clock menu more-arrow no yes mobile

Filed under:

A Little Saber-Rant: Or Dan's Demented Ramblings

On Monday, John Sickels ignited a mini firestorm in the sabermetric community by admitting he's becoming somewhat overwhelmed and tired with the recent advances in sabermetrics.  He still enjoys baseball, but he's beginning to question the relevance of increasing the complexity of our measures for small gains in accuracy.  I'm paraphrasing here, but I don't think I'm too far off from what John is trying to say. 

There have been numerous well-written responses, ranging from Dan Novick and Pat Andriola at The Hardball Times, to Tango over at The Book Blog.  Most seem to grant John's point that at least some of sabermetric study is becoming extremely specialized, and go on to point out how the field in general is becoming more accessible to the average person.

This is my foray into the discussion.  I don't know if I have anything interesting to add, but since it's a topic that's been kicking around my head for a while, you get to hear from me whether you like it or not.

Pushing the envelope in sabermetrics has indeed become the realm of specialists.  I don't think it's possible to argue that (of course someone will, I'm sure).  The last few years have seen the advent of Pitch F/X and the tease of Hit F/X and Field F/X - all of which require some fairly in-depth knowledge in physics to really make sense of.  Many of the advances that aren't physics related are deeply statistical (like SIERA), or deeply computational (like UZR).

This isn't necessarily a bad thing.  And it's certainly not an unexpected thing.  I'll get back to my thoughts on the topic at hand in a moment, but first a historical diversion.

Throughout history, as cultures mature, specialization follows.  The advent of agriculture led to food surpluses, which allowed specialization of labor and the growth of villages and cities. 

During the Renaissance, gentleman scholars like Leonardo Da Vinci or William Herschel, who discovered Uranus and wrote 24 symphonies, led the way - making breakthroughs in many fields.  Fast forward to today.  You need many years of schooling and many years of work to become an expert in just a single field.  Occasionally a newcomer may make an impact on the field, but that seems to becoming more and more rare.

We're at a transition point in sabermetrics.  I'm just too young to have participated in the first big wave of popular sabermetrics - the USENET forums whose discussions were inspired by Bill James' work. (And no, that's not an age joke, Szymborski)  But I have been active in the field for almost a decade now - following Rob Neyer, visiting Baseball Think Factory (back before registration was required), and for the last few years making my own contributions through writing.  I've seen sabermetrics progress from a place where a talented dabbler could make a mark to a place where advanced statistical methods are being used to move the needle slightly forward.

Along with the increase in specialization, we're seeing a tremendous amount of popularizers. Those writers who get the hard stuff and make it accessible to the masses.  These types of writers are essential for the sabermetric community, just as they are for the scientific community.  For every Nobel prize winning physicist, there's the journalist who translates her work for the common people.  For every Patriot, there's a Dave Cameron. Both types of analysts are needed for the field to progress.

And progress it has.  In 2000, Baseball Prospectus published its version of the Hilbert problems for sabermetrics.  For those of you unfamiliar with the reference, at the beginning of the 20th Century, Hilbert, a mathematician, published a list of 23 unsolved problems that he deemed the most important and interesting left to be studied.  Out of the list of 23 unsolved sabermetric problems that BP raised, I count at least 10 that are either "solved" or have had substantial progress made on them.

On the other end of things, sabermetrics has never been more mainstream. Whether it's advanced fielding analysis showing up in the pages of the New York Times, or the MLB Network discussing WPA, saber topics have made inroads into places it's never been before. 

So the combination of specialist and popularizers is a successful one for the field of sabermetrics as a whole.

Why, then, do I feel some of the same ennui as John Sickels?  The easiest answer is that, to some extent, sabermetrics has passed me by.  I'm the dabbler I mentioned up above.  I have the occasional creative idea, and just enough computational and statistical ability to get the idea across. 

But I don't have the knowledge to run a LOESS regression, nor can I calculate the Magnus force on a fastball.  I suppose I could learn how to do those things if I invested the time, but I have a job I like, and a family I love, and I don't have the interest I once did.

Beyond that though, I think there's some staleness in the community right now.  There seem to be three major trends in sabermetrics at the moment: small advances in areas (projections, for example), trying to figure out how to make Pitch F/X data useful, and valuing a player's contribution using WAR. 

WAR may in fact be the holy grail, the one measure to rule them all, but it makes for boring reading when it's the beginning, the middle and the end of the analysis.     

As I'm writing this, I'm in the middle of a Twitter conversation about the value of the Pitch F/X work that's been done to date.  There have been some great pieces, and some flashes of brilliance, but I think we're still looking for the "Eureka" moment that opens up new avenues of study.  Josh Kalk's look at predicting injuries is one of the most innovative things I've seen, and I'm disappointed that he was hired by the Rays and unable to move further on it (it's great for him of course).

Anyway, this has been a long semi-rant, and I do feel very much like an old fogey here - at the age of 30 no less.  After all this, you probably feel entitled to some recommendations.

Unfortunately,  I don't have many to offer.  I'm slightly disillusioned and I don't really know why.  I can tell you what I'd like to see the sabermetric community focus on, but I'm just one man, and my recommendations don't necessarily carry much weight.

I'd like to see a re-emphasis on creativity; creativity while asking the question, creativity while studying the data, and creativity while writing.  Don't just ask how much movement this pitcher's fastball has.  Instead, wonder whether he can control the amount of movement based on his pitch sequence.  Don't just run a regression.  Instead, try to understand the theoretical basis for component interrelationships.  And don't just say Player A had 2 more WAR than Player B so he's the better player.  Examine what A did differently to earn those extra wins.

I'd like to see us focus more on the field of play rather than just the numbers.  That's a strange thing for me to say, because I probably watch a lot less baseball than most of you due to my family commitments.  But we, at times, get so caught up in our calculations that we forget there's a lot outside those calculations too.  Some of those things might be meaningless to our evaluation of the game, but some of them might not be (catcher game-calling for instance) and we just haven't measured them yet.


Finally, I'd like to see a sabermetric community that doesn't become so technically advanced that the average person can't participate to a fairly deep level.  This might be a pipe dream, but I'll say it anyway.  If all of our measures are so far divorced from what the average person can understand, we've lost the battle.  We haven't reached that point yet, but we're within sight of the cliff. We may be forced off that cliff in order to make future advances, but it's not a step we should take lightly.

I know the field is going to continue to move forward, and is likely going to become more and more specialized. It's just the nature of the beast.  We've seen it time and time again in other arenas. And if it leaves me behind, so be it.  I've made my choices and I don't regret not spending more time studying statistics or physics (if only I were still in college).  I know this sounds like a retirement letter, but it's not that.  And it's not a excuse for why I haven't been writing as much over the past 6 months (the twins would be that excuse).  I've had these thoughts for a while but reading Sickel's article made me want to get them out of my head and into a forum for discussion. 

For those who actually made it through these 1500 rambling words, I'd love to hear your thoughts.