Sabermetrics and sexiness

If you’re confused, see the note at the top of this post for an explanation

In a fresh post at his wonderful blog, John Sickels gives his rather timid two cents on the newest advances in sabermetrics, exemplified here:

I enjoyed the first two years of grad school and performed well, but in the third year (as I approached completion of my MA), I grew disillusioned. The further I got into the process, the less I enjoyed it, and eventually I lost my love for history altogether. The topics of study grew so granular and narrow as to lose all relevance, and studying nineteenth century Belgian weavers was just one egregious example. At some point graduate school became just a series of ticket-punching classes that you had to get through, an intellectual treadmill. I felt like it was sucking the life out of my soul… I now wonder if a similar process is underway in my baseball mind. I still love baseball, and I still love studying, analyzing, and projecting minor league players. It doesn’t put a bad taste in my soul the way history did from 1994 through 1997. But when it comes to the most advanced sabermetric stuff regarding major league players. . .that old grad school feeling is returning. The newest stuff is becoming so granular that I’m having problems making sense of it. I’m a humanities guy, and the most advanced math is beyond my ability to completely comprehend. 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.

John, I can relate. I am a liberal arts school-attending history major with a concentration in European History. I understand what it is like to appreciate words, essays, and soft research. And yet at the same time I am completely enamored by the current sabermetric advancements taking place, literally grinning every time I see a friend from Twitter try to put together something new and exciting.

But I think we need to approach sabermetrics different than we approach history. I study history because I find it fascinating. I get a great deal of utility out of reading history and studying it, and it’s fun for me, just like you said it was for you. The difference here is that sabermetrics has never promised to be particularly exciting or sexy. In fact, most of the advancements in pitching and hitting statistics (such as wRC+ and SIERA) are admittedly marginal improvements on already useful statistics. So when you say that they are “more complicated ways to say the same basic truths,” you are, to an extent, 100% correct. However, the questions that remain are: 1) how much an improvement are we gaining over the basic truths and 2) how valuable are those marginal improvements? Maybe you find these advances boring and trite, but many others (such as myself) don’t. I’m sure there are front offices and analysts that clamor over the newest posts at Fangraphs and The Hardball Times, just like I’m sure you find the latest breakdown of a hot prospect’s swing riveting. These are, ultimately, questions of what gives us the most utility (or satisfaction), and are completely subjective.

At one point John says:

Or is advanced sabermetric analysis becoming so specialized that no one but physics and math majors can understand it, leaving us humanities majors behind, let alone the average fan?

I don’t think so. Can I keep up with the math Tango has used to break down SIERA? Not much at all. But do I understand the philosophical elements behind wOBA, FIP, tERA, and other advanced metrics? Absolutely. What’s the best example of this? Well, I’m going to have to reveal a deep, dark secret. You know that crazy talented guy that writes stuff like this and this? He’s an English major. Shocking, but just goes to show you what liberal arts majors can do; advanced degrees are certainly not needed to understand the current trend of advanced statistical analysis.

John eventually says:

why am I reading this? I’m not enjoying it. I want to watch a baseball game.

So don’t read it anymore. Close the Excel sheets and turn on a game. Dissect a swing, not a box score. It’s all a matter of what makes you happy, and you’re free to pursue that. Sabermetrics never made any promises to be sexy (at least to everyone), but there are some people out there that do find beauty in sabermetrics. I’m one of them, and I think it’s great.


Pat Andriola is an Analyst at Bloomberg Sports who formerly worked in Major League Baseball's Labor Relations Department. You can contact him at Patrick.Andriola@tufts.edu or follow him on Twitter @tuftspat
24 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Jeremy Greenhouse
14 years ago

Pat, I’m with you. And good closing paragraph.

What I find odd is that Mr. Sickels and others (Mr. Simmons) seem to blame the field of sabermetrics for making these esoteric advancements. Yet nobody’s going to say that historians have gone too far in their research. But for some reason or another, probably since baseball is supposed to be inherently fun, there’s a tipping point where, if you no longer enjoy a part of it, it becomes *our* fault as sabermetricians for going to that new level of analysis as opposed to his fault for going down a path not suited for him.

I don’t blame Nabokov for writing Despair, even though reading it was a shitty way to spend three hours yesterday. I just realize that this novel wasn’t for me, but I understand it has some utility. SIERA’s definitely not for me, and I’m not going to give the metric the time of day, but I understand Tango for some reason gets a kick out of it, and I’m not upset Swartz and Seidman came up with it. I could go on. Thanks for the rebuttal.

John Sickels
14 years ago

I thought I made it clear in the essay that I wasn’t calling for an “end” to sabermetrics or blaming it for anything. I was asking how do we integrate the new knowledge. The problem with studying 19th century Belgian weavers in grad school was that what we were studying wasn’t integrated into any sort of larger whole.

Jonathan Sher
14 years ago

You write that Sabermetrics never made any promises to be sexy. Your claim suffers from a lack of precision and clarity. Fields of research don’t promise anything—that is the domain of people who practice in those fields, and while you my think that distinction isn’t meaningful, let me show you why it is.

As I understand it, the pioneers of Sabermetrics started in the 1960s but its practitioners and followers were small in number and almost non-existent in the major leagues until Bill James began to write books and essays that examined ways both traditional and sabermetric to examine baseball.

James, as I’m sure most know, was an English major and his work in my view was marked by a clarity of thought and writing rather than being steeped in math, though the two aren’t mutually exclusive.  I was always more impressed by his clarity of expression and focus than I was by his math skills or sense of logic, and I say that as someone who writes for a living but also excels at math.

James made sabermetrics sexy, his work popularizing the field beyond the precious few, inspiring readers such as John Sickels and eventually finding influence in major league front offices. That history should be understood so the current period can be placed in context.

It was sexy because James punctured what had long been conventional wisdom and did so by writing in a way that was fun and provocative.

That sense of fun, and frankly, good writing, is too often missing in more recent sabermetric work, though please understand there are important exceptions. Clear and revolutionary thinking have been replaced with with ever more precise ways to measure the discreet actions that collectively make up a game of baseball. Too much of the work is, for me, a tedious read.

I should also point out that John, in his article, doesn’t dismiss the utility of recent work but rather the utility to him and average fans, in his case because he suspects the advanced sabermetric methods will not be directed at prospects, which is his field of expertise. I found it a bit off-putting for you to paint John as someone choosing changes in a prospect’s swing over sabemetric evidence. John has used sabermetric evidence in his work and has added to that toolbox over time.

Whether John is right isn’t a matter of opinion but evidence—will much of the recent sabermetric work enjoy as wide a field of influence as did the work popularized by James? Time will tell but my guess is no. Too much of the current work is inaccessible to many fans, not only because the math is advanced but because the questions are less involving and the answers are not advanced enough, at least not if writing is the measure.

Jonathan Sher
14 years ago

And to Jeremy—You wrote:
“What I find odd is that Mr. Sickels and others (Mr. Simmons) seem to blame the field of sabermetrics for making these esoteric advancements. Yet nobody’s going to say that historians have gone too far in their research.”

What’s odd is the work by John Sickels that drew your ire says exactly what you say is lacking—that historian in academia had gone too far. It’s what prompted him to abandon history as a career pursuit.

Did you even read John’s essay before posting?

Pat Andriola
14 years ago

Jonathan,

But I think the empirical evidence is that John’s worries/claims are not the case. Advanced sabermetrics is as popular as it has ever been. WPA was being talked about on MLB Network yesterday. Dave Cameron and a host of other analysts are writing for WSJ, ESPN, etc. Fangraphs is insanely popular. If anything, I would say that sabermetrics is currently attracting more fans than ever.

Jonathan Sher
14 years ago

Pat,

(1) Stating there is empirical evidence and showing that evidence are two different things and you have done the former. That doesn’t disprove the latter, of course, but we should rely evidence, and when we lack evidence, we should state so, as I did.

(2) Again I think you aren’t as precise in your thinking and writing as is required. Neither you nor John defined what you meant by “advanced sabermetrics”. I quite enjoy Cameron for his work on Fangraphs and USS Mariner but I would suggest he is mostly using conventional sabermetrics rather than advancing its knowledge. In my view he is using it more adeptly than most, and that, along with his ability to write clearly and with attitude, makes his work engaging, much as it made James’ work engaging. Likewise, on fangraphs, it seems to me there are far more articles that make use of existing stats than there are of those developing new ones. It would be interesting and relatively easy to see which articles get more views.

(3) This is strictly anecdotal but I have heard from friends who are much more excited by work with statistics in basketball than in baseball and I think its because the state of that work in basketball is much like it was in baseball in the 1980s, when universally claimed truths were first being punctured.

Dan Novick
14 years ago

Jonathan—

You wrote:

(2) Again I think you aren’t as precise in your thinking and writing as is required. Neither you nor John defined what you meant by “advanced sabermetrics”.

Required by who, exactly? And you should have been able to figure out what “advanced sabermetrics” meant, based on the article itself or the articles linked to in the post.

Jonathan Sher
14 years ago

Dan –

In Pat’s article he defends advanced sabermetrics by pointing to work about wRC+ and SIERA. Then, when I questioned if such work enjoys the popularity and influence of Bill James, Pat responded by pointing to work by Dave Cameron, who I don’t believe has played a role developing those advanced sabermetrics.

A cursory review of Cameron’s writing that’s appeared in the WSJ, also pointed to by Pat, shows a focus on concepts that have been in vogue since Moneyball, concepts such as exploiting mistakes in he way the market values certain skills, be it relief pitching or defense. All very well done but hardly akin to developing SIERA.

In short, Pat has argued for the popularity of SIERA by pointing to the popularity of a writer whose work has little to do with developing new measure using advanced math.

That doesn’t meat my requirement for a persuasive argument. Whether it meets others remains to be seen.

gary
14 years ago

Sickels is right. A lot of this stuff is splitting hairs. Formula A claims to be .00000000001% more accurate than Formula B. Yawn.

Pat Andriola
14 years ago

(1) I did state evidence, such as Dave Cameron’s writing, the popularity of WPA and Fangraphs, etc.       
(2) If your premise is that people would rather read articles in which they cab think less and simply apply what they know, you are aiming low. It is a strawman to claim that I said articles on new stats get viewed more than those which deal with existing ones.

(3) This is also relatively true. To each his/her own.

Gentleman Jack
14 years ago

I think the author largely missed Sickels’ point.  I thought John’s choice of the word “granular” was perfectly descriptive. That is exactly where it’s gotten for many, many hardcore fans.  These metrics don’t help us better understand what we’re seeing, or better predict what is to come.  They lack context and precision, and often have more design flaws than my ‘08 Toyota.  Designing and implementing new metrics has become an academic exercise without utility.

Dan Novick
14 years ago

Gentleman Jack,

You are horribly misinformed.

TCQ
14 years ago

I don’t get where you could make the jump from “articles on FanGraphs using already existing metrics get more views*” to “better metrics need to stop being developed”. I really don’t understand that: a lot of people use simple math everyday(2 + 2 or what have you), but that shouldn’t stop mathematicians from exploring new concepts.

I think Pat’s main point – and I agree with it completely – is that if something is getting too advanced or granular for your taste…don’t read it. There’s plenty of stuff out there(like, say, those articles at FanGraphs) using already established stats. If you want to read all the stuff Tango has done with SIERA, that’s great. If you don’t, that hardly means there’s something intrinsically wrong about it being written.

*if this is indeed the case, and I know you weren’t stating that as fact

Jonathan Sher
14 years ago

TCQ – I’m not entirely sure to whom you addressed your post as you didn’t preface it with a name and than you enclosed in quotation marks a couple of lines that no one said (quotations being reserved for exact quotes and not a paraphrase or attempt at a summary. But since you post seemed to point to issues I raised, I will assume your post was in response to mine—if that wasn’t the case, I apologize.

You wrote, and I do quote, ” I don’t get where you could make the jump from “articles on FanGraphs using already existing metrics get more views*” to “better metrics need to stop being developed”.”

Well, it’s true I believe that most articles on Fangraphs used existing metrics. But I never stated anywhere that “better metrics need to stop being developed.” In fact, I stated the opposite—that better metrics are valued by more and more people in Major League front offices so there is a clear utility to better metrics. So your inference is factually false. But even if I hadn’t made clear the utility of advancing sabermetrics, your inference would be a logical fallacy.

Not once in my posts did I say that I didn’t find advance metrics useful—in fact, I quite enjoy much of the work of Tom Tango. But one of Pat’s key point — so key it inspired the title of the post – is that sabermetrics isn’t sexy—in other words, it has very narrow appeal. That is a factual claim that can be tested, and the way I suggest testing it is to compare the work of a key figure in Sabermetrics in the 1980s and 90s in Bill James to what has been done more recently. The changes since have left some devoted baseball professionals such as Sickels less interested and I think his response is more the norm than the exception.

And that was Sickels point: That recent advances have left cold many people who previously enjoyed sabermetrics. What you think those people should do about that — don’t read it — may or may not be sage advice. But in either case, your advice does nothing to refute Sickels’ argument —that would require evidence and logic that I don’t see in Pat’s article or in the responses of this defending it.

Pat Andriola
14 years ago

Jonathan,

You are unfortunately missing the entire point, and I don’t now if all the logic in the world can save you.

You say that we should compare Bill James’ work to what is being done today, but this is an absolutely absurd notion (and rest assured, Bill James pissed off a traditionalist or two in his time as well). To compare the writings of James, who at the time was discovering the fundamentals of sabermetrics, to what must be left to work with today, which are the marginal advances made in more specified areas of the game, is a fool’s errand.

The point we are making is that we are not trying to mass appeal sabermetrics into a brand for everyone. While sure, the more the merrier, we shouldn’t have to tell the guys working behind the scenes (the Wyers’s, Davenport’s, etc), that they need to “spicen up” the game in order to satiate the appetites of some mainstream fans. Honestly, that’s mostly the job of guys like Cameron, R.J. Anderson, Tommy Bennett, Carson Cistulli, and others who have done an absolutely fantastic job in the role.

But those guys are out there. So when we have John Sickels saying he just doesn’t get it anymore, that everything seems so granular, we try explaining, like I first did, and then at the end conclude by saying: look, if you really don’t like it, and would rather watch the game, then watch the game! No gun is at anyone’s head.

TCQ
14 years ago

Jonathan, my post was indeed directed at you. I figured you would make the inference – as you did – but I could’ve stated that explicitly.

“(quotations being reserved for exact quotes and not a paraphrase or attempt at a summary.”

It was obvious that those were not exact quotes. The idea that you are going to condescendingly act like I broke some cardinal rule of writing in using quotation marks as I did is laughable.

“But I never stated anywhere that “better metrics need to stop being developed.””

True – not in so many words. But you have repeatedly asserted that the continued development of advanced stats is turning many people off of sabermetrics. Here, I’ll quote you exactly:

“That recent advances have left cold many people who previously enjoyed sabermetrics.”.*

*I do know that this was you restating Sickels’ point, but you certainly seem to agree with him

One: you again make a strange leap in implying that reading articles that are uninteresting to you would render the whole of sabermetrics a bore. Two: in stating this(your actual quote, not the conclusion I just drew), I fail to see how you aren’t saying that new metrics shouldn’t continue to be developed, or published, or posted on popular websites or whatever bizarre point you’re making here. And if you actually *aren’t* saying that, then nothing about your(or Sickles’) point is actionable anyway.

Pat already mentioned how ridiculous the whole James thing was, so I’ll skip to a final point.

“If you really don’t like Sickels’ view on this subject, perhaps you shouldn’t have read it in the first place—no gun is at anyone’s head. And if you find that line of argument distasteful— I do—perhaps you shouldn’t direct it at him.”

This is absurd. I enjoyed reading Sickels’ point of view, and I’ve enjoyed reading and entering my two cents into this discussion. I don’t find his argument distasteful, I just find it a bit silly. But his whole point IS that he finds some of the things he reads distasteful – there’s a key distinction there that I’m afraid you are missing.

Jonathan Sher
14 years ago

Pat,

I understand and agree the marginal improvements in some advanced metrics today aren’t remotely comparable to the groundbreaking argument that was suggested and popularized by Bill James. But it is you who invited the comparison, when, in response to Sickels, you wrote:

“The difference here is that sabermetrics has never promised to be particularly exciting or sexy.”

I also understand and agree when you wrote, “The point we are making is that we are not trying to mass appeal sabermetrics into a brand for everyone.” But it is you, not I, who seemed to dispute that very point just a few posts ago when you wrote:

“Fangraphs is insanely popular. If anything, I would say that sabermetrics is currently attracting more fans than ever.”

You write that “John Sickels (is) saying he just doesn’t get it anymore, that everything seems so granular.” But John didn’t write that EVERYTHING seems so granular. He wrote, ” When it comes to the MOST ADVANCED sabermetric stuff . . . The newest stuff is becoming so granular that I’m having problems making sense of it. I’m a humanities guy, and THE MOST ADVANCED MATH is beyond my ability to completely comprehend.”

Sickels is making a distinction between new sabermetrics that makes use of an advanced knowledge of math and sabermetics that does not—the very same distinction that you made in trying to explain your position.

Finally, you wrote that, “You are unfortunately missing the entire point, and I don’t now if all the logic in the world can save you.”

Your argument is premised in a misstatement of Sickels’ position. It is supported on conjecture when evidence can be easily obtained.  And it is defended with arguments that are contradictory. In short, if I must depend on your logic you’ve shown in this thread to be saved, I’m afraid I’m a goner.

If you really don’t like Sickels’ view on this subject, perhaps you shouldn’t have read it in the first place—no gun is at anyone’s head. And if you find that line of argument distasteful— I do—perhaps you shouldn’t direct it at him.

Jonathan Sher
14 years ago

TCQ –

(1) As to your misuse of quotation marks, I wasn’t condescending, I was confused as to whether your post was in response to mine.

(2) You claim I am not interested in advanced sabermetric articles. That isn’t true. I quite enjoy Tom Tango, for example.

(3) You assumed I’m mot interested because I argue that many people are not and YOU make the leap — the inference — that I must not be interested myself.  It’s clear your understanding of logic, of what can and can’t be deduced from a set of facts, is flawed.

(4) You claim that anyone who believes that many people have been turned off by advanced sabermetrics must oppose that advancement. This again is an inference you make that is without merit because you can’t infer the latter from the former. There are countless trends I can observe without endorsing—rising obesity in children, for example.

I agree with you the exchange between you and me is absurd. I enjoy reading Sickels and I enjoy reading Tom Tango and checking his math. But you continually make assumptions about my views that have no basis in fact or logic.

TCQ
14 years ago

Jonathan,

“(1) As to your misuse of quotation marks, I wasn’t condescending, I was confused as to whether your post was in response to mine.”

Fine, it’s really semantics either way. Not a big deal at all – to either of us, I’m sure.

“(2) You claim I am not interested in advanced sabermetric articles. That isn’t true. I quite enjoy Tom Tango, for example.”

Quote me. That’s patently false.

“(3) You assumed I’m mot interested because I argue that many people are not and YOU make the leap — the inference — that I must not be interested myself.”

This is correct, to a degree. You’ve made yourself a proxy for Sickels’ in this discussion and I’ve treated you as such.

“(4) You claim that anyone who believes that many people have been turned off by advanced sabermetrics must oppose that advancement.

No. I said that if you are not opposed to said advancements, then your analysis is in no way actionable. Which it isn’t. That makes your points, while not invalid, a bit petty. If you aren’t offering a solution, then you are essentially whining.

“But you continually make assumptions about my views that have no basis in fact or logic.”

Personally I think you’re wrong, but at the very least you’re a hypocrite.

Jonathan Sher
14 years ago

TCQ—I’ve become bored by you inability to make an argument grounded in logic rather than its opposite and I don’t want to further prompt responses from you that only serve to further illustrate that point.

TCQ
14 years ago

“TCQ—I’ve become bored by you inability to make an argument grounded in logic rather than its opposite and I don’t want to further prompt responses from you that only serve to further illustrate that point.”

Christ, you’re arrogant. I mean, now you’re gonna break out the “take my ball and go home with it” tack? I think I’ll live…

Jonathan Sher
14 years ago

Pat,

(1) As I wrote to Dan, pointing to the popularity of Dave Cameron does little to sport your contention about the popularity of advanced and developing stats such as SIERA. And if you look at fangraphs, you will find most posts and responses deal with projecting players based on much less cutting-edge sabermetrics. 

(2) Again, there is a difference between stating evidence and showing it—the latter would be measuring the number of hits on Fangraph postings to compare what elicits more interest and what elicits less. I would think that would be a useful measure.

(3) Please don’t mistake my comments about the popularity of advances sabermetrics as an opinion about their utility, I would think that if advanced measures prove useful to those who can manipulate the data, that major league front offices will be very interested—in fact they already are. But the fact they are useful to those investing millions of dollars in major league franchises doesn’t make them useful to or even interesting for most fans.

(4) You seem to be saying there is less thought involved using measures we already know than in creating new measures, that somehow, this is “aiming low.” Einstein proved general relativity by using astronomical measures that had been in use since Galileo, measuring the curvature of light during an eclipse. Darwin came up with his theory of natural selection by observing the practice and beliefs of farmers in breeding that had been done for millennia. No one would suggest they avoided thinking by creating a theory, in Darwin’s case, or gathering evidence in support of a theory, In Einstein’s case.

MVD
14 years ago

I’m not sure where I fall on the Casual Fan-Sabermetrics Analyst spectrum. I attend and watch a large number of games, read quite a few blogs, news sites, and magazines, enjoy analytical articles, consider myself knowledgeable about the game’s history and present, and, to an extent, love stats, so I’d guess more on the Saber side than most. I’m not a major in anything nor do I have a degree, but I did excel at Math (including Calculus) in High School.
When it comes to specific stats…I’m glad that Hot Stove’s discussion of WPA was mentioned because I feel WPA is a very solid stat. How much closer to a Win did this player’s performance get us? I also find WAR to be a very solid stat, in that it make a lot of sense that players create Runs which translate into Wins, and Player X is this much better than a Replacement Player. These aren’t the only two I like, but I think I made my point. On the flip side, I still think its important how many Wins a Pitcher ‘earns’ though its not the most important thing to consider when we determine who the best Pitcher is. I also dislike FIP and its kin because there literally is no such thing. The batted ball and the defense behind a pitcher are essential to the game. Yes, it matters that a pitcher gets hitters out, but inducing a GIDP is just as important, and is sometimes better, as a Strikeout. I’m not saying ERA is perfect, but too many people act like FIP is. Math is great (Hits divided by At-Bats!) but when we are taking stats we already have, multiplying them by seemingly random numbers (yes, I know that the individual stats have varying importance and are therefore weighted, but still) and then adding them together, that is kind of ridiculous. Then on top of that, we throw a little x in front of a stat and it becomes an “expected” stat. Its nice to make predictions and theories and have expectations and possibilities and what ifs, but every game that any stat is based on has been played. It doesn’t matter what a player should have done, or could have done without the rest of his team around, it matters what he actually DID. I’m proud of someone like Zach Greinke for paying attention to his FIP because, in essence, he is paying attention to his K’s, BB’s, Fly Ball Rate, Ground Ball Rate, Home Run Rate, but he is going to get hit, and he can’t control what happens once the ball is hit, nor should he have to. Isn’t it enough to tell me that Pitcher X struck out a ton of batters and walked very few? He had a higher ERA than he should have? No, I don’t believe that. Those hits still happened (Yeah I get it, his BABIP was crap, but defense is a necessary evil) and he still gave up those home runs. It seems like some stats aren’t based on what actually happened anymore. There are some stats that I love, but it seems to me the stat community keeps making up new ones that don’t make any sense.

TCQ
14 years ago

MVD,

I’m not going to get into your main post, because that’s really just a philosophical split, but in reference to FIP, it isn’t ERA multiplied by some numbers*. It’s an entirely separate stat that’s just put on an ERA scale(you might know this already, but the quote below is what made me think you didn’t)…

*“I’m not saying ERA is perfect, but too many people act like FIP is. Math is great (Hits divided by At-Bats!) but when we are taking stats we already have, multiplying them by seemingly random numbers (yes, I know that the individual stats have varying importance and are therefore weighted, but still) and then adding them together, that is kind of ridiculous.”