clock menu more-arrow no yes mobile

Filed under:

The Crimson Hour: Usquebaugh towards a Genealogy of BABIP

Wherein I offer a challenge to some of the new old saws of public baseball discourse.

Branch Rickey gazing beyond the maddening crowd.
Branch Rickey gazing beyond the maddening crowd.
Doug Pensinger/Getty Images

When fans' feet on the pavement quicken toward the gates, we cease to measure the length of the day with clocks, whose ticks and tocks now distend ever more noiselessly until we count beats and rhythms in pitches and swings. We frame periods in innings and recover a more intimate acquaintance with beginnings and endings, with the origin of finitude. This is the hour for hushed vitality, for precise attention, for active sensation—sight, sound, touch, but most of all smell and taste. This is the crimson hour.

The All-Star break is a good time for reflection. The hustle and bustle of simultaneous games, the bombardment of highlights, the dispersion of attention are all briefly curtailed. So, I thought today would be a good day to publish a selection of untimely thoughts about the state of art in baseball knowledge.

When I reflect on the sabermetric revolution in baseball, I'm often confronted by the thought that the now up-turned baseball wisdom must have had its origins in genuine insight suited to the game at some point in its history. And given the strength of tradition and lore in the game, these insights were probably garnered early in its history, perhaps even in its pre-history. Players did not just make up stories about how to play the game well and pass them down. However naïve to the sciences, they sifted through their experience to extract advice, tactics, and strategies for success. So, how could they have gotten some things as wrong as sabermetricians have claimed? Is it as simple as human error unaided by scientific restraints?

Before I speculate on an answer, you might want to grab a drink. For abstract reflection I prefer whiskey, a word derived from ‘usquebaugh' a Gaelic translation of ‘aqua vita', the water of life, the Latin name given to distilled alcohol, most likely by monks not Romans. Certainly, there is no better moniker for the best spirit neat. By mixing spicy-sweet grain with a more than moderate amount of alcohol, whiskey prevents the inwardly turned meditator from coming untethered from the world. A sip of whiskey yanks the thinker back when the temptations of rushing from concept to concept into the deep dark night of empty abstraction.

To wit, here's a story passed down in philosophy about H.L.A. Hart, father of contemporary legal positivism, and his then star mentee, Joseph Raz, who is now the foremost legal positivist. Raz is known for his impenetrable prose. That is to say, the people in the world who have been trained to decipher impenetrable prose often cannot penetrate his prose. Anyway, Hart had to read the rough drafts of Raz's dissertation. According to the story, Hart would have two whiskeys per draft-read: first, a preparatory shot; second, one to sip so as to remember how to apply concepts to the world.

So, as I am about to speculate about the origins of our baseball concepts, I advise that you pull a bottle of whiskey—bourbon or rye are most appropriate—off the shelf, grab a rocks glass, drop a cube or sphere of ice in the glass, and sip. If you want to spice it up shake a couple dashes of aromatic bitters into the whiskey.

To return to the question: where did traditional baseball come from? Was it just a matter of naivete or ignorance? In short, the answer here is yes and no. Since the "yes" answer has been given lots of air over the last decade, I'll offer a conceptual defense of the "no" part of that answer.

At bottom, the "no" answer is grounded by the difference between the first-person and third-person perspective on the game of baseball. The first-person perspective is that of the player. Within the first-person perspective we find what a player must believe and do in order to succeed. From the third-perspective we find how various baseball events relate to successful baseballing and develop models for quantifying that success. Because the third-person perspective discounts the mental states of the players, what the players believe about their success and what scientific inquiry tells us about their success can conflict. For baseball players, this conflict raises the question: should I revise my beliefs about what I do in light of this new research?

All this is terribly abstract. [sip] Let's focus on a particular case: the value of walking. A cornerstone of the sabermetric revolution was overthrowing the traditional emphasis on batting average and placing on-base percentage at the center of offensive evaluation. In other words, sabermetricians discovered we should value walks much more than traditional wisdom suggested. This insight was touted as a revelation and resisted by traditionalist like Joe Morgan. And both the resistance and the fact that it was a revelation seem silly. Any brief reflection on how the game works makes it apparent that a walk is in many cases as good as a single and that the difference in value between them in most cases is very slim. So how did the traditional wisdom so thoroughly misvalue walks?

It is implausible to treat the traditional valuation as a mere irrationality; moreover, it is uncharitable, charity being a central value of hermeneutic reconstruction. Instead, we should look for a good reason that the players themselves would have created the traditional wisdom in the first place. It seems to me the distinction between the first-person and third-person perspectives allows us to reimagine the origin of baseball wisdom. For, while from the third-person perspective, the observer can clearly see the value of walks; from the first-person perspective, the player cannot give walking its proper weight in his list of hitting priorities. That is to say, when a hitter goes to the plate, his approach should be designed to produce the best possible outcome: homeruns, specifically; but extra-base hits, generally. The hitter, of course, must take what is given and do with it what he can. He should not strike out to spite the walk, and no hitter would. But rarely would it make sense for a hitter to go to the plate with the walk as his primary goal. It is a tertiary goal, one to accept but not pursue.

To the hitter, then, the discoveries from the third-person perspective sound like the advice that he ought to try to walk. That was sound advice for me in Little League, as an undersized but speedy slap hitter. It is terrible advice for the crème of the hitting crop. Of course, that is not the sabermetrician's advice to the hitter. It is something closer to: design an approach that maximizes getting on-base. To which the hitter will reply, thanks but I'm already doing that.

We still, however, have not sufficiently explained why traditional wisdom so thoroughly ignored on-base percentage in favor of batting average. Even hitters who prioritize hits far above walks should be able to see that the value of their hits is still capture in on-base percentage but their walks are included as well. So, why was on-base percentage ignored? Here the history matters. In the early seasons of professional baseball walk rates were miniscule: less than 2% of plate appearances ended in a base on balls. Indeed, more than 95% of plate appearances ended with a ball in play. (For comparison, these days we're around 72%.) So, the difference between batting average and on-base percentage was negligible. And given how rarely walks occurred it would seem reasonable to conclude that walking was not a skill. (Reciprocally, the infrequency of strikeouts would lead to the conclusion that inducing weak contact was the primary pitching skill. More commentary on this to follow.)

[sip]

So, while traditional baseball wisdom has fallen by the wayside, we shouldn't treat it as a mere irrationality and we shouldn't ignore the conditions under which it would rationally arise. Attending to those conditions can nuance our sabermetric insights and point toward further, deeper research.

To exemplify this conclusion, let's turn to pitchers. We all know today that pitchers do not have a skill that allows them to control how many balls in play turn into hits. Of course, put that way, the sabermetric insight about BABIP is false even by sabermteric lights. But the public understanding of pitching BABIP usually treats pitchers that way, and even the sabermetricians tend to hypostasize the recent conclusions about pitching BABIP as if the very rules of the game of baseball makes pitchers unable to prevent hits on balls in play. However, such hypostasization is a mistake and a cursory review of the history of BABIP demonstrates this. Pitching BABIP is the result of as yet undiscovered historically contingent conditions. And it is quite possible that in previous eras pitchers have had a significant skill for preventing hits on balls in play. Whether this is the case requires more research than I am presenting here.

[sip]

Here's an overview of the history of BABIP. If you look at the MLB average BABIP for each season since 1871 (excluding 1897-1909 because I do not have strikeout numbers), you will see that there are 6 clearly divided BABIP eras with perhaps some transition seasons between them. I feel confident that even though I cannot offer a justification for quantifying how much of a shift in BABIP causes an era break, any reasonable observer would agree with the divisions I have drawn. So, here, look for yourself. In what follows I provide eras, their BABIP average, and the standard deviation among the seasons.

Eras of BABIP Mean
BABIP
Standard
Deviation
The Dawn of Baseball
1871-1892
.277 .010
The Dead Ball Era
1910-1919
.284 .008
The Live Ball Era
1920-1939
.299 .006
The Golden Era
1940-1972
.276 .0034
The Modern Era
1973-1992
.284 .0027
The Post-Modern Era
1993-2015
.298 .0027

[Note: 1893-96 show a sharp uptick in mean BABIP. It looks like a new era had begun but I don't have the data to be sure and I didn't want to include a 4-year era. So, I truncated the data. I acknowledge this is not good statistical practice, but neither is my historiographical model framing the data in the first place.]

And here is a visualization of the eras. The solid blue line tracks the average BABIP of each season and the dashed green line represents the mean for that era.

BABIP Eras

In these six eras we can see two interesting trends. First, the six eras have a repeated cycle within them. The first three eras have average BABIPs that steadily increase and the second three eras start over from the nearly the same starting point, repeating the incremental increases almost exactly. I find this bizarre and in need of both speculation and research, neither of which I will do here. Second, the eras have a steadily decreasing standard deviation among average BABIPs from season to season. In other words, the average BABIP—whatever its quantity—has become more and more standardized over time. Here I would speculate that this is the result of an increase in the pool from which MLB drew its talent. It is probably not a coincidence that the standard deviation of average BABIP makes a jump to a nearly contemporary level after the color-barrier was abolished. By enlarging the talent pool the average talent level of an MLB player increased and the differences in talent level between the best and worst decreased. The result was more standard outcomes on balls in play. This speculation needs testing; however, I'm confident it would be confirmed.

I should emphasize that the changes in the standard deviation does not license the inference that pitchers used to demonstrate a more pronounced skill for affecting whether BIPs become outs. The wild fluctuations in the early game could very well be the result of conditions unrelated to pitcher skill, such as the quality of the balls used, decisions of official scorers, and the quality of bats used, among others. We can only conclude that the .300 BABIP that varies little from season to season is a contingent historical moment produced by a constellation of factors that we as yet have not pin-pointed.

So, the obvious question is: why are there such distinct eras of BABIP? What changes in the conditions of the game altered the likelihood of balls in play turning into hits? There are obvious candidates such as new pitches and better equipment, etc. But we need to study the changes in the conditions of the game more closely to say anything responsibly on this.

To discover that the phenomena of BABIP are historically contingent is to open up the possibility that those phenomena could change. Adjustments by pitchers or hitters might destabilize the present constellation and reveal a new BABIP regime, perhaps even one in which pitchers have a recognizable skill to induce weak contact. But even if the conditions never change, it seems worthwhile to recline for an evening and try to imagine how the future of BABIP might excrete a new era.

[sip]