Imperfect Game

To distract ourselves from the relentless midpackness of our NFBC Main Event Team, we have been ruminating about loftier matters. What, we wonder, should be the stat categories and rules of the ideal full-season roto-style Fantasy Baseball League, which let’s call the Universal Baseball Association? We invite you to play along at home, or, as is more likely the case, at work.

First, let’s consider categories. There are five criteria by which we judge our choices of categories. These are:

Fidelity. We’ve often said that, in Fantasy Baseball, you’re not drafting “players” or “teams,” but rather proper names that represent collections of stats. That’s too facile, though. What you’re drafting (or claiming or FAABing) is a bunch of stat collections that, in the aggregate, are supposed to look something like a Reality Baseball team’s aggregate stats do. For one thing, you want your league to reward player performance that corresponds to performance that succeeds on the field. For another, you don’t want your teams to lack the balance that successful Reality Baseball teams have. You don’t want to leave out anything important, except to the extent that Reality teams do, and you don’t want to overemphasize any stats. For example, you don’t want categories that encourage you to do without relief pitchers, or (important point) relief pitchers who happen not to be closers, or players who can run.

(Speaking of which: are we the only people in the world who think that human ability has overtaken the geometry of baseball? The relationship between how fast a guy can run from first to second and how fast a catcher can throw the ball from home plate to second has changed, we think, even since we were striplings, and certainly since the dimensions of the diamond were set. It seems to us, when we examine replays of stolen base attempts, that virtually every runner—95%, say—who tries to steal a base is actually [if the term has any meaning in this context] “safe,” that the Baseball Gods regard this as a disturbance of the equilibrium of the universe, and that accordingly a runner is now out, as opposed to “out,” when the ball arrives in the second baseman’s or shortstop’s glove, wherever located, at about the same time the runner’s spikes hit the base.)

Economy. Up to a point—and we have no idea where that point is—the more categories you use, the better your league’s reflection of Reality Baseball as it’s actually played is going to be. We know there are 8×8 leagues, and for all we know there are 80x80s. But the more interesting question is: how few categories can you use and still get the simulacrum you’re after? And to achieve economy, you have to avoid redundancy. If we can avoid double counting, we should. OPS is in many ways a lovely stat, but it double-counts hits. Why do that if you can avoid it?

Accessibility. We’re being a little disingenuous here. Any stat, combination of stats, or 12-inch dance remix of stats that the human mind can concoct is out there on the Internet and free for the taking. And since you’re reading this, then, like us, you’re probably too far gone to be saved, and you not only know what UBR is but also (unlike us) wish you could (or already do) play in a league that uses it as a category to supplement SB-CS in the Catcher’s Defense category. Or perhaps you play in a league that relies on microscopic distinctions—the 17 varieties of Quality Start, say—among things that look the same to the naked eye of the non-initiate.

But we’re designing the UBA for Normal People as well as for us. Suppose it’s a week before your draft and an owner defects. Everyone you know who’s interested in owning a team already has one; everyone else you know hurries away when you start talking about it. Suddenly, you think of your brother-in-law. He’s a big sports fan, or anyway he watches lots of sports on TV. Baseball’s not his favorite–he roots for certain teams but doesn’t really understand the niceties of the game, or even pay much attention to the game itself when he’ watching it, and he’s absolutely not a stat geek. But you’re desperate, so you give him a call, and he’s good-natured and agreeable (come to think of it, this is all a pretty good description of our brother-in-law), so he asks you to explain the rules. Well, you say, our categories are wRC+, RngR, and xFIP, which we tweak because we think that home runs are more like 11% than 10.5% of fly balls, and we multiply relief pitching performance by a factor reflecting the leverage situation in which it takes place, plus we use REW, though we have no idea what it is or even what the letters stand for, and…. Click. So, to encompass the world’s brothers-in-law in our affectionate embrace, if ESPN doesn’t report the stat, the UBA can’t use it. (ESPN doesn’t compile Holds, presumably because they’re not an “official” MLB stat, but they report them in the box scores, and brother-in-law’s familiar with them—they’re just saves when the guy doesn’t stay in the game–so we can use them).

Elementalness. An awkward word, but let us explain. All stats are “derivative” in an obvious sense: they aren’t the same as the action on the field that gives rise to them. But some stats seem primary. We can quibble about what the precise definition of a “hit” should be, for scoring and statistical purposes, and we don’t even have to call it a “hit,’ but whatever we call it, it’s at the primal core of the game, the I-get-on-base-or-you-get-me-out nub. Less primary are “label” stats: numbers, some very useful, some not so much, that describe an event or outcome, that nobody needed until someone thought them up, and that may themselves have changed the way the game is played. Bill James made this point about “sacrifice” bunts: they (and the way they were scored) changed once someone called them a “sacrifice.” The most notorious modern example is, of course, Saves. And even apparently canonical stats, like RBIs, are of relatively recent vintage. And then, beyond that, there are what let’s call “blender” stats: well-meaning attempts, like WAR or Win Shares, to emulsify many things or everything into a single number. We not only applaud these attempts, we also find them useful in assessing player performance. But in designing the UBA, we want to get as close as we can to the building blocks of the game while still (see above) keeping our brother-in-law on the phone.

Gladiatorialness. Yeah, it’s another awkward word, plus it has some ugly connotations, but all we mean to suggest is the irreducible one-on-oneness of baseball. It ain’t Mixed Martial Arts, but it ain’t football either, and (as virtually everyone who’s ever written about the sport, most notably Roger Angell, has recognized) it boils down to the hitter-vs.-pitcher confrontation. And beyond that, baseball looks like a more individual sport than the other team sports—an “error” by, say, the shortstop, looks more like an individual mistake to the spectator than an error by, say, the left tackle (though that’s changing, too, as videography progresses and Safermetrics catches up with Sabermetrics). All we mean to suggest is that we prefer stats that aren’t situational (RBIs, let’s say) or outcome-dependent (Saves, let’s say, or for that matter Wins), and that reflect purely individual achievement, to stats that are.

So where does that leave us? It leaves us, meaning “us,” as opposed to “you and we,” about to embark on a week of solitary prayer, fasting, and meditation aboard a Disney cruise ship. Really. We’ll be back with you at some point during the week of August 17th. But we don’t want to leave things, as Melville did in Billy Budd, hanging at the end, so here are the categories we envision for the UBA:

Hitting: OBA; ISO; SB-CS

Pitching: (2x QS), and we expect to have more to say about what a “Quality Start” should be; (SV+H-BS); Component ERA.

Fielding: We have an idea, but we’re not sure it’s a good one. That’s one of the things we’ll be praying and meditating about next week. If you need to get started without us, use FPCT, but we’re wondering whether we might be able to do better.