a Mets reliever for . . . anything.

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=""> <s> <strike> <strong>