Russell A. Carleton wrote for Baseball Prospectus from 2009 to 2010, and prior to that (2007-2009) wrote at Statistically Speaking. He served as a consultant to a team in Major League Baseball for two years. Beginning today, his work will once again be appearing at BP on a regular basis.
Dear Mr. Morgan,
I owe you an apology. No, not a snarky, sarcastic, "Haha this will get a lot of pageviews and I'll smack him down at the end—Big laughs all around!" sort of apology. A real one.
I'm sorry.
Mr. Morgan, I'm a sabermetrician. I'm one of those new-wave guys who like to look at baseball through the numbers. I never did really play the game, unless you count seventh-grade community center summer softball. Instead of that major-league career I dreamed about when I was a kid, I got an advanced degree and a background in statistical analysis. But when I should have been working on my dissertation, I was reading in Moneyball and Baseball Between the Numbers about other guys with advanced degrees and backgrounds in statistics who were working in the game, and saying to myself, "Hey, I know how to do stuff like that!" I figured that if I wasn't any good at actually playing the game, I could at least do the nerdy next-best thing. I could study it.
About five years ago, I started publishing my own research as a little hobby, and I found out some interesting things about baseball. It was fun, not only in a nerdy way, but in the sense that I really felt like I was contributing to a better understanding of the game.
Had it stopped there, I wouldn't be writing this letter of apology. But it didn't. Something happened to me over time. I'm not sure when it happened, but gradually the work became less about having fun by talking about the game of baseball and more about proving that I knew more than anyone else. I somehow convinced myself that the sabermetric way was the only true way to understand the game… and I laughed at those who said otherwise, including you, Mr. Morgan.
A couple years ago, I had an opportunity that not too many people get. I had a chance to work as a consultant to a real, live team in Major League Baseball. In addition to the tingly feeling that comes with saying “I’m a major leaguer (sorta)!” there was another benefit that I didn’t appreciate at first, but came to treasure. My employers asked me politely to refrain from publishing anything while I worked for them (for obvious reasons), and so I had to say goodbye to BP. While I did miss BP, my departure meant that I didn’t have to have something new ready every Sunday night, which meant that I could go deeper into topics than I had ever gone before. It was the equivalent of entering a sabermetric monastery to contemplate some of the deeper mysteries of baseball.
In my reflections (and uh, Gregorian number-crunching) I came to some rather interesting conclusions. I can’t get into specifics (so please don't ask) but I will say this: there are things that are generally publicly held as sabermetric doctrine—in some cases, crucial underlying assumptions—that are demonstrably false. Statistical models are wonderful things, but they are only as good as the data that power them and the understanding of the programmer who defines them. When I wrote for BP, and before that for Statistically Speaking, in my rush to get something ready, I often went with a very simple statistical model of whatever topic I was studying. There's something to be said for one of my favorite lines, "direction before precision," but ultimately, a simple model assumes a simple reality, and baseball, as I found, is not a simple game.
In other words, I discovered that I was capable of being… wrr… wrrro… wrrrrr… incorrect.
Mr. Morgan, when you said that you weren’t a fan of sabermetric theories, I dismissed you as a stubborn and foolish man who wouldn’t listen to a reasonable argument. I rushed to point out that the "old school" way of doing things—the old “eye test”—was prone to all sorts of biases. People often see patterns where none exist, they can be influenced by a number of outside factors, and they are predisposed to fall back on old folklore and moralistic explanations for behavior. All of those are actually still true, but that's not the point.
My problem was that I didn’t look in the mirror and examine my own methods for the flaws inherent in them. The variables that I chose and the statistical techniques I used reflected my own biases and preconceived notions about what I thought was important, and my own assumptions about reality. When others suggested looking at an issue from another angle, I acted like a stubborn and foolish man who wouldn't listen to a reasonable argument.
In justifying my own views, I talked about how numerical models could take vast pools of data into account (entire decades of games!) and how my models could give unbiased estimates of how important various factors really were. And they can… if the models are good reflections of reality. That's a major point in the sabermetric method’s favor, and to this day, I believe it to be its primary strength.
But had I been a more charitable man, I could have pointed out that the eye test has its benefits too. There are a lot of baseball lifers who have been around the game so long they instinctively— sometimes subconsciously—know to look for things that I don’t even know exist. Not all of their theories and beliefs are going to be right, but I acted as though, by dint of their non-outsider, non-Ph.D-having status, anyone who wasn’t quoting the latest sabermetric research was automatically wrong.
My goal in writing this letter isn't to say that you were right all along about sabermetrics. In fact, Mr. Morgan, I still disagree with you on plenty of issues, most notably in that I believe sabermetrics can offer a lot to the game of baseball. I’ve come to the conclusion that sabermetrics is a young, toolsy prospect. There’s a lot of potential there to be a game-changer, but maybe, just maybe, there’s something to be gained by sitting down and listening to a wise man who’s been around for a while.
Mr. Morgan, I was arrogant and believed that I had the power to answer all questions. I indulged in the idea that someone who didn't speak about baseball in the same language that I did was somehow beneath me. I delighted in pointing out your flaws while ignoring my own. And yes, I laughed along with that website when it made fun of you.
For this, Mr. Morgan, I am sorry, and I humbly request your forgiveness.
peace, love, happiness, banana pudding,
Russell A. Carleton
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now
Welcome back Russell, great to have you back on board!
But, may I point out a similar process that has taken place in public. DIPS has gone from "there is no difference between pitchers in preventing base hits on balls in play" to "Well... sorta... but it's a lot more complicated than that." (Shout out to Mike Fast, among others, for a great deal of work on that.)
Purely wild speculation: Russell's last employer has finished near bottom in UZR/150 the past few years. I'd assume the smart guys with a lot of data there believe some things that oppose current conventional wisdom about measuring defense and/or the amount of value defense has relative to offense.
And just as Joe Morgan might argue that we shouldn't clog the base paths with slow runners, he also knows how to read a change-up a lot better than most.
The point, which some newly-converted zealots seem to miss, isn't that crunching numbers is the One Path to Truth. It's that to give yourself the best chance at finding the truth, an effort to minimize bias is crucial.
Sabermetrics is better at this than old-timey folk wisdom, but as Mr. Carleton says in this fantastic article, there's still a lot of work to be done. That's the best part if you ask me.
Welcome back!
Perhaps this applies to SABRmetrics as well?
There may be some people that believe the numbers are always right and old school baseball knowledge is a myth, but
a. those people are few and far between in 2012; and
b. they are helping us gain insight at a far greater level than anyone that thinks they should "take their nose out of a spreadsheet and actually watch a game".
Well played, sir.
I am reticent to give too much ground back to the Joe Morgan's of the world. I know Kevin Goldstein hates it when we talk about the "war" between stats and scouts and I understand his reasons. I get what Russell is trying to get across in this piece. But seriously, as far as I'm concerned Joe Morgan is the one who needs to apologize for treating an entire class of baseball fans and analysts like shit. From his position as the color commentator on the premier baseball broadcast at the time, he lent his weight to the people who get their jollies by pointing and laughing at the nerdy kids in class. Fuck that.
Granted... the problem is that playing the game can also hide things. Things that are obvious to the sabermetrician, who is able to take a long view and gather a lot of data, are often hidden from the player, who is most interested in the day to day grind of the game.
The great irony of Joe Morgan was that he played the game incredibly intelligently. He was almost the perfect sabermetric player. And then he spent his entire broadcasting career defaming the people who most appreciated his accomplishments. Trust me, wade into the comments on a mainstream site like ESPN. I've actually heard idiots argue that Joe shouldn't be in the HoF because... "look at his RBI's" and "he didn't even hit .300 for his career." Those are the people who Joe Morgan enabled with his senseless animosity towards sabermetrics. Which is a damn shame.
Welcome back Russell - can't wait to wade through the math and read your wonderfully clear writing!
That said, I've always believed that the Saber numbers don't come close to telling you everything. And for all the reasons Russell stated. I've never been able to articulate this nearly as well as Russell did here and certainly not from the unique position he is in. Being very close to the actual inner workings and having access to stuff we don't gives him a very unique view point regarding this type of thing. It's great to have him back.
Regarding Morgan. He is pretty typical of ex-star players and their attitudes towards the "new stats." He believes he was a great player because he worked much harder than most others and he had special winning skills and the ability to do things that win games and championships that can not be captured by some nerd with a computer.
And really, he's right about that I believe. Morgan played on championship teams. He also played on teams that weren't so great in San Francisco, Houston Philly and Oakland and those teams far outperformed what they were expected to do. So he has an idea about things like intangibles and stuff that can't be quantified.
The way he dismisses "statistics" -- as he calls everything regarding numbers -- isn't really correct. But no one should dismiss everything he talked about either. He wasn't all wrong or always wrong or really even close to it.
Example: Higgs Boson. Scientists are sure it's there and it is a factor in mass though they can't measure it directly and can't seem to find one.
Example Two: Slumps. Sabremetricians are pretty sure players can get bad luck and sabremetricins are pretty sure players can slump, though they can't measure how much of it is luck and how much of it is a slump. They also don't know if it's a difference in mechanics, approach, mentality, or a hundred other factors that are hard to measure.
Here in Boston, as you may have heard (teehee!), the media are going apeshit over the supposedly dysfunctional clubhouse and how it's bringing down the team. Yet I can't think of a more dysfunctional clubhouse than the 77 Yankees (managers punching star players, a team captain who hated everyone, most starters asking for trades, etc.), and all they did was win the World Series. How come that team never gets mentioned?
I also don't think sabremetrics has found the Holy Grail of fielding metrics either and the continued confounding debate about it, especially if the mechanics are behind a black box, makes me take a good chunk of sabremetrics with a grain of salt.
Both of those points being said, it doesn't detract from my enjoyment of discussing intangibles or discussing sabremetrics.
Some players have an easier time of this than others -- for example, there's a reason many players will go out drinking after a night game; it makes it (a little) easier to sleep afterwards, to wind down from that three-hour adrenaline rush. Which, obviously, can have an effect on the next game. (And explains to some extent the use of greenies.)
All of these -- and more, such as any individual's changes in focus and effort in "clutch" situations -- are "intangibles" that are difficult if not impossible to measure, but which do have an impact on statistical outcomes. (I think fielding is a different order of difficulty, as it's inherently subjective to say a fielder "should have made" any particular borderline play, and I would think that's an essential part of any granular analysis of fielding.)
That certainly does not mean that we should stop exploring statistical correlations and causations. Just that we need to recognize that the nature of the endeavor is that there will NEVER be a "grand unified theory" that can take statistical results and predict future results -- for a game, a season, or even a player's career -- beyond the typically-accepted 70% accuracy range. Maybe we can improve on that a little, but my instinct tells me that the "human factors" may well be that 25-30% that we can't get beyond.
Sabremetrics gives a good framework for discussion and a vocabulary and a set of rules for comparing and evaluating players. Yet, it's not the end all and be all.
I think many smart baseball fans used sabermetrics to classify other fans and internet interlocutors - if you said 'FIP' then maybe you know what you're talking about. This quick little rubric might have actually had its uses, but like many, I want to actually understand how baseball works, not "win" an argument with/stylishly mock GoTeam25457 on the intertubes. People like you (and Mike Fast) make me think we're still on the right track - we're still learning.
It just seems like if there are huge underlying assumptions that you now know are wrong, but you can't talk about them you will have to just pretend they are right.
I'm in the numbers camp, but Joe Morgan has seen literally thousands of major league games in his life and had a reputation as a brilliant player. I think it's less likely that he doesn't know what he's talking about than it is that he just doesn't have lay terms for much of what he knows.
For instance, "consistency." I don't know what that means, if it means something besides not going into a slump. But it COULD mean the thousand little parts of your routine as a professional, from practice to in-game adjustments. I've got that in my job -- why wouldn't it be the same in baseball?
I don't know. Maybe Joe or someone like him could find a way to tell me what I'm missing not having played the game.
Maybe I'm wrong, but it seems to me that Morgan truly doesn't recognize that the things that made him such a great player are the same things that sabers value so highly. And it's certainly clear from his regular on-air comments that he has never understood the underlying point of Moneyball (hint to Joe: it's not "don't steal").
P.S. I like your comments about consistency.
As much as I dislike Joe Morgan, being able to get insight from someone who has seen thousands of major league games can be invaluable because it can help put data elements and metrics into context.
So why the better real ERA? Well, there are some factors outside of what an estimator considers. First, park effects. Hellickson played in St. Petersburg which allowed 33% less runs last year than Toronto where Reyes pitched most of his games. Half the games are away, but already you should expect Hellboy's ERA to be 16.5% lower.
Next Tampa had the best defense in baseball. They got to 2.5% more balls in play than Toronto (Baltimore, where Reyes played for a short time was even worse). That's good for almost a half run of ERA.
Sure, if you were a GM you'd prefer Hellickson: he's younger, has a better injury history, and he has better stuff (which points to a better future), but last year it was mostly an illusion that he pitched better.
I did see Joe play a lot (when the Giants traded him away, I was totally pissed), and he was definitely one of the headiest players out there -- he observed his opponents, and utilized those observations to help his team win that game. (Bill James wrote about his ability to smell out a pitchout, a great example.) Not often mentioned is the fact that he -- and many of the Reds of the 70s -- were hard-core bench jockeys, who would do their best to verbally intimidate those players they could get to. Just a point that he would do whatever it took to win each and every game. Sabermetrics tends to look more to long-term results. IMO, both are relevant -- I think there is something to Joe's belief that one-run strategies are called for more in the playoff context, for example, when strategies that will play out over a season are dicier in a best-of-seven series.
I enjoyed the book "Sixty Feet, Six Inches," the "conversation" between Reggie Jackson and Bob Gibson. Would love to be a fly on the wall for a similar conversation between Joe Morgan and Earl Weaver . . . I have to think Joe would respect Earl enough to engage him, and maybe acknowledge some things he might not acknowledge otherwise.
Of course, by then, we hope that you'll have another baseball consulting gig, if that's your desire.
Biggest thing I see with hardcore sabermatricians is a tendency to reify useful analytical tools--like, say, WAR/WARP.
Meet the new boss, same as the old boss.