ROTOWIRE.COM NFL MLB / Minors NBA NHL GOLF RACING    COMMUNITY FORUMS PODCASTS MYROTOWIRE ASK AN EXPERT GAMES

Is Pitcher BABIP Luck?

There's a school of thought that pitchers primarily control strikeouts, walks and fly balls allowed, but don't much control what happens once the ball is put into play, i.e., whether a batted ball finds a fielder or an empty space. Sure there are exceptions like knuckleballers who typically induce weaker contact, but for the most part - the theory goes - batting average on balls in play allowed is largely random for pitchers. As a result, it's should be a safe bet that pitchers like Felipe Paulino (.341 BABIP in 2011) , Ricky Nolasco (.331) and Edwin Jackson (.330) should all normalize toward the league average (.295) in 2012 - or at least their teams' average, assuming their defense is roughly equal to last year's. (The Marlins, for example had a team BABIP allowed of .297). But is it?

Assuming non-knuckleballers don't have a lot of say in their BABIP, the pitchers with the best and worst career BABIPs should be roughly equal, with the best BABIP's having a slight advantage simply due to their good luck on balls in play. But over an entire career, that luck (a) shouldn't be too extreme because the sample is so large, and (b) fluctuates so much over time.

There are 176 pitchers who have thrown at least 1000 IP since 1995. Let's take a look at the best-50 and worst-50 by BABIP:

Best 50 BABIP IP BABIP Worst 50 BABIP IP BABIP
Mariano Rivera 1211.1 0.262 Glendon Rusch 1477.1 0.326
Matt Cain 1317.1 0.265 Zach Duke 1041 0.323
Barry Zito 2252 0.268 Paul Quantrill 1015.1 0.319
Orlando Hernandez 1314.2 0.268 John Burkett 1651 0.316
Ted Lilly 1911 0.27 Aaron Sele 1898 0.315
Tim Wakefield 3006 0.273 Charles Nagy 1227.2 0.314
Jarrod Washburn 1863.2 0.273 Shane Reynolds 1631.1 0.312
Jeremy Guthrie 1020.1 0.273 Jimmy Haynes 1200.2 0.312
Ryan Franklin 1201 0.273 Edwin Jackson 1079 0.311
Johan Santana 1908.2 0.275 Jeff Fassero 1604.1 0.31
Scott Elarton 1065.1 0.275 Jeff Francis 1065.2 0.31
Carlos Zambrano 1826.2 0.276 Carlos Silva 1241.2 0.31
Jered Weaver 1131.2 0.276 Paul Maholm 1143.2 0.31
Woody Williams 2120 0.276 Jaime Navarro 1012.1 0.31
Jamie Moyer 3019.1 0.277 Andy Pettitte 3055.1 0.309
Kerry Wood 1371.1 0.278 John Lackey 1876 0.309
Rick Helling 1474.1 0.278 Sidney Ponson 1760.1 0.309
Pedro Martinez 2567.2 0.279 Brian Moehler 1567.1 0.309
Tim Hudson 2503.1 0.279 Jason Jennings 1128.1 0.309
Bruce Chen 1164.2 0.279 Esteban Loaiza 2099 0.308
Cole Hamels 1161.1 0.28 Zack Greinke 1279.2 0.308
Tom Glavine 2891 0.281 Aaron Harang 1622.1 0.308
Al Leiter 2052 0.281 LaTroy Hawkins 1261.2 0.308
Rick Reed 1296.1 0.281 Julian Tavarez 1365.2 0.308
Bronson Arroyo 1874.1 0.282 Mark Hendrickson 1169 0.308
Freddy Garcia 2076.1 0.283 Doug Davis 1715.2 0.307
Denny Neagle 1565.2 0.283 Aaron Cook 1312.1 0.307
Eric Milton 1582.1 0.283 Livan Hernandez 3121.2 0.306
Greg Maddux 3097.1 0.284 Chuck Finley 1564 0.306
Hideo Nomo 1976.1 0.284 Pedro Astacio 1779.1 0.306
Jon Garland 2083.1 0.284 Jeremy Bonderman 1176 0.306
Pat Hentgen 1626.2 0.284 Scott Karl 1002 0.306
Ismael Valdez 1799 0.284 Scott Erickson 1469 0.305
Wilson Alvarez 1221.2 0.284 Joey Hamilton 1232 0.305
Dustin Hermanson 1283 0.284 Nate Robertson 1152.1 0.304
Kirk Rueter 1740 0.284 Ryan Dempster 2042.2 0.303
Brian Anderson 1434 0.284 Darren Oliver 1756.2 0.303
Justin Verlander 1315.1 0.285 Shawn Estes 1678.1 0.303
Randy Wolf 2110.1 0.285 Mark Redman 1238.2 0.303
Kevin Brown 1977.1 0.286 Dontrelle Willis 1221.2 0.303
Brandon Webb 1319.2 0.286 Kelvim Escobar 1507 0.302
Jake Peavy 1581.1 0.286 Carl Pavano 1725.2 0.302
Steve Trachsel 2335.1 0.286 John Thomson 1270.1 0.302
Paul Byrd 1697 0.286 Kyle Lohse 1762 0.302
Brett Tomko 1816 0.286 Chad Billingsley 1013.2 0.302
Russ Ortiz 1661.1 0.286 Scott Kazmir 1022 0.302
Ron Villone 1168 0.286 Pat Rapp 1150 0.302
Chan Ho Park 1989 0.287 Jon Lieber 2089.1 0.301
Andy Ashby 1444.1 0.287 Brad Penny 1871 0.301
Andy Benes 1389 0.287 Jeff Weaver 1838 0.301

As you can see, there's arguably only one (marginal) Hall of Famer (Andy Pettitte) among the bottom 50, and one current star (Zack Greinke). Among the top 50 are the following: Mariano Rivera, Matt Cain, Johan Santana, Jered Weaver, Tim Hudson, Pedro Martinez, Cole Hamels, Tom Glavine, Greg Maddux, Justin Verlander, Kevin Brown, Brandon Webb and Jake Peavy.

Maybe I'm underestimating how much BABIP "luck" spares pitcher arms, giving them better health and more confidence, but even so, that would be reason enough to take it seriously, even if it were pure luck initially. More likely, though, while BABIP - like batting average- fluctuates greatly week to week and even year to year, for the most part it's a greatly associated with more skillful pitchers.

That's not the same as saying Nolasco, Paulino or Jackson can't bounce back - or can't as Pettitte and Greinke (so far) have overcome bad luck/inability to induce weak contact - or that *some* of BABIP can be explained by luck. Just that a poor BABIP very likely has a repeatable skills component to it and that you should be wary targeting outliers for bounce-backs without some other compelling reason.

Comments

By: Scott Pianowski
On: 3/8/2012 2:54:00 PM		 I completely agree with your premise. But it's more fun to play in roto leagues where otherwise-smart people drink from the cup of "this one amazing stat explains everything!"

Batters steer batted-ball outcomes a lot more than pitchers, but the pitcher isn't completely irrelevant in the discussion. It's a shame Mike Fast left BP for the Astros, because he was doing some amazing research before he departed. Now it belongs to the Astros, and it's private.
 
By: Scott Pianowski
On: 3/8/2012 2:59:00 PM		 It's part of a longer thread so it's coming out of context, but here's a great point Mike Salfino made on Patton & Co:

It's so obvious that all batted balls are not the same when it comes to measuring hit probability but we treat them all the same and thus flyballers are generally undervalued by the sharps who outsmart themselves with FIP and BABIP, while groundballers are overvalued.
 
By: Zenguerrilla
On: 3/8/2012 3:08:00 PM		 It doesn't take into account quality of defense and the ballpark factor. It's a nice guide but not a core stat.
 
By: lvtdude
On: 3/8/2012 4:25:00 PM		 A guy with good late movement like Mariano Rivera has to be better at keeping hitters from squaring the ball up, even when they are lucky enough to hit it. The BABIP guys that ignore this are, well, ignorant.
 
By: kroyte
On: 3/8/2012 5:43:00 PM		 The Mike Fast story that offered convincing proof that pitchers did exert some control over how the ball was hit can be found here: http://www.baseballprospectus.com/article.php?articleid=15532#commentMessage.

But Chris, I think you're underplaying the effect of contextual factors on a pitcher's stats, which is what we're interested in when putting together our fantasy teams. McCracken's observation was that when you take fielding and park effects out of the equation it turns out pitchers don't control much of the outcome of what happens to batted balls. His calculations for DIPS are fantastically complex, and try to neutralize out every bit of context except for the pitcher's skills. What he found when he completed this exercise was that pitchers' DIPS ERA swing wildly from year to year, from positive to negative and back again, indicating a certain randomness, which he then sold as a pitcher having no control over balls in play.

But Fast's work shows that there is a basis for finding there is some control, which is important. But the amount of control is overshadowed by park effects, opponents, and the defense behind him. So while a pitcher has some control, and some are better on balls in play than others, the results swing wildly because of outside factors.

The reason not to put undue faith in BABIP is because a pitcher on the same team is still dealing with the same defense and ballpark and opposition, for the most part, as well as his skills, which also explains some very real percentage of the divergence in your charts above.
 
By: Chris Liss
On: 3/8/2012 5:55:00 PM		 Kroyte - I agree the contextual factors are huge, but as you point out (1) They usually change only gradually from one year to the next (for short-term fantasy purposes) and (2) These are 1,000 IP samples over which they probably changed drastically. That's not to say some pitchers on this list didn't have advantages or disadvantages based on context, but don't you think those are far smaller than the disparity in talent?

Also as for BABIP (or DIPS ERA) swinging wildly from year to year, isn't that a bit like any other small sample stat fluctuating? But when you look at it over 1000 IP, the correlation between skill and preventing hits seems pretty strong over the last 17 seasons.

 
By: kroyte
On: 3/8/2012 6:20:00 PM		 I don't know the answer about weighting the skill versus the contextual factors. Fast says that pitcher control over contact is about half as powerful as hitter control, which gives us a clue, especially when we look at the range in defensive efficiency across a league.

As for No. 2, you looked at 176 pitchers with more than 1,000 innings. For some of those 176, some years with a great defense will be cancelled out by some years with an awful defense, and they'll end up in the middle group you don't show. It is persuasive that the better than expected list contains better pitchers, but what if it turns out those better pitchers are considered better because they had more better defenses behind them most of the time?

I don't want to make a big deal about this, because I mostly agree with you about BABIP, but I think it's important to emphasize that the way your lists are compiled makes it very difficult to judge how much of the differences above are based on pitcher skill. What they do show is that better pitchers do better, for whatever reason, which is what we really need to know. (Also, it should be noted that Fly Ball pitchers have a significantly lower BABIP than Ground Ball pitchers, which is another factor in all this!)

Don't mean to be a grouch.
 
By: Chris Liss
On: 3/8/2012 8:57:00 PM		 Appreciate the debate Kroyte, so no worries about that. I'd just say that when you look at how many greats are on the low BABIP list vs. the high one, it's hard to ascribe that to park factors and defense over 1000 IP. In other words, even if we assume that Maddux and Pedro had advantages in those areas, and that helped them a little, they'd still have been great even if they didn't. So you'd expect some Pedros or Maddux's to show up on the unlucky list. But they don't. Granted, Clemens and Randy Johnson don't show up on either list, but why are almost no stars on the unlucky list unless it's rare for greats to be below average at inducing weak contact.
 
By: kroyte
On: 3/8/2012 9:24:00 PM		 I think when you look at how many guys who tilt Fly Ball are on the low BABIP list, you have your answer. It's a systemic issue. Well, I mean to say part of it is systemic. Fly Ball pitchers tend to have a lower BABIP. That's the way it works.

They allow more runs, however, because they allow more home runs. But not so many more that it offsets the prejudice against them. So a fly ball pitcher may not be a bad thing. In fact he is often a good thing when people are overvaluing ground balls.

It makes me wonder how someone like Tim Hudson got on that list. He's swimming up the river on that one.
 
By: Chris Liss
On: 3/8/2012 10:08:00 PM		 But Rivera, Hudson, Greg Maddux, Brandon Webb, Tom Glavine, Zambrano, Kevin Brown were all ground-ballers.
 
By: ScoresheetOwner
On: 3/9/2012 10:14:00 AM		 Chris, nice stuff. Though I wonder if Maddux and Pedro had been on teh unlucky list if we would have considered them greats? After all, if they'd been unlucky then the things that made them great (low ERAs and/or high win totals) would have not been nearly as good.
 
By: kroyte
On: 3/9/2012 10:17:00 AM		 True. But I looked at the top 20 on the overachievers list and got this:

GB: Rivera, Zambrano, Hudson, Hamels
Neutral: Guthrie, Franklin
FB: Cain, ZIto, Hernandez, Lilly, Washburn, Santana, Elarton, Weaver, Williams, Moyer, Helling, Wood, Martinez, Chen.
Knuckleballter: Wakefield

Let's look at the top 20 on the underachievers:

GB: all of them except
FB: Esteban Loaiza
 
By: Chris Liss
On: 3/9/2012 12:46:00 PM		 It's true FB skew toward low BABIP and GB toward high, but that just means you could separate them into two camps - GB top-20 and bottom 20 luck, and FB top-20 and bottom 20 luck. You'd still find that the "lucky" pitchers were much better in both camps, and better beyond the boost they got from their supposed luck.
 
By: gooklaw
On: 3/9/2012 12:59:00 PM		 Maybe this all comes out to: "It's good to be lucky and it's lucky to be good."
 
By: Chris Liss
On: 3/10/2012 1:43:00 PM		 Scoresheetowner - sorry forgot to respond but I'd say Maddux and Pedro were good enough to have overcome a little bad luck, so you'd expect at least a couple great pitchers to be on the unlucky list, but the best we could do was Pettitte and Greinke.
 

Leave a comment

Commenting is restricted to registered users only. Please register or login now to submit a comment.

Tell Someone

  • Digg it
  • submit to reddit reddit
  • Add to Mixx!

Recent Favorites

Interesting Spring Performances, Part One: The NL
As spring training ambles nonchalantly towards the finish line, it's always fun to take a look at spring stats to see who really jumps off the page and might be making some kind of statement or other...
My AL Tout Wars team
We did the AL Tout Wars draft Saturday morning, once again in "The Bubble" at the Sirius/XM studios. It was a different Tout experience than my usual drafts at Tout - lately I haven't paid for closers there, or spent much on my outfielders. But I'm also one to take the values where they fall. So to that end, here's my squad:
My NFBC Recap
I have never been one to hold up to peer pressure well. Since Peter, Jeff, and Chris did a NFBC challenge, I had to join the cool kids and do one too.
SiriusXM Host League
My team:
Beat Peter Schoenke - Draft Results.
My "Beat Peter Schoenke" Draft Results.

My entry in the RotoWire Fantasy Baseball Online Championship was Monday night. Here's my quick take on the draft - this is a 12-team mixed league, 5x5 categories, 30 rosters spots (23 active), with no trades in the league. It has a top prize of $50,000 and a league prize of $1,400.

RSS Feeds