Sabermetrics And You: An Introduction
"'Is there another explanation?' is really, in my view, the heart of science. It's so easy to accept the explanation you have because it sounds good. It's so difficult to say, 'Well, it sounds good, but is there a better explanation?' and to keep looking and to keep testing and to keep trying to refine your explanation. That's something that's really hard to do." --Eugenie Scott
"I've never said, never thought, that it was better to be an outsider than it was to be an insider, that my view of the game was better than anyone else's. It's different; better in some ways, worse in some ways. What I have said is since we are outsiders, since the players are going to put up walls to keep us out here, let us use our position as outsiders to what advantage we can. Let us back off from the trees, look at the forest as a whole, and see what we can learn from that. Let us stop pretending to be insiders if we're not. Let us fly over the forest, you and I, and look down; let us measure every tract of land and map out all the groves, and draw in every path that connects each living thing. Let us drive around the edges and photograph each and every tree from a variety of angles and with a variety of lenses; and insiders will be amazed at what we can help them to see. Or maybe they won't; who knows. But anyway, we'll have some fun." -- Bill James, The Bill James Baseball Abstract 1984
Bill James coined the term 'sabermetrics', a derivation of SABR -- the Society for American Baseball Research -- and defined it as "the search for objective knowledge about baseball." For some people, sabermetrics is merely baseball statistics, which is a lot like saying biology is merely test tubes and beakers. We use statistics predominantly as a means of collecting data, because we can't experiment with baseball players the way we can with plant cells or bacteria. We can watch those players play, and we can design simulations and experiment with those, but we can't make Joe McEwing bat 5,000 times against Randy Johnson to find out if he really does own him.
Make no mistake: we study baseball because we love it. Absolutely love it. We watch a lot of games. Personally, I watch maybe 200 baseball games a year. In many respects, sabermetricians are a lot like all baseball fans. We all learn a great deal by actually watching the games. There are a lot of things we miss, though, not because we're imperceptive or because we aren't paying attention, but because our eyes tend to see what is obvious and miss what is not. We tend to best remember things that happened most recently and forget or minimize things that happened a little while ago. We are so-called pattern-seeking mammals. We consume and embrace evidence that confirms our biases and often reject or ignore evidence that does not. For some people that's good enough.
For the rest of us, though, merely interpreting the game through our eyes and the eyes of those who play it and cover it professionally is insufficient. We love the game and we want to know everything we can about it. The more information we have the happier we are. This ceaseless desire for more and better information doesn't take anything away from our love of baseball; in fact, it only strengthens our bond with the game and our respect for the tiny fraction of human beings who are talented enough to play it for a living. We spend hours a day studying and reading about baseball, because it's awesome and fascinating and comforting and humbling.
One fundamental precept of sabermetrics is that we should always be skeptical of baseball "truths" that are revealed to us by those who play (or manage or are otherwise close in proximity to) the game. Some time ago, if you were told that Roger Maris was a clutch hitter you'd have little recourse but to accept the assertion on face or to rebuke it based on a handful of times you remember Maris striking out with the game on the line. Now you can test the veracity of the "Maris is clutch" statement in any number of ways depending on how you wish to define clutch. Is pitching really 90% of winning baseball games? We can test that. Is bunting in the first inning with your #2 hitter a smart play? We can test that, too. All of this information makes us smarter, more knowledgeable baseball fans. Ignorance may be bliss, but information empowers, emboldens, and enlightens.
The Mets have a new front office that craves information, that is skeptical of institutional baseball axioms, and that is driven to learn as much about baseball as it possibly can. That's what sabermetrics is really about, and the overarching goal of this series is to bring you up to speed on what we know and how we know it (plus what we don't know) about the game of baseball. I urge you to stick around, enjoy yourself, and by all means ask questions. This is a learning process for all of us.
32 comments
|
5 recs |
Do you like this story?
Comments
Yay! Tutorial!!!
I was way wrong about Jeff Franceour . . . but at least I wasn't paid for it.
I'm not yet at all used to the idea that my Mets front office is one
“that craves information, that is skeptical of institutional baseball axioms, and that is driven to learn.”
It’s very disorienting!!
Excellent
I’ve been looking forward to this since the “Cohen remarks.” I’ll try to remind myself that the only stupid question is the one not asked and hope you agree with that. Now if you’ll excuse me I have to go open my mind.
Mostly agree....
Though some stupid questions are asked, to wit, “is Oliver Perez worth $36 million over 3 years?”
It's a triumph of number crunching over the human spirit...aaaaaand, it’s about time. -- Play-by-Play Announcer, The Simpsons.
by MookieTheCat on Nov 11, 2010 9:03 PM EST up reply actions
I prefer Yogimetrics
Baseball is 90% mental, the other half is physical.
That said, this an excellent primer. As someone who is still learning sabermetrics, this is appreciated.
by Russ on Nov 11, 2010 11:55 AM EST reply actions 2 recs
Nicely done
You had me at Eugenie Scott.
That being said, until you basement dwelling SABR-dorks come up with a stat to better understand GRISSION, I cannot fully trust anything you say. I mean, why don’t we see monkeys give birth to humans? You have no proof for anything.
Holy shit, you're right!!
NO STATS ON GRISSON!! HAHAHAHAHAHHAHAHA
This joke old.
I was way wrong about Jeff Franceour . . . but at least I wasn't paid for it.
The true meaning of this
In many respects sabermetricians are a lot like all baseball fans.
I’m not the grammEr policy, I’m just amused how it turned out.
Apologies for my obtuseness, but I don't really get this comment.
Is it the word “sabermetrician”? I was pretty sure that was a generally accepted term for someone who practices sabermetrics.
See: Sabermetrics. Among numerous instances of the word:
Craig R. Wright, a statistician for the Texas Rangers, was the first front office employee in Major League Baseball to work under the title “Sabermetrician.”
Maybe I’m missing something obvious.
I read that as "in many aspects", but then
The sequence of “respects sabermetricians” then “baseball fans”, gave me a twisted interpretation that “respect sabermetricians, baseball fans must” (or the other way around).
Obviously that wasn’t what you meant. I’m just amused by my silliness.
or
In many, respects sabermetricians are a lot, like, all baseball fans.
I.M. Forme
"When you get yourself into trouble is when you feel you have to do something, and then you get yourself in trouble." --Omar Minaya
by itsmetsforme on Nov 11, 2010 4:07 PM EST up reply actions
This post has now been Cerrone-ized.

Chamption of the R.A. Dickey Face contest and "Cromulent Photoshopper Extraordinaire" of Amazin' Avenue!
by Steve Schreiber on Nov 11, 2010 9:19 PM EST up reply actions
the bigger issue has to do with the logic of the count
As written, the logic of the sentence requires either that sabermetricians not be baseball fans, or that they be like a set of people that includes themselves. May I suggest: “In many respects, sabermetricians are a lot like OTHER baseball fans.”
But I disagree. I think statheads are generally better fans, more appreciative of nuance and of beauty. IMHO.
Hopefully, if "people who like advanced statistics like baseball all the same" part is repeated enough,
it’ll sink in, and we’ll stop being accused of not actually liking baseball.
"Blinding ignorance does mislead us. O! Wretched mortals, open your eyes!"
Gil Hodges IS a Hall of Famer.
AA Gamethread Embiggening Record Holder- 453 posts (10/03/10)
by Brooklyn Dodgers Mets Fan on Nov 11, 2010 12:43 PM EST reply actions
awesome
Great introduction, im excited to see how that new front office acts in the off season,
And any extra info to help those of us still getting deeper into sabermetrics is heartily appreciated.
by scubaeric on Nov 11, 2010 2:06 PM EST via mobile reply actions
I think there is a lot of interesting information that comes from sabermetrics (I read this blog religiously). However, I think what turns a lot of people off about the sabermetrics crowd is the fact that many proponents do not necessarily treat advanced statistics with the same amount of skepticism as they do traditional baseball axioms. There is a tendency to accept the newest statistical measure as gospel. This is especially problematic when some of these new measures build off of other somewhat dubious measures. For instance, some present WAR as an absolute truth. But WAR is dependent on UZR, a statistic with questionable merits at best. Most of the contributors on this blog do a good job qualifying any judgements. Many proponents of sabermetrics are not as judicious.
by irishmets on Nov 11, 2010 3:17 PM EST reply actions 4 recs
Correct on all fronts....
It's a triumph of number crunching over the human spirit...aaaaaand, it’s about time. -- Play-by-Play Announcer, The Simpsons.
by MookieTheCat on Nov 11, 2010 9:06 PM EST up reply actions
Not to be pedantic
although I don’t think it’s pedantic. “What we know” is not much. Sabermetrics has provided better evidence which means that our provisional knowledge has improved. We’re wrong less often by smaller amounts than we were when we still used batting average as our primary measure of a hitter’s ability.
One of the best and most important parts of the sabermetric revolution has been getting comfortable with doubt and uncertainty. In “The Meaning of It All,” Richard Feynman argues pretty convincingly that the most important result of the enlightenment was earning the right to doubt authority.
The important part is how we know it. We know it by doubt and proposition and tests. Science, and the transformation of human knowledge—even knowledge about baseball—is about making better guesses.
by jadelane on Nov 11, 2010 4:18 PM EST reply actions 2 recs
"Sabermetric Revolution"
and “The Enlightenment.” Yeah, pretty much equivalent.
I was way wrong about Jeff Francoeur . . . but at least I wasn't paid for it.
I agree that sabermetrics has revolutionized how we think about baseball in a lot of ways. I am just more skeptical about the heavy reliance on statistics to form absolute judgements. Baseball is still a game played by human beings. As with any human endeavor, there are millions of variables that impact performance and decision making. With so many variables, the insight gained by scientific inquiry is limited. Our understanding of baseball should not be approached in quite the same manner as you might approach a hard scientific field, like biology or physics. Now, that does not mean that we should abandon advanced statistics or other innovations in how we measure performance. But rather, we must temper the results of these mathematical evaluations with the recognition that any scientific approach to human interactions is inherently problematic. Not all truth can be ascertained through empirical evidence. Finding new ways of looking at baseball is a good thing; becoming dogmatic about such innovations is not. Absent from such hubris, I think sabermetrics would become much more palatable to more traditional fans.
by irishmets on Nov 11, 2010 4:52 PM EST up reply actions 4 recs
Where have you been all my life?
This is exactly right. Any statistical analysis of human endeavor is an inherently poorly designed study, and in many of the social sciences subjective analysis is used to support or explain whatever “hard” facts one can still distill from human behavior. Add into that variables based on weather and environment, and you have a set of numbers that are inherently fallible. Are they better than your gut? IMO they are. But to argue absolutely that any such system is perfect is just silly. What I think many people miss, however, and as you point out above, is that many people who really know how these numbers operate are careful to qualify themselves when discussing the operation of the numbers. Something I see fairly commonly, however, and in particular when the argument gets away from the numbers themselves and to individual players, is an over-reliance on certain stats, notably WAR, and what might seem—rightly or wrongly—to be a dogmatic adherence to these numbers by a newcomer to the conversation (or the field in general). I’m not sure it’s hubris on the part of those using the numbers, but perhaps lapses in full appreciation of the fact that qualification should be made in application of numbers outside of conversations focused on how they work.
It's a triumph of number crunching over the human spirit...aaaaaand, it’s about time. -- Play-by-Play Announcer, The Simpsons.
by MookieTheCat on Nov 11, 2010 9:14 PM EST up reply actions
irishmets for President
farm team in County Cork.
I was way wrong about Jeff Francoeur . . . but at least I wasn't paid for it.
I find this meatloaf rather shallow and pedantic!
http://www.youtube.com/watch?v=QguWdSsI8S8
It's a triumph of number crunching over the human spirit...aaaaaand, it’s about time. -- Play-by-Play Announcer, The Simpsons.
by MookieTheCat on Nov 11, 2010 9:17 PM EST up reply actions
I think people conflate results analysis and process analysis
There’s a fundamental difference between the two.
results analysis = OPS, UZR, line drive or swinging strike rate
process analysis = breaking down delivery and swing mechanics
I’m not sure where Pitch/FX data lies between these two, though.
Image credit for Jerrysaurus goes to astromets
i think the pitch FX data on a per pitch basis is results analysis (how a single pitch was executed)
as an aggregate of all pitches for a PA though is process (pitch sequence, location vs. hitters strengths/weakness, etc…)
Hey, wait! I'm having one of those things. You know? A headache with pictures?
by KeithsMoustache on Nov 12, 2010 2:00 PM EST up reply actions





























