Pitching by the Numbers: On the fly

The supposedly sabermetrically savvy are outsmarting themselves when it comes to fly ball pitchers.

Let's start with the king of all the new-age pitching stats – batting average on balls in play (BABIP, but if you pronounce that as one word, you are a dork). I like the stat, generally. Our pitching projections can be much sharper as a result of it. But it's not only less useful in projecting extreme fly ball pitchers, it's actually harmful.

The reason is that fly ball pitchers should be expected to have a much lower BABIP than league average for the simple reason that fly balls are significantly less likely to become hits than grounders. How less likely can be gleaned from this chart of the most extreme fly ball pitchers in baseball last year (minimum 20 starts):

Player

Team

GB/FB

AB

H

HR

K

SF

BABIP

Guillermo Moscoso

Oak

0.49

469

98

14

71

3

0.219

Jered Weaver

LAA

0.66

857

182

20

198

5

0.254

Josh Collmenter

Ari

0.70

530

129

16

89

2

0.266

Ted Lilly

LAD

0.71

723

172

28

158

5

0.268

Colby Lewis

Tex

0.71

767

187

35

169

5

0.270

Brandon Beachy

Atl

0.75

529

125

16

169

5

0.317

Jeremy Hellickson

TB

0.77

695

146

21

117

2

0.224

Scott Baker

Min

0.77

500

125

15

120

2

0.301

Bruce Chen

KC

0.79

589

152

18

97

5

0.283

Javier Vazquez

Fla

0.82

732

178

21

162

6

0.286

J.A. Happ

Hou

0.82

592

157

21

134

8

0.311

Brett Cecil

Tor

0.83

476

122

22

87

5

0.272

Michael Pineda

Sea

0.84

629

133

18

173

3

0.263

Tim Wakefield

Bos

0.87

545

148

22

83

7

0.286

Wade Davis

TB

0.88

712

190

23

105

7

0.286

Alexi Ogando

Tex

0.91

628

147

16

124

3

0.268

Jason Vargas

Sea

0.91

787

205

22

131

4

0.289

Shaun Marcum

Mil

0.92

753

175

22

158

6

0.267

Freddy Garcia

NYY

0.92

564

151

16

95

7

0.298

Tommy Hanson

Atl

0.93

484

106

17

142

3

0.274

Brandon Morrow

Tor

0.94

683

162

21

203

9

0.307

Randy Wolf

Mil

0.95

805

214

23

134

6

0.295

Bronson Arroyo

Cin

0.96

793

227

46

108

5

0.283

Justin Verlander

Det

0.99

904

174

24

250

3

0.238

Josh Tomlin

Cle

0.99

634

157

24

89

3

0.255

Ian Kennedy

Ari

0.99

818

186

19

198

9

0.278

Josh Beckett

Bos

1.00

693

146

21

175

5

0.252

Aaron Harang

SD

1.00

651

175

20

124

2

0.306

James McDonald

Pit

1.00

657

176

24

142

5

0.310

The BABIP expectation for this group should be .276 (their average) and even lower (.270) for the those lower than the average GB/FB ratio of 0.87 (it's .280 for those with a higher rate, but still 1.0 or better).

But yet all these pitchers are generally recalculated to have the generic, league-average BABIP and resulting expected ERA. (I do realize that fly balls that are hits are about 10 times more likely to result in extra bases, so let's table xFIP for now.)

I promise you that I am not representing Jeremy Hellickson. I had no shares, at least not until doing this research (after which I acquired him for Rickie Weeks in a dynasty league). But I understand your suspicions after last week's column on missed swing rates. Commenters noted how that piece didn't address his lucky BABIP. (Why would it?) Well, here you go. Hellickson wasn't so lucky after all – especially when you factor in that 16.2% of his fly balls were infield pop-ups (second highest rate in the league, behind only Ted Lilly). That's about 20 more automatic outs than we could expect – about as good as Ks (versus the 16 or so outs we should have expected with regular old fly balls).

Another key note about extreme fly ball pitchers: Their expected rate of homers allowed as a percentage of fly balls allowed is typically better than the league average rate, too. Perhaps I'll expand on this in a future column. But you need to know it right now if you are drafting because those xFIP ERAs recalculate homers and unfairly inflate them for many of these types, I do believe. Think about it logically – the fly ball pitcher is exerting control over the at bat when the hitter hits a fly ball. He's basically won. When a ground ball pitcher allows a fly ball, something by definition has gone wrong and the hitter is presumably significantly more likely to make solid contact.

So discount the reported "luck" factor for the pitchers above. Many of your owners and most stat heads will be undervaluing them. Don't be like them. Also, circle the guys who had a high BABIP despite their GB/FB suggests. They were likely far more unlucky than it first seems (though check their respective line drive rates, too). And, yes, by implication ground ball pitchers are unfairly rewarded by the luck adjusters. But more on that later, too.

Michael Salfino (Twitter @MichaelSalfino) is a quantative sports analyst whose writing regularly appears in the Wall Street Journal. His New York sports musings can also be found at SNYWhyGuys.

What to Read Next