June 24, 2016

Prospectus Feature

Baseball and the Chinese Room

by Trevor Strunk

In a fit of writer’s block, I asked a friend what he wanted to read about baseball. My friend, who’s a pretty incisive guy, tells me to write against the sabermetric orthodoxy and argue for the importance of personalities and character on baseball teams. Basically argue against the stats-for-all approach to team building and management and make a case for the oft-beleaguered “character” types.

I thought about this for a while and at first I was going to go in a different direction, because I think the baseball community has pushed back against some of our earlier blindspots about the human element. Managers like Joe Maddon have shown the power of not only solid game planning, but also good people management. (Writers like Russell Carleton have shown the power of those managers.) And the rise of fun, young players that galvanize and characterize teams—think, for instance, of Jose Fernandez’ smile or Trevor Story’s precociousness—have shown us that, even if they can’t be quantified in terms of “wins,” players’ characters matter to the overall aesthetics of the game. So it isn’t like we haven’t grown some nuance as a group: We no longer, to paraphrase the Great Old Man of the game Joe Morgan, root for computer numbers.

But we still have our biases and it might be worth reconsidering how we think about things like character and personality, about “the human element” as opposed to the pure optimization of efficient team-building. As I suggested in my piece on international free agents, these are in fact people being drafted, signed, used, and released by our favorite teams, and there is a story behind all of them. One reaction to this realization is digging into these stories, trying to put more of a face on the game; this generally leads to depression or despair, a lack of true alternatives to recognize the mass of humanity that makes up baseball. What’s that second option, then?

Well, our newly more attractive reaction is to say yes, of course players have personalities, and luck willing they’re nice people who don’t commit violent crimes. And yes, managers are always better when they're fun and don’t commit their own DUIs, et al. But, that notwithstanding, the point of baseball is to win: The point, our second reaction continues, is to make sure your team is in the best possible position to do that. And if the people are good and they win, great; but if you make up a winning team full of Milton Bradleys, Lenny Dykstras, and Bobby Valentines, then so be it. Flags fly forever, and aiming toward a winning clubhouse is preferable to aiming toward a peaceful one.

See, I told you—totally a tempting position, even for we, the enlightened baseball fans. The problem with it, though, is that the human element isn’t distinguishable from a winning baseball team. In much the same way that “character” in an organization doesn’t really come from the GM, the manager, the players, or the scouts, but from some larger conglomeration of those four and more, the distinction between “good group of people” and “good group of baseball players” is murkier than we want to admit. I think John Searle’s famous “Chinese Room” experiment might clarify this a bit.

Searle posed the Chinese Room as a thought experiment to disprove what he called “Strong AI” proponents, or people who believed that computer intelligence could progress to a point at which it was identical to human consciousness. Searle set up a hypothetical situation in which a person is in a room, alone and with a series of books given to them by some authority. On the other side of a wall, there are a group of people composing messages to send over to the first person. The catch, however, is that these messages are in Chinese, and the first person—the one in the room with the books—can’t speak Chinese. Luckily for that person, the books they have give them all the contextual knowledge they need to understand the messages they received and send Chinese messages back. So there’s a sort of communication, and the first person can convince the shadowy others sending the messages that they know Chinese.

This, Searle suggests, is what a computer would be doing when it “communicated” in the voice of a human consciousness—doing a lot of translation back and forth without really knowing the actual “language” it’s speaking. And while that can be convincing to a degree, anyone who’s had an experience with translating or even learning foreign languages will know immediately that there are nuances that are beyond translation. Not just turns of phrase, but entire words and signifiers that are illegible without a working knowledge of a particular language. And so while a computer—or a beleaguered government worker translating Chinese—might be able to spoof the language for a while, the stitches are going to show eventually, naturally. Because a dictionary isn’t the sum of language, and data is not the sum of consciousness.

You can begin to see how this translates back to baseball’s human element. While it’s deeply compelling to imagine that we can privilege data over people and be done with baseball’s Gordian knot, it’s just not that easy. Each number that is attached to a position player, or each point on a GM’s CV, or each pennant attached to a manager’s resume does not exist in isolation. While we need not become Ecksteinian about this—and let’s be clear, much of the “character” arguments that sabermetrics and baseball analysis have attacked have been racist and inane—we unfortunately cannot throw the baby of the human element out with the bathwater of the past. Because good play on the field is not just indistinguishable from the qualities of the people playing, but homologous to it.

The upshot of this all is that personality counts, but not in the ways that we’ve traditionally understood it. Hard workers and gym rats are well and good, but in a sport where everyone works out constantly or washes out, that’s not nearly as important as we might expect on the outside. But the .300 average is paired with a particular person, as is the 1.150 OPS or the sub-2 ERA. And while “headache players” might be double-speak for “non-white” at times, there are players, managers, and GMs who simply cannot overcome their personality issues. Sometimes that inability is paired with numbers that, on the surface, look great, and then we have to have the conversation. But perhaps, more often than we expect, quality numbers on the field are quietly supported by a good personality off the field. It’s a vaguely conservative stance, and one that I’m a little gunshy making, but it’s one I think I believe.

Trevor Strunk is an author of Baseball Prospectus. Follow @hegelbon
Click here to see Trevor's other articles. You can contact Trevor by clicking here