The league average BABIP for 2007 was actually .316, while this year to date (at the time of this article) it is .306. The three-year average of all players with at least 100 at-bats is .312. Shandler, et, al, has used .300 as a league average for comparison purposes, but it changes from year to year, and we have to compare the stats to what is happening overall at the current climate of the season. Also, using BABIP stats of the many players with less than 100 at-bats probably doesn’t work because many of them were likely part-time players and minor league call-ups that skew the stats of the players we are concerned with. I do not know how Shandler came up with the .300 BABIP to use as a benchmark, but it is more realistically around .311 for the relevant and applicable hitters. I have developed a statistical comparison of batters in 2007 and 2008 who each had at least 100 at-bats in both years, and who each had 2007 BABIPS at least 10 points higher than the league BABIP average of .311.

Fact #1: There were 135 players in 2007 with BABIP’s at least 10 points higher than the league average.

Fact #2: Of those 135 players, 27 of them have higher BABIP’s currently in 2008, nearly 70% through the season. Also, 48 of those players had higher BABIP’s than .300 (Shandler’s league average stat). Therefore, between 20-36% of the players whose 2007 BABIP’s were “abnormally high”, as Shandler would call it, actually have higher BABIP’s this year. Now that is not a high percentage, but significant enough to warrant flaws in the BABIP theory about regressing to the league average. What happened to those among the 135 players that saw a decline in BABIP? While 64-80% of them had lower batting averages, 28% of them had higher HR rates and most of that 28% had higher run-produced rates.

Fact #3: Of the overall 320 hitters, 126 of them (39%) have higher BABIP’s this season, meaning that there is a downward trend in BABIP overall anyway. Why? Perhaps there are more home runs, among other things. (Remember that HR are subtracted in the BABIP formula).

Fact #4: Of the 126 players with higher BABIP’s this season, 24 of them (20%) have lower batting averages. The difference is much more apparent for the pitchers, as the BABIP stat as originally created by it’s inventor was intended as an analysis tool for pitchers, not hitters.                                                                                        

Sgt. Hulka would suggest that I have made his point by showing that a higher percentage of the hitters whose BABIP increased also had an increase in batting average, thus making a synergy between BA and BABIP. I guess he would also suggest to forget about those 20% of the players where the reverse is true. To pick out a hitter and predict that his future batting average performance will increase or decrease based upon his BABIP is risky business, as some hitters have shown to have consistently high or low BABIP’s. You will have about a 72% success rate. BUT here is the kicker about that last statement:

Based on the three-year averages, 30% of the hitters had BABIP’s at least 10 points higher than the league average. For a fantasy team that employs 13 starting hitters, that means it is likely that four (4) of your 13 hitters may have high BABIP’s. The question becomes which of those four will regress towards the average, thus becoming sell or bench candidates? Based on a 72% success rate, it is likely that three of those will have their batting average regress, and one of those is likely to have their HR rate actually increase (see Fact #2). I hope you chose correctly. Perhaps one of those is having a career year. Perhaps a couple of your remaining nine hitters are notorious second half duds, or have poor match-ups remaining. There is a lot to consider, but using the BABIP stat to make your choice for you should not be the #1 indicator, especially by itself.

To visualize this, let’s look at four full-time players who had the highest BABIP’s in the first half of this season.

By all accounts, and according to Shandler and Sgt. Hulka, a “crash” should be coming. Are you going to trade away Jones or Holliday regardless? Could they possibly keep up that pace? Probably not, but does that mean they are going to stink. Pick one to trade, as you have a 72% success rate. Better yet, trade them all, and you should improve on three of them, right? (if you trade for the right player). Here are their current second half stats: 

Hmmm. Whattaya know. Everybody’s previous “abnormally high” BABIP actually increased accept for Chipper, who has been injured, but who still had a higher RP and HR rate than two of the other hitters. Injuries have certainly slowed him up, but you still have the other three players stats who all increased! Holliday’s BABIP was practically the same, yet his BA increased by 33 points! In fact, all three other hitter’s power numbers increased. How can that be? Where’s the crash?

My Work Quoted: “If you buy into the fact that hitters (or pitchers) will eventual regress (or increase) towards the mean of a .300 BABIP (as written by Ron Shandler), whereas a .300 BABIP has been stated as the average benchmark to which all players will gravitate towards (with the law of averages), then I guess the carrer BABIP’s of Wade Boggs (.344), Rod Carew (.359) and Ichiro Suzuki (.357) are just abhorations…”

Hulka’s Response: More misrepesenting what the studies have shown. Once again, choosing to call out Shandler, the author suggests that we believe a hitter should also regress toward the .300 BABIP league average, just like pitchers. This is totally false as Shandler has actually written that a hitter will establish his own personal level in which his BABIP should revert toward, and typically one should look at the hitter’s previous 3 years to determine this level.

If Hulka is so sure that this is a misrepresentation of the studies, then please quote the studies and provide a link and backup that claim. Shandler's article on B.J. Upton did reference a hitter's personal BABIP as a benchmark, but he has also written other articles that indicate otherwise. In fact, I did not wish to cite Shandler’s articles in order to prevent the appearance of bashing him. However, since Hulka has erroneously called me out on this one, here you go. By my count, Shandler has written at least four (4) columns in USA Today Sports Weekly using BABIP stats, and here is a quote from the June 18-24 issue regarding a hitter:

“Ludwick [Ryan] had been the beneficiary of a lucky BABIP: .350. In his 637 at-bats from 2002-2007, he batted .251 with a normal .301 BABIP, so we can see how unusual his averages are.” Shandler went on to say: “but he’ll finish with a batting average around .285. Note that this means he will bat around .265 for the remainder of the season.”

Note that Shandler referred to a “normal .301 BABIP,” which equals that “30% of balls put in play fall for hits” which he cited McCracken. Over the last three years (which Shandler had said a hitter’s personal BABIP will regress towards), Ludwick’s personal BABIP is actually .296, not .301, which means the .301 Shandler referred to as “normal.” Was he referring to .301 as Ludwick’s “normal” level, or was he referring to .301 as THE normal level for all hitters. Based on other quotes regarding pitchers and citing McCracken’s 30% stat, it is logical that Shandler was referring to .301 as a league normal level, especially when Shandler finished the discussion by saying “he will struggle to exceed a .280 batting average, assuming normal BABIP luck”. In any event by the way, since his Ludwick prediction in that June 18 article, Ludwick is currently 23rd in the NL with a .303 batting average and .336 BABIP, not yet in the same country as the .265 predicted. While his BABIP decreased by 14 points, his batting average decreased by only 7 points, a 50% difference. Why the disproportional decrease in batting average versus BABIP? Oh yeah, he has a higher HR rate and a higher walk rate, both mathematical components of BABIP.

From an April 2008 issue regarding pitchers:

“Pitchers whose BABIP is significantly different from .300 can be expected to regress.”

And Shandler again from the July 23-29 issue:

“So if a pitcher has a significantly lower BABIP (batting average of balls in play) than .300, random variance should correct itself and negatively affect the pitcher’s future performance.”

In the article above, Shandler cited McCracken (the BABIP inventor) saying that “30% of balls put in play fall for hits, regardless of a pitcher’s skill.” While that interpretation of McCracken’s work is not completely accurate, Shandler’s number of .300 came from the 30%, NOT a pitcher’s personal level.

Do the research and check your sources, Hulka. More erroneous assumptions and statements by Sgt. Hulka:

Quoted my work: Interestingly, last April of 2007, Bannister posted a similar .243 BABIP, but had an ERA of 4.91. Also, in September/October, he had a “low” BABIP of .253 but had an ERA of 7.30. In fact, the month (August) with his second worse BABIP (.290), he had his second best ERA (2.90). What gives?

Hulkka’s Response: Seriously? This is akin to saying that park factors are meaningless because Barry Bonds posted a higher home OPS in every season since 2002. Obviously the above simply tells you that if Bannister had a higher than .253 BABIP when his ERA was 7.30 in Sept/Oct, his ERA would be even higher, all else equal. The author implies that because Bannister had a high ERA with a low BABIP, then that renders BABIP useless as a metric. The author also makes the mistake of assuming we think BABIP is the only metric worth looking at, as if we simply ignore K/9, BB/9, LOB%, HR/F, and GB%. As a reminder: Bannister’s 2008 ERA = 5.37 and BABIP = .320.

Hulka completely misunderstood the point I was making, as he omitted my basic premise (and primary point), which is “there is no direct correlation between a pitcher’s fantasy-relevant stats (ERA/WHIP) and their BABIP.” In fact, Hulka referenced my point in his first remark, but apparently he forgot. Bannister’s stats that I provided completely supported that a rising BABIP does not result in a rising ERA, or a falling BABIP does result in an improving ERA, as the stats demonstrated that the reverse was true in Bannister’s case. Those stats were contrary to Shandler’s theory that “Pitchers whose BABIP is significantly different from .300 can be expected to regress. A high BABIP (bad luck) means a pitcher’s performance is expected to improve. A low BABIP (good luck) means a pitcher’s performance is expected to drop off.”

Please show me where I said that “a high ERA with a low BABIP, then that renders BABIP useless as a metric.” I never said, or implied that. What should have been clear by reading the entire article was my principle that BABIP should not be used as the #1 primary indicator of predicting future player performance, which is being commonly done by current and contemporary articles, which I have cited here. Hulka also goes on to say “The author also makes the mistake of assuming we think BABIP is the only metric worth looking at, as if we simply ignore K/9, BB/9, LOB%, HR/F, and GB%.” Really? Here is my concluding remarks:

“Far too much over-analysis is spent on a pitcher’s and hitter’s BABIP, including a hitter’s contact rate, especially early in the season. Variables that are likely to have more of an affect on pitchers’ (or hitters’) effectiveness and production include, but are not limited to monthly splits, home-road splits, ballpark effects, opposition (schedule), team-mates, and defense. A pitcher’s “stuff” and skills, and a hitter’s natural skills, have more affect, while “luck” has the least effect.

Quoted my work: In fact, I would be more concerned if a pitcher had a high BABIP (say, higher than the league average of .300), because by THAT thinking, his BABIP should eventually lower (if you trust the law of averages), and that could mean more walks and more home runs along with less singles.

Hulka’s Response: Whaaaaaaaaaat? Where did this idea come from that a high BABIP that regresses toward league average could mean more walks and home runs? One has no affect on the other! A drop in BABIP simply means fewer balls falling for hits, because well, that’s what BABIP tells you (the frequency at which balls in play fall for hits).  

Here’s my answer to that loud “whaaaaaaat?” 1 + 1  = 2, Sergeant. I guess I must once again show the remedial, grade-school level math: Below is the BABIP formula between two pitchers. The set has HR as the only variable (didn’t we do this already?):

Pitcher A: BABIP = (150 H – 20 HR) / (500 BFP – 20 HR – 50 BB – 100 SO – 5 HBP) = .400 (High)

Pitcher B: BABIP = (150 H – 40 HR) / (500 BFP – 40 HR – 50 BB – 100 SO – 5 HBP) = .360 (lower)

The math shows that a pitcher can lower his BABIP by giving up a higher percentage of HR/H. Is giving up more HR’s a good thing for a pitcher’s ERA? Again, my point was that just because a pitcher has a high BABIP and eventually lowers his BABIP over the rest of the year, it does not, in fact, mean that his ERA was also fall. An increase in walks, while raising a pitcher’s BABIP, will lead to a higher WHIP. Do we have to give a math lesson on the WHIP formula too?

Quoted my Work: The other interesting comparison is between ERA and BABIP alone, where pitchers with the lower BABIP have had higher ERA’s. Therefore, if a current pitcher has an abnormally low BABIP, if his BABIP actually does concede to the law of averages over time during a season (which I don’t believe in all the time), it doesn’t necessarily mean that a rising BABIP corresponds with rising ERA or QS%, in fact, the reverse may be true!

Hulka’s response: I have no idea where he gets this conclusion that pitchers with the lower BABIP have had higher ERAs. And then the author follows that up by making the preposterous claim that a rising BABIP might actually mean a declining ERA. As I said earlier, there are many other metrics that need to be examined, so a rising BABIP could certainly result in a lower ERA, but all else equal, this is just wrong.

How can Hulka have no idea where I came up with that conclusion? How can he say it is wrong when I provided the following research and stats as support: Read my entire article, people.

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

* MLB Average

Hulka stated that he had “no idea where he gets this conclusion that pitchers with the lower BABIP have had higher ERAs.”

I guess I need to explain the stats in the tables so that a three-year old can understand them. In the first table, groundball pitchers had the higher BABIP and lowest ERA, while the flyball pitchers had the lowest BABIP and highest ERA. Isn’t that what I said? In the second table, the stats show that 26% of all types of pitchers had higher BABIP’s with lower ERA’s, while 30% of all  pitchers had lower BABIP’s with higher ERA’s. Isn’t that what I said again? The stats show that some pitchers have High ERA/High BABIP’s and some have High ERA/Low BBABIP’s. The same is true for pitchers with low ERA’s (some have high BABIP’s and some have low BABIP’s). You just cannot ever predict whether a pitcher’s ERA (or hitters’ BA, HR, RP’s) will be high or low based upon BABIP alone, or as the #1 indicator – the entire point of my article. And yes, fantasy analyst;s such as Shandler and others have used BABIP as the primary indicator to make predictions about future performance, as I have shown via citations herein.

Finally, it’s almost over.

Hulka’s Closing Remark: Well that wraps up my deconstruction of an article I feel was poorly researched, not backed up by any statistical support, and proved once again that the real reason BABIP and other saber-stats are so criticized is due to a complete lack of understanding and misconceptions about its use. I quoted far more than I expected to, but I just couldn’t possibly pick only a couple of lines to comment on.

In fact, I indeed provided research and stats as backup support in my original article, while Hulka provided none. Hulka clearly misunderstood the purpose of my article and how the BABIP formula works, and he was unable (or unwilling) to comprehend the stats that I provided (along with a lack of basic math). I hope my readers do not make the same mistakes.