Lately, I've had baseball on the brain more so than usual. I always have the Astros on the brain because it's my pseudo-job to do so. The last week or two though, I have been engaging in more meta-conceptions of the game we all know and love. It definitely started with Will Carroll's BPro Unfiltered piece about how to make stats more accessible to the huddled masses. What has solidified my current obsession with thinking about how baseball is digested by us as fans was a conversation I had with my girlfriend's father on last Friday.
Like I did with my own father, I gifted him a copy of Michael Lewis' Moneyball as a gateway drug to my way of thinking about baseball. Prior to me baiting him into the statistical revolution with Moneyball, I believe it's fairly accurate to describe him as a decidedly old school fan of the game. When we spoke on Friday night, it was clear that Moneyball gotten through, but the man still had deserved doubts about cursory knowledge he gleaned through reading half a Michael Lewis' novel.
His doubt centered on Moneyball's meme that a walk is just as good as hit. He agreed in principal, but was fraught by the omission that a walk doesn't allow for a runner already on base to advance more than a single base; so while it is mostly as good as a hit, it is not always. The conversation was reminiscent of my own father disputing sabermetricians' claim that there is a lack of streakiness in baseball.
My dilemma with my girlfriend's father was that in order to begin absolving sabermetrics of the overly-simplified concepts presented in Moneyball, was that I very quickly found myself in the territory of Linear Weights, wOBA, and WPA. These are subjects that I literally spent days digesting for them to ring true to me: How am I supposed to convey all of this information is a casual 5-10 minute conversation and make it seem credible in the least? Moreover, I went into such a tangled web of acronyms to simply explain that considering OBP as the best player evaluator isn't a requirement for sabermetric belief That there is far more nuance in the understanding of offensive metrics and that OBP is just ticket into rabbit hole. Who would want to go down a rabbit hole that crazy (aside from most of us...)?
With my own father, it was streakiness. I started citing Markov-chain based studies that determined there is no meaningful correlation between varying samples of previous at bats and performance in varying samples of future at bats. My dad, a stats-savvy man, rolled his eyes as he relived junior college baseball glory and prognosticated on the truth of a streak. Much like any conversation that arises in regards to clutch, I was essentially left to concede that measuring something like that was difficult-if not impossible-because a hot streak, like clutch, doesn't necessarily have to be a measurable skill to be real to some.
After reading the updates from the Sloan Conference via Twitter this weekend, I was struck by the constant commentary of the need for evangelism of the sabermetric/stat-geek message. My knee-jerk reaction-especially after struggling to explain advanced offensive metrics-was full on concurrence. But after considering the implication for a while, I realize that stats don't need zealous brow-beaters, they need a better message.
My favorite example of how the message of sabermetrics is generally conveied is BPro's authoritative, Baseball Between the Numbers: Why Everything You Know About the Game is Wrong. If you were weary of stats-laden analysis, but were a passionate baseball fan, wouldn't you turn your nose up at that? Don't get me wrong, I love that book, but that's not a very effective message for the general masses (which it admittedly was not likely intended to be).
Heretofore, this is kind of been a stream-of-consciousness exploration of my feelings about how to make statistics "sexy." There is a method to this madness. The first is that I hope it sparks discussion amongst us, a group of knowledgeable "sabermetricians" without such a flair for some of the implied dogmatism. The second is to put forth my own suggestion of how to best sell stats.
The one suggestion I have is largely based on my own efforts at evangelizing the good book. The thing I have noticed is that no one wants to hear a longwinded description of what a stat is measuring, how it's derived, or what it's correlation coefficient is. So my suggestion is let's stop making it about the stats. Rather, I think we should sell the idea, or the strategic implications of the idea. That's just about the only way I've ever succeeded.
There are so many other approaches we can take to our "truths." Odds, game theory, etc. All of which can be boiled down in general of terms to make sense on a broadcast, in print, or over a few beers at a sports bar. If reading an array of sabermetric writers (and the discussions the occur amongst like minded thinkers) has taught me anything it's that we don't even hold the stats to be the primary goal amongst ourselves-really. We debate their methodology and qualify their intended purpose, but what never debate is that there are pointing us to certain truths. A truth which doesn't need a four or five letter acronym to be sold to the masses. We quibble about methodology and purpose because we can glean for statistics and our own knowledge of the game that each stat is pointing us somewhere. Yet, when we try to present the map we've made of somewhere, we spend more time fussing over the vehicles that take us there, than the destination itself.
That is my suggestion/observation. It is probably not the greatest. I do, however, know that there are sure to be other suggestions, and that I'd like to see what we can come up with.