keyboard_arrow_uptop

Several articles ago, I took a look at the possibility that knuckleball pitchers might exert some “hangover” effect on opposing batters extending beyond their own appearances on the mound. The notion that facing the knuckleball might possess some potential to “mess up” a batter’s swing has long been a part of baseball lore, causing players and managers alike to ascribe post-knuckleball slumps to the mysterious force of the flutterball. But could the rumors of a knuckleball hangover hold up on a baseball edition of “Mythbusters”?

In part one of this two-part series, my cursory initial review of performance following exposure to Tim Wakefield seemed to reveal some sort of lingering penalty, but I mentioned that I’d like to subject the knuckleball hangover to a more rigorous investigation, controlling for additional factors in an attempt to determine whether the effect was real. Readers expressed interest in a follow-up, and with the help of our newest full-timer, Colin Wyers, I’ve managed to settle the matter (for the most part).

Here’s what I found last time for the performance of opposing batters in games following starts by Wakefield:

Time Period

PA

AVG

OBP

SLG

OPS

SOr

UBBr

HRr

1 Day After

11105

.265

.332

.416

.748

.168

.080

.025

2 Days After

10460

.271

.338

.429

.767

.168

.078

.028

3 Days After

10274

.280

.345

.435

.781

.163

.078

.030

Seasonal

 

.274

.344

.440

.784

.159

.083

.030

With the old methodology, there seemed to be something to the notion that Wakefield’s floater impaired opposing batters for at least a day or two after his starts. However, as we’re about to see, our new methodology isn’t quite so sanguine about his ability to reach beyond the box score in an effort to suppress offense. This time around, we used “newTAv instead of OPS, which allowed for a more accurate assessment of offensive performance. In addition, we controlled for the prowess of the pitchers following Wakefield, as well as each batter’s seasonal performance, using the Log5 method, which gave us an expected TAv for each batter on each day removed from a confrontation with Wakefield.

Time Period

Expected

Actual

Same Game

.261

.260

1 Game After

.264

.258

2 Games After

.265

.270

3 Games After

.268

.267

That “Same Day” figure includes both plate appearances against Wakefield and plate appearances against the relievers who followed Wakefield in his starts; isolating the performance against the relievers alone would be ideal, but were a sizeable effect out there, it would likely still show up in the cumulative figure. Batters did perform slightly worse than expected in their first game after facing Wakefield, but exhibited no handicap in either of the following two games. If there is an effect, it doesn’t seem to be one of a magnitude large enough for the Red Sox to go out of their way to exploit it, which they could consider doing by manipulating their rotation in order to start Wakefield more often in the first game of a series, or choosing to pitch him in the first game of a doubleheader wherever possible (neither of which they appear to have done with any regularity over the course of his career).

“New” TAv has not yet been park-adjusted, so it’s possible, if unlikely, that a hangover effect associated with Wakefield’s knuckler could have been disguised above by the fact that a substantial portion of the games following his starts took place in Fenway Park, a favorable hitter’s environment. With the aid of The Neyer/James Guide to Pitchers and several other sources, I compiled a list of starters who featured the knuckleball and were active at any point from 1972 on (the period during which Retrosheet’s records are complete enough for a study of this nature). In cases where I could establish that a pitcher adopted the knuckleball mid-career, any data from before the pitch became a critical component of the converted knuckleballer’s repertoire was dismissed. That left me with the following list:

Name

Usage Notes

Tom Candiotti

Post-1984

Bruce Dal Canton

 

R.A. Dickey

 

Eddie Fisher

 

Charlie Hough

 

Ken Johnson

 

Joe Niekro

Post-1971

Phil Niekro

 

Steve Sparks

 

Dennis Springer

 

Tim Wakefield

 

Wilbur Wood

 

So, what happens when we run the numbers for that list of 12 knuckleballers, a sample consisting of almost 32,000 data points?

Time Period

Expected

Actual

Same Game

.257

.259

1 Game After

.263

.263

2 Games After

.262

.266

3 Games After

.262

.267

 Those may not be the exciting findings we were hoping to generate, but the results seem fairly conclusive: batters need not consider consuming a draft of knuckleball-flavored hair of the dog after facing the flutterball. In short, it’s possible that the knuckleball hangover effect exists for certain batters, or even for certain pitchers, but the evidence suggests that it’s neither a major factor, nor an inherent trait of knuckleballs on the whole. The knuckleball may be a nuisance to hit, but its widespread reputation for fouling up swings is likely the result of confirmation bias and selective memory. Teams may not be pleased to see an opposing knuckleballer on the mound, but blaming an entire road trip’s worth of offensive ineptitude on one’s offerings seems like quite a stretch.  

You need to be logged in to comment. Login or Subscribe
Michael
8/19
Not that it would budge the data, but I believe that R.A. Dickey began primarily relying on the knuckleball from 2006 onward. He had a more traditional pitch selection before that.
ckahrl
8/19
It's interesting to me that this confirms the results of a similar, albeit much less robust study we did 15 years ago. Good work, Ben!
flalaw
8/19
Was any consideration given to the quality of the pitchers that followed the knuckleballers in the rotation, or was that simply considered to be a wash over the years?
bornyank1
8/19
Yes, this analysis accounted for the quality of the pitchers who followed each knuckleballer.
tbwhite
8/19
What about breaking it down by the quality of the knuckleballer's start ? Perhaps facing a knuckleballer who really has it working messes up batters, while facing a struggling knuckleballer is nothing but batting practice.
bornyank1
8/21
That's possible. Maybe I'll get around to a Part Three someday.
StatFreak101
8/20
Awesome stuff. This is something that I have always been curious about, and this series of articles provides an answer.
Mountainhawk
8/21
This might be asking too much, but do these stats include only the hitters that faced the knuckleballer, or the whole team?
bornyank1
8/21
They include only the hitters who faced the knuckleballers, not the entire teams.