Today FanGraphs came out with their early projections for the 2016 season. As is almost always the case with anyone’s statistical projections, there is a lot stuff that doesn’t seem to hew close to human expectations and/or observable reality. Stuff like the World Champion Royals slated to win 79 games or the 78-win Red Sox to be the second best team in the game. Could those things happen? Sure! Anything can happen in baseball. But they’re not necessarily the safest bets. And they’re gonna anger a lot of people, especially Royals fans.
They seem to anger baseball writers a lot too. Each year as soon as the projections come out from FanGraphs or whoever, there is a goodly bit of snark from the ink-stained wretches. “Ha! They have the Royals stinking and the Nationals winning! Silly computers!” says the baseball writer. Never a word, however, for the scores of living breathing “baseball experts” who predicted exactly the same thing last year. Projections are mocked. Predictions — which almost every writer either does on his or her own accord or is elbowed into doing by their editor — are forgotten.
Which isn’t to say I have a lot of use for projections or that I am somehow defending them as superior. I don’t pay too much attention to them and couldn’t begin to do any of my own. I do know enough about them, however, to know that they are not meant to be guesses as to what will actually happen. Rather, they are attempts to approximate what is most likely to happen given variables one can reasonably ascertain and figure. They’re a baseline, really, of what might happen before the totally unexpected and unprojectionable crap that happens every year comes into play. All things being equal, Shlabotnik is going to decline this year because he’s older, Oppenheimer is going to improve a bit because he’s no longer a rookie and if we do this to all 25 guys on the roster, we might be able to make some assumptions.
I think maybe the folks who do projections could be clearer about that when they release them. That all things aren’t equal, ever, and that they are just assumptions built on assumptions, subject to the unfolding of actual events. I know they build that into the numbers too, in the form of error rates and whatnot, but maybe the people who produce the articles presenting the projections could do a better job of clearly stating it in human terms as well. Doing so may seem unnecessary for the projectors — I get an image of a research scientist being told to talk to the popular press about his research, which is always kind of awkward — but it might help with popular adoption of projections by those who currently prefer to mock them. Or, at the very least, would make them think twice about mocking given that it’d be more clear then than it is now that they are, in fact, attacking a straw man.
It’s a humility thing, really. A humility which, to their credit, the writerly class has tended toward in recent years when it comes to their predictions. In their heart of hearts I still think that baseball writers consider themselves experts and think that their predictions are better-informed than that of most people, but they no longer act that way publicly when it comes time for predictions. They preface all predictions with self-effacing humor and verbiage in which they 100% acknowledge that they’re going to look foolish in seven months (and if they’re right, they brag anyway), and I think that makes talking about baseball easier. In my own practice I’ve done that for several years now, and I no longer get nearly as much angry “YOU HATE MY TEAM!” responses as I get “man, that’s a reach, but I guess we’ll see!” responses. The former stops conversation. The latter encourages it.
Anyway, the projectors won’t stop projecting. The predictors won’t stop predicting. Neither of them will stop being wrong a lot of the time because sports are random and unexpected things happen. I would just hope that, at some point, we’d stop mocking folks who do projections and stop claiming that anyone has the market cornered on this stuff.
