What Is Sabermetrics? And Which Teams Use It?

It is a simple question.

What is sabermetrics?

Not the history of it, but what is it, right now? What is, in our nerdiest of lingoes, its derivative? Where is it pointing? What does it do?

Last Tuesday I created no little stir when I listed the 2012 saber teams, delineating them according to their perceived embrace of modern sabermetrics.

Today, I recognize I needed to take a step back and first define sabermetrics, because it became obvious quickly I did not have the same definition at heart as some of the readers and protesters who gathered outside my apartment.

I believe, and this is my belief — as researcher and a linguist — that sabermetrics is not statistics. The term itself has come to — or needs to — describe more than just on-base percentage, weighted runs created plus, fielding independent pitching, and wins above replacement.

Sabermetrics is the advanced study of baseball, not the burying of one’s head in numbers.

Personally, I see sabermetrics as breaking down into three separate, equally important distinctions, and one massively important, yet fully amorphous unknown element. Again, fastening on our mathematics overalls, we would describe this as:

f(Sabermetrics) = statistics + scouting + business + ε

In words: Sabermetrics the study of baseball statistics, baseball scouting, baseball business, and anything yet-known or missed by myself (which is the “ε” epsilon).

More specifically:

Scouting is the more subjective study of baseball. All study is somewhat subjective — if it was not, then economists and statisticians would always agree, yet they most certainly do not.

Scouting analyzes the physical attributes and the medical attributes of a player — is he a big-bodied slugger (the type that fades early in the MLB)? does his pitching motion preclude potential elbow injuries or is he a problem waiting to happen? should he have an more open stance? is he tipping pitches? is he stepping early to first base? These are the questions scouting and only scouting can answer.

For amateur players, international free agents, and even guys working through the minors, scouting must be a major component — if not the major component of any analysis concerning them. Not only are these players likely years away from the MLB — and thereby likely to change physically and run the risk of injury — but they are also coming from leagues and schools where the available statistics and league environments produce numbers that are relatively unreliable. There is no UZR, no FSR, no Pitch F/x, no Home Run Tracker.

And at the major league level, scouting can find a problem faster than statistics can. The stats are subject to random fluctuation, but if a pitcher starts falling off the mound in a different way, a scout (or pitching coach in this case) can identify and correct the problem before the statistics — or the team’s record — can even notice.

Scouting can also provide valuable insight into a player’s mental or non-physical attributes. The statistical profile of John Outfielder may say he’s a free-agent catch:

Career: 970 PAs, 31 HRs, 18 SB, 102 wRC+, 26 years old.

But the mental scouting says: Elijah Dukes.

Dukes may be the hyperbole of a player profile in mental shambles, but the need for mental analysis remains. Teams can head off major problems well in advance if they can early on recognize destructive or dangerous patterns in a player’s lifestyle.

Statistics, in my humblest of opinions, breaks down generally — very generally — into three areas league-wide, player-specific, and game strategy avenues of research. These three distinctions, of course, overlap so much as to nearly claim they are all one. And that is fine.

The important note here is that statistics was long-neglected. Branch Rickey and F.C. Lane were about the only two true sabermetricians from baseball’s beginning to modernity because they understood that the contemporary branch of statistics was sorely under-utilized and sorely out of date.

So, in the early 2000s, with the help of the Internet and Michael Lewis’s Moneyball, the world of baseball statistics boomed into a full-fledged fan obsession. Of course, Moneyball, both the book and the movie, helped create a dangerous pendulum swing — the arena of baseball study shifted so heavily into the realm of statistics that scouting began to garner a negative connotation — which is not good.

Lastly, we have the business aspects baseball study. Sadly, because our information is limited, we cannot dive as fully into this field. We may have attendance numbers, ticket prices, contract payouts, and general economic data, but we still lack a vast amount of information and must reverse-engineer answers to our most pressing questions — say, how much does Albert Pujols or Ichiro Suzuki actually bring in annually outside of their on-field production? We do not know for certain, but we can guess.

For clarity’s sake, I see the business branch breaking down into at least four subsets: contracts (which we have almost full information on and have made great strides in understanding), media deals (of which we know very little — only the briefest of press releases give us information on this matter, to my knowledge), stadiums (which include a whole bucket of issues ranging from national to local), and economics (or how the game relates to, is affected by, and affects the national and international economy).

Are these distinctions I am making somewhat arbitrary and not entirely universal? Yes. Hell yes. I have done my best to analyze the general wings of baseball, but I fully recognize this is just one man’s logic. As more voices offer feedback and more fields are uncovered, the breakdown should grow and shift and be thrown out entirely and resurrected later. At current, though, this breakdown makes sense to me.

To neglect any one of these three known branches (as well as the pursuit of the unknown) is to be an incomplete and sub-optimal organizations. It is not to be a losing team, mind you. It is my understanding the Philadelphia Phillies are very much a scouting-heavy and statistics-light organization, yet they have been wildly successful. Likewise, the Seattle Mariners and Cleveland Indians are widely considered fully sabermetric teams, but they have failed to reach a level of sustained success like the Phillies.

In the saber team lists, I used the terms Highly Analytical Organizations, In Between Organizations, and Old School. I confessed in the comments how I longed for better terminology — because I imagine all MLB teams are highly analytical, even if they employ zero statisticians.

So, let us use this three-branch break down to redefine our list of 2012 sabermetric teams. If we look at the teams once again, and analyze where their analytical leverage seems to come from, whether it is coming from just business and scouting or all three branches, we get this:

Alphabetical by mascot. And remember, this is as of 2012.


3 Branch Organizations

Oakland Athletics
Houston Astros
Toronto Blue Jays
Chicago Cubs
Arizona Diamondbacks
Cleveland Indians
Seattle Mariners
San Diego Padres
Pittsburgh Pirates
Texas Rangers*
Tampa Bay Rays
Boston Red Sox
New York Yankees


2-to-3 Branch Organizations

Milwaukee Brewers
St. Louis Cardinals*
New York Mets**
Philadelphia Phillies
Chicago White Sox


2 Branch Organizations

Los Angeles Angels
Atlanta Braves*
San Francisco Giants
Miami Marlins
Washington Nationals
Baltimore Orioles
Cincinnati Reds
Colorado Rockies
Kansas City Royals
Detroit Tigers
Minnesota Twins


1 Branch Organizations

Los Angeles Dodgers**

*I’m especially unsure about these teams.

**Given the Dodgers’ and Mets’ financial problems,
they are the only two teams who appear to be without
analytical leverage from the business branch.
This may have changed recently.

I would like to reiterate (from the last piece) that these are merely my perceptions (coupled with community suggestions). If you work or worked for one of these teams and I have gotten the team’s placement wrong, please feel free to alert me to the oversight. Naturally, I weigh insider’s perspectives more heavily.



Print This Post



Bradley writes for FanGraphs and The Hardball Times. Follow him on Twitter @BradleyWoodrum.


Sort by:   newest | oldest | most voted
ussdavidprice
Guest
ussdavidprice
4 years 4 months ago

could you specify which teams are lacking which branches?

Bill King
Guest
4 years 4 months ago

The “Business” end of baseball is the part we all tend to be in the dark on, and until you’re in a front office with hands on knowledge of where and why these decisions are being made, it’s incredibly hard to find reasoning. All we can do is look at player acquisitions/moves from a purely value standpoint but often times there are so many influences involved from marketing/PR, to ownership preferences, a million different point of views, and then the GM/scouting side.

It does show you, however, when an organization like Toronto gets serious about bringing in someone who’s bright and trustworthy like Alex Anthopoulos and they hand him the keys, that a lot of great things can be done in a short period of time. The teams that don’t adapt are merely pissing away dollars whether they be hard dollars in the bank, or equity %’s left on the negotiating table.

Bill
Guest
Bill
4 years 4 months ago

We may be in the dark about things on the business end, but we do have the ability to make reasonable assumptions. We can say with 100% accuracy that Vernon Wells does not bring in enough in ticket and merchandise sales to make up for his terrible on field play.

CircleChange11
Guest
CircleChange11
4 years 4 months ago

So, basically sabermetrics is “everything baseball”. when a term means everything, it essentially means nothing. Now, at least, with this deifnition we cna claim that sabermetrics has taken over the world and every team does use sabermetrics as its guide.

But, that’s not accurate. Sabermetrics primarily describes situations where teams use data analysis to guide decision-making versus the traditional methods of using scouting, observations, and perceptions. Or to get more nerdy, it’s using data-based performance evaluation over talent-evaluation and performance.

To me calling everything ‘sabermetrics’ is cheating. I don;t think most consider scouting as being “sabermetrics”, and I would think that many would see them as polar opposites in definition and practice. I think I saw a movie not long ago that illustrated the conflict between scouting and sabermetrics, or data-based evaluation and analysis.

Rather than look at the 3 branches as being part of sabermetrics, I would view sabermetrics as being a component of the 3 branches. sabermetrics is definitely, IMO, not the umbrella.

———————————–

It’s amazing how a few hires can drastically change the “stature” of an organization. Do we know whether HOU’s GM is even going to listen to the sabermetrically-inclined hires. Or is it going to be like many other facets of life where Mike Fast comes in with sound analytical cxonclusion based on mounds of data and research and someone’s going to replay (like a politician) and say “Well, my gut tells me differently.”?

If we’re talking game strategy, player usage, etc I would move TEX down the list. There couldn’t be a less sabermetrically-inclined manager than Ron washington.

Really, and I’m not saying this as an insult or anything, but I don;t we really have good ideas on what goes on in each front office, and this is more of an exercise in percepetion.

Mostly, I wanted to object to naming the umbrella “sabermetrics”, when I consider it to be a gear in the machine, but not the machine itself.

ussdavidprice
Guest
ussdavidprice
4 years 4 months ago

agreed, with both parts. And reading your previous post, Bradley, I saw that a lot of the comments suggest that all front offices use quantitative analysis to some degree. For a team to adroitly utilize saber statistics (i.e. not scouting), we know the management hierarchy requires:

1) A quant team that can proficiently collect and quantify data
2) A gm and inner circle that can translate this data into player valuations, and apply it to the player market. i.e., make the most out of the resources at their disposal- money, current players and prospects, draft
3) An on-field manager who efficiently allocates playing time and utilizes in-game strategy

And we can really only see maybe one and a half of those things- we know what managers on the field do, we think we know based on interviews how competent front offices are. There is a lot that goes on closed doors.

So my point is this: while I appreciate the article, Bradley, and can definitely appreciate the thought that goes into it, I don’t think it can be underscored enough that this article is 95% personal speculation, not the result of any sort of rigorous analysis, and not to be taken as anything close to fact or gospel. One of the stereotypical, yet somewhat accurate, complaints one hears about sabermetricians is that they are outsiders who focus on their statistical narrative with no real world context.

It may set the sabermetric movement back fifty years if the community were to prematurely claim that the 56 win Astros are quantitative gurus!

FFFFan
Guest
FFFFan
4 years 4 months ago

I had the same reaction. Metrics refer to to things that are measurable (i.e. can be assigned a number). Much of scouting is complementary to the measurable quantities. Defining sabermetrics too broadly renders it redundant. Many aspects of on field performance can be measured, as well as many aspects of business performance can be measured. Putting those two together into baseball value generation would probably be closer to a useful definition. I do applaud the effort to create a standard definition though, it would save a lot of pointless arguments.

AdamM
Guest
AdamM
4 years 4 months ago

I agree that the definition of sabermetrics used here is too broad. Sabermetrics has always been, and continue to be, the study of baseball statistics, which are necessarily observable and quantifiable.

Front-office operations and scouting are neither of these.

Baltar
Guest
Baltar
4 years 4 months ago

I agree with circlechange. The author’s definition of “sabermetrics” is wrong. His definition is perhaps some ideal of what he would like “sabermetrics” to mean, but “sabermetrics” means, basically, “the correct usage of better statistics.”
When the average ignorant fan raves against sabermetrics, he most certainly does not mean that baseball teams should not scout nor use business judgement.
When I come to Fangraphs, I do not want to primarily see scouting reports and business discussions, though I do not object to having that mixed in with sabermetric results.

channelclemente
Guest
4 years 4 months ago

What intrigued me was the attempt by teams such as the A’s to reduce important, subjective data/impressions from scouting hand waving, to something numerical that can be analyzed objectively. Not a statistic, but a numerical parameter nonetheless. I know it’s not entirely kosher, but if you use the Rsquared determined in most regression models as an index, those models rarely explain more than 20-40% of the variance in a regressed parameter or collection of parameters. Somehow a model with an R squared of .25 doesn’t seem all that satisfying.

supershredder
Guest
supershredder
4 years 4 months ago

Glad you got the Astros high on there – Luhnow and the new owner have just added some cool positions to their staff that deal with sabermetrics.

Rob
Guest
Rob
4 years 4 months ago

It strikes me that the dodgers have been very active on the financial branch, exactly because of their difficulties. They’ve been backloading deals to lure free agents for the past 3 years for one thing. They also back loaded Kemp’s extension. Not sure if that qualifies, but frankly every team, every business makes strategic and analytical decisions with cash flow. I have to agree with just about everything said by circlechange above.

The Dude Abides
Guest
The Dude Abides
4 years 4 months ago

But the Dodgers also threw a crap-ton of money at aging, mediocre to below-average veterans like Mark Guerrier, Juan Uribe, Dioner Navarro, Mark Ellis, Juan Rivera, Adam Kennedy, Mark Treanor, Aaron Harang, Chris Capuano, Mike MacDougal, and Todd Coffey. Not to mention signing an expensive free-agent pitcher immediately after he lost significant velocity on his fastball AND AFTER his MRI showed he had a partially torn rotator cuff (Jason Schmidt), signing an expensive but punchless outfielder with no arm, no on-base ability, and an average steal percentage (Juan Pierre), and an expensive, fat outfielder who hit .222 the season before (Andruw Jones).

Rob
Guest
Rob
4 years 4 months ago

Why stop there? The Dodgers’ mistakes are legion (though there are some good moves too that get less notice). What I was trying to say is that the “financial” branch is somewhat meaningless, or at least it’s not well defined in the OP. What organizations make no attempt to maximize revenues and be creative with their cash flows?

AL Eastbound
Guest
4 years 4 months ago

When I think of sabermetrics I naturally think of Joe Morgan.

CircleChange11
Guest
CircleChange11
4 years 4 months ago

It is amazing that the player/person who might personally benefit most from sabermetric player valuation opposes it the most.

IMHO, Joe Morgan is a great example og why sabermetric-friendly people need to temper what they say and how they say it. Rather than examine the contents of sabermetrics, too many non-saberists get hung up on aspects like sabermetrics saying scouts aren’t important or that computers can do a better job.

I’m sure if you ask Joe Morgan is walks are important, he’s going to say yes. He used them very well in his career. I’m sure if you asked Joe Morgan if stolen base percentage is important, he’d say yes … with the idea that you have to steal safely a lot of the time or you’re hurting your team.

But, that’s not what happened. All Joe Morgan heard was that Moneyball said computers were better than scouts and that baseball was full of idiots and the guys that never played anything more than fantasy baseball before now have all the answers.

We’d sooner take financial advise from the homeless than believe that. non-players and managers know more about baseball that those that excelled at it for decades.

We need to do a better job at explaining ourselves. When the player that perhaps personifies sabermetric principles better than any other player in history opposes sabermetrics that’s a “we” problem not a “they problem”. If we cannot convince Joe Morgan that so much of what he did is under-valued (walks, defense, SB%, power at 2B position, etc), even as a 2-time MVP, then we are failing at our communication.

WilliaminMaine
Guest
4 years 4 months ago

Brilliant. Loved it.

TK
Guest
TK
4 years 4 months ago

In the end, how SABR a team is will come down to how much of what they do the self-described SABR writer agrees with.

TK
Guest
TK
4 years 4 months ago

I would also like some explanation for why you believe Philly is a SABR team? I know you have them on the fence, but between their oldest-school Manager and their financial and personnel decisions (signing Papelbon and Howard; playing Ibanez at all much less over a quality player like Dom Brown), I really want to know. I don’t think you need to subscribe to SIERA or xFIP to think Halladay and Lee are good acquisitions.

Toz
Guest
4 years 4 months ago

The flaw here, TK, is that you are making value judgments on the Phillies managerial and personnel decisions.

By the way, could you let me know which quality Domonic Brown you are talking about, because I am clearly not familiar with him. I know a Domonic Brown who seems to excel in the minor leagues, but can’t hit water out of a boat in the major leagues and is sub-par defensively.

Ibanez, of course, had a difficult to trade contract and was in the third year of a contract in which we knew he would vastly perform in year 3.

For an article about sabermetrics, this comment seems more about perception than statistic.

Richard
Guest
Richard
4 years 4 months ago

“I know a Domonic Brown who seems to excel in the minor leagues, but can’t hit water out of a boat in the major leagues and is sub-par defensively.”

This is the same Dom Brown who posted a wRC+ of 101 in the season during which he was recovering from a broken hamate injury?

TK
Guest
TK
4 years 4 months ago

Unless you actually have real hard insight into each of the 30 teams decision-making structure, this is actually more about perception than statistics, in my opinion. That was basically my point.

JC
Guest
JC
4 years 4 months ago

KLaw has said on the Baseball Today podcast that Philly is one of the less sabermetric oriented teams actually.

mettle
Guest
mettle
4 years 4 months ago

You’re a linguist?!

In what sense?

Jack Weiland
Guest
Jack Weiland
4 years 4 months ago

Subject verb agreement ftw!

John
Guest
John
4 years 4 months ago

lol

RMR
Guest
RMR
4 years 4 months ago

I don’t know why Sabermetrics is complicated. It’s simple:it’s the application of rigorous, evidence-based decision making. That’s it. Sometimes that means stats; often it doesn’t.

CircleChange11
Guest
CircleChange11
4 years 4 months ago

According to Bill James it is the search for objective baseball knowledge.

Other definitions include the computerized measure of baseball statistics, and the analysis of baseball statistics.

Regardless, we can’t just change the definition to fit whatever purpose.

CircleChange11
Guest
CircleChange11
4 years 4 months ago

I would be interested in learning how teams use sabermetrics to make decisions,

Granted we have 2 major books, Moneyball and The Extra 2%, that detail how 2 teams use(d) sabermetrics, but how are teams like the Astros going to utilize MIke fast?

AA and TOR are reputed to be heavily into sabermetrics, by how are they using it? Certainly teams didn;t need sabermetrics to know that Wells was over-paid, nor did they need sabermetrics to know that Bautista could be extended before he became a FA.?

We know that certain teams used sabermetrics to target high OBP, low/med BA players with the idea that getting on base is under-valued.

But, how are sabermetrics being used now? A lot of the small market teams are using sabermetrics because they have no choice in regards to big contracts to stars. They never “make that mistake” because it’s not an option.

If we were to strip away all of the names of the players, teams, GM, etc and just look at talent and contracts, could we tell who is “sabermetric” or not?

Daniel Stern
Guest
4 years 4 months ago

Absolutely brilliant in stating sabermetrics is a work-in-progress; in a weird way it’s like the Constitution: always open to adaptation & interpretation

Bill
Guest
Bill
4 years 4 months ago

Saber-metrics can only be changed through a deliberately rigorous amendment process? I don’t see that.

sc2gg
Guest
sc2gg
4 years 4 months ago

I’ll agree to support your addition of scouting to the Sabermetric umbrella, but only if you support directing all Fangraphs traffic to my website for a three month period. It’s in section 117.1b, on page 562.

TD
Guest
TD
4 years 4 months ago

Interesting, there doesnt seem to be a correlation of wins to how many branches of sabermetrics

Baltar
Guest
Baltar
4 years 4 months ago

I agree. At least in part, the use or lack thereof by the manager seems to be neglected in sabermetric ratings.
I’m not quoting Bochy exactly, but he has in effect said that the Giants supply him with all kinds of numbers which he ignores. He will always play the veteran over the younger, better player, or let a player like Tejada talk him into letting him bat leadoff.

regfairfield
Guest
regfairfield
4 years 4 months ago

Holy crap you actually believe there are teams that employ zero statisticians.

jorgath
Guest
jorgath
4 years 4 months ago

No, just teams that employ zero statisticians who pay attention to stats other than AVG, HR, SB, Field%, W-L, IP, K, BB, and ERA.

regfairfield
Guest
regfairfield
4 years 4 months ago

No every team employs ridiculously smart people, down to the lowest level. Their scouts are well aware of anything that you can find on this site, let alone any kind of proprietary work their stats department may listen to. A team might not way the opinions of their scouts and their numbers guys equally but to think there’s a team that has no idea what FIP is just goes to show how unqualified anyone on this site is to write this article.

I Agree Guy
Guest
I Agree Guy
4 years 4 months ago

The Twins, up until a year ago, did not employ a statistician. Now they have a whole one. Rob Anthony, their assistant general manager, is on record not having the foggiest idea of what FIP is.

Xeifrank
Guest
4 years 4 months ago

Could I see the Org Chart of each team that you used to make these claims?

Mr. Red
Guest
Mr. Red
4 years 4 months ago

Off the top of my head, I would break baseball down into “Quantitative” and “Qualitative.” Both groups would include scouting/player development, team structure, salary structure, business, etc.

There is qualitative scouting, e.g. “Lefty Five Tools has a quick, compact swing that will produce opposite field power.” There is quantitative scouting, e.g. “Freddy Fastball” has an average fastball velocity of 96 MPH and a swinging strike rate in the minors of 17.6%.”

There is qualitative major league player evaluation, e.g. “Ken Clubhouse always keeps his teammates happy and focused on baseball.” There is quantitative major league player evaluation, e.g. “Sam Slugger” has a .393 wOBA over the past 3 years.

There are qualitative business decisions, e.g. “Our marketing campaign needs to focus more on families.” There are quantitative business decisions, e.g. “Our rate of return on the new party deck is 14%.”

I think sabermetrics applies to all the quantitative information pertaining to baseball operations, which would include scouting, player development, player evaluation, and even quantifiable medical information. I don’t think any of the qualitative information qualifies as sabermetrics. I don’t think any of the non-baseball quantitative information qualifies as sabermetrics either.

Mr. Red
Guest
Mr. Red
4 years 4 months ago

And in the last paragraph, I’m referring to scouting, player development, etc. that falls under the quantitative branch of baseball, not all scouting, player development, etc.

Keystone Heavy
Guest
Keystone Heavy
4 years 4 months ago

Didn’t Bill James, the man who coined the term, define it as “the search for objective baseball knowledge” or something to that degree. Wouldn’t that throw out traditional “scouting” out the window.

Tony M. Fernández
Guest
3 years 11 months ago

Is there no objective value that you can get from scouting? You can determine speed, pitch velocity, and bat speed. Furthermore, you can determine build, height, and even mechanical approaches.

And if you have in-game situations, you can scout reaction speed to balls in play, pitch recognition, tendencies, etc. The problem is that we have to make sure that scouts don’t get caught up in things that don’t matter, such as “the look in their eyes” and “competitive spirit”. It’s not to say that mental makeup isn’t important, it definitely is, it’s just that it’s hard to judge that from watching some kids play a game. Leave that part to the professionals.

Snowman
Guest
Snowman
4 years 4 months ago

You can stop being unsure about the Braves. We have a team President, John Schuerholz, who has publicly made mother’s basement jokes about sabermetrics. While sneering. And being absolutely serious.

wpDiscuz