I’ve felt the same way about some of the new metrics for a few years now… I have embraced some SABR stats (runs created, win shares, WAR), but a lot of the new ones seem to contribute more to clouding my understanding of the game than contributing to it

But I bet you’ll see pitch f/x in minor league stadiums within 10-15 years. Baseball has become a game of information, and teams will want as much as possible.

I’m like you. I played baseball for a long time and now I’m 27 and a HS coach. I love the game. Then someone posts something calculating some new stat that has just been released. Like that new ERA predictor that I can’t think of… What good does that do me? If I watch the game, I’ll tell you what a play can and cannot do. I’ll tell you if a guy can steal a base or hit the fastball on the black away… and I’ve got my own values on those things especially in the HS game.

I get most of my information from a combination of old school stats and watching guys play. I really hate the defensive metrics to this point as I feel they are to many variables on defense to measure everyone fairly… positioning, a pitcher missing locations, a hit and run play on that takes a player out of position, how the groundskeeper feels about the home team SS… There are just to many variables.

To me, you either have the skill of reading a bat, ball, angles, swing types, pitch types, velocities, and several other items or you don’t. If you don’t, you won’t be in position. If you do, you’ll field far more balls than the counterpart even if you aren’t half as wizardous with your glove and range. You get it done or you don’t.

To me, I like the FIP and BABIP and some others… but GRAR? To complicated for me. That is devil-fingers formula for GRIT by the way…

Trust the process.

Sorry, stupid Royals joke.

As a fan who does not play fantasy baseball, but enjoys following prospects, I like to be aware of the latest metrics without wading too deep. I think most casual fans would benefit from knowing about BaBIP, DIPS theory, etc. so that they can have a better understanding of the large and usually uncredited role that luck plays in the game. I think it is the responsibility of beat writers and columnists to help educate casual fans on this stuff a bit. Some are, some seem to be willfully ignorant.

John, I think you have a good balance, and scouting will never cease to be just as important as stats, especially for prospects, where the small sample, poor defense/field conditions, and other problems abound.

My personal opinion is that the many of the newest metrics (at least in regards to hitting and pitching) are just more complicated ways to say the same basic truths.

I’ve begin to embrace defensive stats like runs saved/UZR, BABIP, and even tRA for pitchers, but most other stuff I just don’t both with, and not because they’re not useful, but just because I don’t think they’re necessary.

Good read! Also nice pic—the original Hank Hill-lol!

It appears that everyone actually read the whole article, thought about it, and gave reasonable responses. I expected at least half of the comments to be something like

“Sickels, you’re an idiot! You suck at math so you hate those of us who are revolutionizing the game! I used to think you were a 5-win writer, but now it seems your WAR has fallen to 2.5 AT MOST”

Well done, John. It takes quite a pair to reasonably doubt the utility of anything SABR-related, especially on the internet with all the killer trolls lurking under every bridge.

The best use of Sabermetrics is determining what actually happened. Any time you record an event, it’s great to have as much data as possible for the interpretive process. There’s the rub, you can have all the data in the world, you can build a nearly pefect picture of a players performance. But, how he did it and why he did it can’t be answered by simply recording the results. Not all .850 OPS’s are equal. You need a skilled observer there with a deep knowledge of the psychological and physiological side of baseball. The problem with that skill is its difficult to quantify which makes people tend to dismiss it. Or at least it sure is tough to win an argument when the other guy can produce a number that makes you look stupid for not understanding that seems to bolster his opinion. And yet, all these guys that love to spring these numbers on you aren’t any better at predicting future results. Most of the one’s I know are flat terrible at it.

Let’s take Lars Anderson for example. I believe he won’t be a home run hitter in the majors. You can tell me his stats all day long, but I spotted a fatal flaw in his swing right away. His swing plane sweeps downward. He’s gonna hit a ton of groundballs. But to hit home runs, he’s going to have have to shave the lower half of the ball just right and get lucky, essentially, to hit home runs. Now you could come up with a metric to measure it of course. And you could painstakingly record the data with program and collate the data into a scientific paper. Or you could just ask me, and I’d tell ya.

There’s so many reasons a player makes it or doesn’t. Many are just unknowable, and completely unpredictable. Like a quiet coachable moment on a long bus trip. Or when Roger Clemens chewed out Curt Schilling for his childish behavior and changed Curts attitude about the game. If I had to gamble, I’d bet those moments are more important than all the scouting and sabermetric data combined.

But I find that a lot of the math-y stuff rarely goes over my head. Sure it’s often slow and I have to reread and rethink certain portions, but I also do that with my history reading (which is fortunately not as granular as what John has described). Now, I’m not saying I could engage in my own original saber research or that I can spout explanations like “how to derive linear weights” or anything complex like that off the top of my head, but there are articles and tools available online that help you do that if you have the time and desire.

Also, like in history, even if the content is boring, the methods by which the author(s) came to their conclusions contribute to my understanding of one way someone thinks like an historian/saber-person.

from finishing the book process and all that, though I share you general feelings about where sabermetrics are right now. You don’t really have much of an offseason with the work on the book taking you through the winter.

John,
I completely agree with your article, but the scary thing for me is that I feel this way about everything these days. Am I turning into one of them guys that will take kids baseballs and never return them??? LOL
Seriously though, From the back of a baseball card it was easy to see who was a stud, and what their ticket to the bigs was. I think the infinite statistical roads are much more valuable to define the gray areas, but totally agree that enough is enough.
I am considering returning to school for a Phd just to decipher the statistical side of my dynasty fantasy team LOL

I certainly understand where you’re coming from. A lot of the higher-level sabermetric stuff being done on offense these days is simply re-hashing the things we already know, adding another layer of complexity to provide a minimal amount of additional utility. We have a really good idea of the relative value of the outcomes of a player’s at-bats, and most attempts to go further tend to lose the forest for the trees. Hell, I’m probably guilty of it myself (see here ).

In essence, you’re looking at exercises of (trying to be tactful) mental self-abuse by people who express themselves better through numbers. That’s not meant to be derogatory, I’m one of those myself. But, a humanities guy like yourself would be much less interested a numerical expression only slightly different from conventional wisdom, much like I check myself out of any Baseball Think Factory thread discussing WWII strategy (you know, the ones that always contain phrases like “Well, that’s a common misconception, but actually…”). Our reactions to pieces of analysis outside our “comfort zone” are likely similar. I read the WWII strategy stuff, and think, “there’s no way these guys can be that certain of their position; way too many variables, way too much noise.” You read the sabermetric stuff you described above and think, “This has so many variables and so much noise, it’s almost worthless to apply to actual baseball”

I think, as you note, that the next step in statistical analysis will come about once we learn more about the application of Pitch F/X and batted-ball data. The sabermetric movement will naturally head in that direction. We just need more data, first. There will be ground-breaking research using these models; maybe we determine certain hitter profiles more likely to translate Minor League success into Major League success. Maybe we identify pitcher profiles more susceptible to injury.

But, for now, the best you can do is to let us numbers guys have our fun, though we may not be offering anything of much use to you in your field. We’ll let you have your WWII discussions in our baseball threads, and we’ll get back to you once we have some idea what the hell this Pitch F/X stuff actually means.

Anyway, I think you do a great job with the site, and are typically one of the few baseball writers who strikes the correct balance of being open to new ways of looking at the game and remembering that it’s still the same game they played for a hundred years. I always find your writing, whether about baseball or other topics, entertaining and well thought-out, and find your insight to be genuine and original. Thanks for doing what you do, and I hope that this “burn-out” period passes.

Regarding Pitch f/x data, it’s likely that it will be implemented in the minor leagues at some point (although I don’t think it will be made public like it currently is).

However, I do think that major league Pitch f/x data can have some relevance to your work. For one, it allows you to take a much more granular look at prospects who have a few hundred pitches in the major leagues, but who’s expected level of performance is still well up in the air. It can allow you to cross reference scouting reports on the guy, and get a rough look at how major leaguers react to the guys strengths and weaknesses. Furthermore, you can use Pitch f/x to study trends in what makes major leaguers successful and compare that to scouting reports on prospects.

It reminds me a lot of recent discussions about “saltwater/freshwater” economists and the various failures of their models to explain or foresee market failures in 2008. I am occasionally mystified when some otherwise rational stat-head uses “science” to prove why some 100 AB performance of some player proves more than a different set of 100 AB selected by equally-arbitrary criteria or his year X results mean more than year Y, when there’s no injury history or other relevant way to distinguish them.

On the other hand, I really am thankful that the spreadsheet community has taught me why RBI and ERA don’t measure… well, whatever people think they measure. And I am ever hopeful that some of this will get through to the Giants’ front office, so they will stop holding interviews in which they claim their offense will be improved by bunting better/more. I definitely glaze over at the advanced metrics, and really don’t have the time to devote to tinkering with them, but sometimes just the principle inspiring the formula or metric or debate is enlightening enough on its own.

And as to your experience in academia, it’s too bad the “granularity” overwhelmed you. The last class I could find to complete my history BA was the history of the working class in industrial revolution-era England, with pretty much every reading involving the great search for the emergence of class consciousness, mostly by Marxist historians (just after the fall of the USSR, lending an air of extra surrealism to the proceedings). At the time I remember talking to my uncle about the class and grad school, and he observed: you go to school more and more and learn about less and less until you know everything about nothing. I suspect he lifted that quote from somewhere, but it certainly summed up a fair bit of my experience, and apparently yours too.

for MiLB players, the research coming from the ever-evolving saber tools will help John and others (306008, for instance) understand what can be or may be going on behind the scenes. The same way learning that ERA, RBI, and errors are poor measures of absolute performance, the newer tools will also help fine-tune the scouting-eye by providing a different perspective.

But when a hot new baseball metric comes out, my initial impulse is to be conservative about using it. Not to ignore it or dismiss it, but to wait-and-see how it pans out, how it tests out after a couple of years, before jumping on the bandwagon.