BABIP BS – Part II
	by Steve Fonzo Wayne August 10, 2008

 

Back in May, I wrote a column BABIP BS, which was an attempt to demonstrate that the wrongful use of Batting Average on Balls In Play (BABIP) is a bunch of BS. Apparently, I struck a major nerve (more like a Carotid Artery or Spinal Cord) with some in the fantasy industry. The article may also have been taken personally, as a writer prepared an in-depth criticism (link here) which was littered with a litany of attacks, insults, and sarcasm, many of which were expressed with exclamation points.

 

In what was probably the writer’s fantasy sports version of “Revenge of the Nerds”, it seemed like someone was offended, went and cried to momma, and sicked Fantasy Sergeant Hulka on me, or perhaps even Weird Science’s big bother Chet: “Do you realize it's snowing in my room goddammit!” Armed with a 22-caliber scientific calculator, pocket thesaurus, and a Maxwell Smart spy pen, anybody who uses the words “vitriol” and “metric” in a baseball article likely didn’t get invited to play pickle in the neighborhood.

 

Now that I got that off my chest, let’s get on with the important stuff, the BABIP BS – Part 2.

 

The Rebuttal

 

The easiest way to respond to the critic’s remarks is to cite his exact comments in the same fashion he presented. Please keep in mind this important notion - this rebuttal is not intended to defame, ridicule, or prove the critic wrong, but rather to help you and the critic better understand my points about the fallacy of BABIP and it’s ineffectiveness as a predictive indicator of future player performance.. One main problem with his criticism is that he conveniently cited bits and pieces of my article without accounting for the statistics and back-up support that I provided.

 

Clearly, any analyst can select and twist statistics and counter-statistics that better serve their purpose, so we could go round and round forever without coming to a logical and reasonable agreement - but you cannot deny basic, grade-school level mathematics, and many of my important conclusions were base on 1+1=2, so it is clear that the critic misunderstood some premises, as I’m pretty confident that he can count to 10.

 

The BABIP Formula

 

BABIP = (H - HR) / (BFP-HR-BB-SO-HBP)

 

I have also seen the formula expressed as (H – HR) / (AB – SO – HR + SF), which yields the same result. Please remember that walks (intentional and non-intentional) and hit by pitch (HBP) do not count as an at-bat, and therefore cannot be defended, thus omitted from the equation (other types of plate appearances also cannot be defended, but walks are important, as I will get too later).

 

Quoted My Work: “Probably the biggest concern is that there is no direct correlation between a pitcher’s fantasy-relevant stats (ERA/WHIP) and their BABIP, so it serves no purpose unless fantasy leagues use BABIP as a scoring category.”

 

Critic’s Response: “This was the author’s 1st concern about the application of BABIP for fantasy purposes. He is completely misunderstanding the use of the metric if this is truly how he feels. This is very strange since right above he actually defined the term, explained the theory, and even discussed Voros McCracken’s famous work on DIPS. As we’re all well aware, the use of BABIP is to determine which players are most likely to see their ERAs and WHIPs rise or fall depending on whether their BABIP is lower or higher than the league average. And of course there is a correlation between a pitcher’s BABIP and their ERA/WHIP! A pitcher with a low BABIP is obviously more likely to have a lower ERA and WHIP than they “should”, while a high BABIP would signify a higher ERA and WHIP than they “should”. It’s equivalent to saying that GB% serves no purpose to analyze unless fantasy leagues use it as a scoring category.”

 

I’m not sure what is so strange. Yes, I defined the BABIP theory and how contemporary fantasy analysts use it to evaluate future performance in terms of ERA and WHIP, but I also continued to describe its intended purpose by the inventor (which I explained is different than the way contemporary fantasy analysts perceive it’s theory and use). So, I stated why and how it is commonly and currently used, as the critic understood, but it should have been clear that I disagreed with the contemporary fantasy analysts’ own twisted theory and their use of it. Let’s make my opinion clear, so that even Fantasy Sergeant Hulka can understand what I am saying:

 

1. The inventor of the Defense Independent Pitching Statistics theory used a sub-stat which had been identified and defined by contemporary statisticians (and fantasy analysts) as “BABIP,” and the inventor’s  basic premise was that a pitchers DIPS stats, of which BABIP is an element of, created a new stat line which “removes virtually any effect the defense could possibly have on the pitchers stats.” He also pointed out that he was not drawing the assumption that pitchers can't have any effect on singles, doubles, triples etc., but he was “simply saying that the defense CAN affect these things and we're often uncertain to what extent these effects are.” That is it! He made no assertion (or any hints) that a pitcher’s or hitter’s ability to control what happens to a ball hit in play (e.g., a hit or an out) is mostly random and a result of luck, or that a pitcher’s BABIP will eventually gravitate towards his career BABIP or the league average – which brings me to #2.

 

2. It is many contemporary statisticians and fantasy analysts that took the inventor’s DIPS research and made leaps of their own and extrapolated an expanded theory which correlates BABIP and “luck,” and the resulting belief which the critic states “the use of BABIP is to determine which players are most likely to see their ERAs and WHIPs rise or fall depending on whether their BABIP is lower or higher than the league average.” That is the belief of many contemporary fantasy analysts which I disagree with, and they therefore improperly use the BABIP stat for purposes other than it was intended – to evaluate a players performance as lucky or unlucky and make predictions about his future performance.

 

Contrary to my statement, the critic emphatically stated “there is a correlation between a pitcher’s BABIP and their ERA/WHIP!”  He further stated that “a pitcher with a low BABIP is obviously more likely to have a lower ERA and WHIP than they “should”, while a high BABIP would signify a higher ERA and WHIP than they “should”. The critic conveniently left out my statement and supporting stats that demonstrate otherwise. For example, immediately following my first concern, I stated “Further down in this article, I show that, on average, pitchers with low BABIPs, which you would think is a GOOD thing, actually have higher ERA's and lower percentages of quality starts, while pitcher's with a high BABIP (a bad thing) had lower ERA's and better percentages of quality starts.” I backed up that statement with actual three-year average stats, and posted the following tables:

 

Statistical Comparison among GO/AO, ERA, BAA, and BABIP
3-yr Avg 2005-7	INN	BBd9	Kd9	ERA	BAA	GdAO	BABIP	QS%
Groundball Pitchers	20915.9	3.693	6.112	4.29	0.269	1.558	0.306	51%
Fly ball Pitchers	19704.7	4.347	6.961	4.53	0.262	0.673	0.299	44%
TOTALS-OVERALL	66411.4	4.065	6.574	4.38	0.267	1.052	0.304	49%

 

ALL Starters – 3-Yr. AVG
ERA	BABIP	% Above .302 BABIP	% Below .302 BABIP
< 4.00	.292	26%	74%
> 4.00	.306	70%	30%
4.42*	.302*	52%*	47%*

* MLB Average

 

Stats shown in blue text are the best, while stat numbers in red text are the worst.

 

The interesting comparison is between ERA and BABIP alone, where groundball and flyball pitchers with the lower BABIP have had higher ERA’s and vice-versa, whereas stats for the group of all pitchers combined show otherwise. The three-year averages of all starters also clearly indicate that roughly 1/4 to 1/3 of pitchers defy the correlation between BABIP and ERA.

 

Quoted My Work: “Pitchers who give up more home runs will also have lower BABIPs, and thus higher ERA’s, just like the effect of walks.”

 

Critic’s Response: “This is just plain wrong. Rather than simply looking at home runs, we’re betting off looking at a pitcher’s FB%. The higher that is, the lower the BABIP since a fly ball falls for a hit less often than a ground ball or a line drive. More HRs allowed by a pitcher doesn’t necessarily mean a higher FB%, and thus lower BABIP, because a low K pitcher will also allow lots of HRs since he allows more raw fly balls. And last, I have no idea where he came up with the notion that more walks lead to a lower BABIP. Was this concluded in some study I’m unaware of? How should a walk affect the frequency a ball falls for a hit when a batter puts the ball in play?”

 

He states that my statement is just plain wrong, eh? How about that remedial math I talked about. Below is the simple BABIP formula, and I will insert identical stats for two pitchers, with the only stat being different is home runs allowed:

 

Pitcher A: (100 H – 15 HR) / (400 AB – 150 SO – 15 HR + 5 SF) = .346 BABIP

Pitcher B: (100 H – 5 HR) / (400 AB – 150 SO – 5 HR + 5 SF) = .380 BABIP

 

The plain simple math of the BABIP equation shows the pitcher who gives up more HR’s has a lower BABIP. How can that be wrong? Does 1+1 not = 2? Hello, McFly? We MUST look at HR’s because it is in the gosh darn BABIP equation. So next, how does that correlate with ERA? Let’s look at the three-year averages of pitchers who allowed more and less than the three-year average of 0.117 HRA/HA below. We KNOW that the high-HR rate pitchers will have lower BABIPS because they MUST, according to the math (but not according to Sergeant Hulka):

 

Player-Three-Year Averages	INN	ERA	BABIP	QS%	HRA/HA	WHIP
Pitchers < 0.117 HRA/HA	14363.8	4.202	0.307	54%	0.096	1.39
Pitchers > 0.117 HRA/HA	12789.8	4.649	0.297	48%	0.139	1.40

 

I did not hand pick these statistics to suit my purposes. They are the three-year averages for all pitchers 2005-2007 with at least 10 starts. So how is my statement “Pitchers who give up more home runs will also have lower BABIPs, and thus higher ERA’s, just like the effect of walks” just plain wrong, as the critic eloquently states?

 

Lets add in the walk rates, shall we, because Sergeant Hulka has “no idea where he came up with the notion that more walks lead to a lower BABIP.”

 

Player-Three-Year Averages	INN	BBd9	ERA	BABIP	QS%	WHIP
Pitchers > 3.132 BBd9	10727.2	3.896	4.703	0.301	45%	1.46
Pitchers < 3.132 BBd9	16561.1	2.468	3.733	0.302	56%	1.21

 

Why should we be looking at FB%? That is not part of the discussion or the BABIP stat.

 

I’ll now answer his question “How should a walk affect the frequency a ball falls for a hit when a batter puts the ball in play?” This is fundamental baseball. A walk is obviously another way for a batter to reach base, in addition to hits. With runners on base, say first base for example, 1) the first baseman usually holds the runner on, thus making the opening between first and second base wider for a hit ball (line drive or grounder) to pass through. 2) Ever hear of a “hit and run” play? 3) on a regular play, if a base runner attempts to steal second base, either the SS or 2B has to cover the bag, thus opening the gap wider for a hit. You want stats to back that up? 4) in close games with a runner on 3B, sometimes managers will draw the infield in to try and make a play at home, thus leaving less reaction time for the defense and also having a wider space between the infield and outfield for a ball to drop for  a hit. Again, fundamental baseball. Let’s add runners on base to our three-year average stats:

 

1.268 MOB/9 League Average (three-year averages all pitchers 10 or more starts)

MOB = Men on Base = HA + BB + HBP - HRA

 

Player-Three-Year Averages	INN	ERA	BABIP	QS%	WHIP	MOB/9
Pitchers > 1.268 MOB/9	11287.6	4.89	0.313	43%	1.51	1.384
Pitchers > 1.268 MOB/9	16000.7	4.07	0.294	57%	1.29	1.175

 

The pitchers with the higher men on base rates have a higher BABIP, but that should be obvious, and just plain common sense. These are not my stats – I didn’t make them up.

 

Critic’s Comment: Well of course, no fantasy analyst has ever said that a rise/fall in BABIP alone would be synergistic with batting average, HRs, etc. This is a perfect example of the straw man fallacy as the author argues a point no one is arguing against. A rise in BABIP signifying a rise in batting average would only occur if the hitter’s contact rate and HR/F ratio remained the exact same.

 

OMG. I guess Sgt. Hulka had read every piece of work written by every fantasy analyst, and I guess he can speak for every single fantasy analyst. Yes, indeed, if anybody would like to  pick up a USA Today Sports Weekly and read 3-4 issues in the month of June (2008), and read a major weekly column on Sportsline fantasy sites (two major venues, by the way), you will observe that there are “experienced” and “high profile” fantasy analysts that do, indeed, make that claim, and they even attempt to back up their claim by citing BABIP stats. Furthermore, in his following comments where he criticized my piece on B.J. Upton, he even agreed with another columnist that inferred the exact same thing! I even quoted the analyst who made that claim, and the critic responded to say that he was correct! This is completely contradictory to his statement “no fantasy analyst has ever said that a rise/fall in BABIP alone would be synergistic with batting average, HRs, etc.” What’s going on around here, are gringos falling out of the sky? J’yes, El Guapo.

 

Critic’s Comment. The author obviously meant B.J. Upton, not Justin. I find this quite comical that the author cherry-picked an example of a player that Shandler, and every other fantasy analyst, said would regress in batting average and was actually right about, yet is using this as an example of how the “BABIP theory doesn’t work”. Upton ended up finishing the year with a ridiculous .399 BABIP. Guess what happened this year…Upton is now hitting .263 as his BABIP has regressed to .333, and that has come even with a much improved contact rate. So maybe us BABIP-users were right after all? And just a hint, you’re not going to get anywhere by choosing Ron Shandler as the expert to bash.

 

I was certainly not attempting to bash anybody, rather I was merely referencing a high profile analyst’s work and providing counter-arguments, especially since Shandler is one of the primary advocates of BABIP, and has written many columns in major periodicals about the BABIP stat. Fantasy owners are not obliged to believe everything they read from all high-profile analysts as if their stuff is the Bible.

 

Clearly, I stated “finishing at .300 surely meant that he hit below .300 for the remainder of the year, but nothing near a “crash,” especially in the HR department. My point was that Shandler recommended trading Upton away for the remainder of last year because he said “a crash is coming.” My point was not to buy into the BABIP stat alone to determine overall future performance and the critic also conveniently omitted my following stats for Upton’s post-all-star rankings among all second baseman:

 

B.J. Upton – 2007 Post-All-Star Break Stat Rankings
HR	RBI	Runs	OBP	BA	SB	OPS
1st	2nd	3rd	6th	.285	7th	7th

If fantasy owners took Shandler’s advice and traded away Upton because of Shandler’s prediction that a “crash” was coming due to a high BABIP, then they were not happy. Maybe it’s just me, but do those numbers look anything like a crash? Yeah, I guess we all should trade away a player with second half numbers like those (sarcasm).

 

And why is Sgt. Hulka doing citing Upton’s stats this year? The argument was about Upton’s second half last year. Nobody said anything about this year. Additionally, does the good Sargeant also intend to say that every player who had a BABIP above the league average last year would have lower production this year? Let’s look at more actual stats, shall we.

 

Continue to page 2.

 

Steve “Fonzo” Wayne is the Editor of the Baseball Department at Barracuda Fantasy Sports and is an approved member of the Fantasy Baseball Writers Association. He is also President of the Fantasy Sports Commissioner Training Institute (www.fscti.com). Ask questions or send comments to The Fonze in the Baseball Help section of the forum in Ask the Fonze or email to Fonzo@barracudafantasysports.com.