ReOpening the Old Stats versus Eyes Discussion

[Brisco]

Baseball is back, but until we get to roster decision time, there isn't much to talk about, and since I do not feel like getting any work done, i thought i would reopen the old Stats versus Eyes discussion/debate.

At heart, I love the general concept and idea of sabrmetrics/statistical talent evaluation, but I thought I would bring up the darker side of this phenomenon.

Simply put, I believe that sabrmetrics creates a false self-view of knowledge/competence/credibility in the common fan such as myself.

First, the sum total of my actual baseball expertise is 2 years in little league, over 35 years ago. I can't tell one type of pitch over another unless one of them is a knuckle ball; I would not recognize a 'hole' in a swing if i was standing in it. In short, I am a total ignoramous in terms of actual baseball.

But you know what? I can add, subtract, multiple and divide and do all that other stuff you find on a basic calculator. I even have an informal background in statistical analysis. The other thing I can do is argue... I have done that for a living most of my life. As a result, I can come up with and somewhat effectively advocate just about any baseball conclusion out there... EVEN THOUGH I DON'T KNOW A DARN THING ABOUT BASEBALL.

I would say that those three skills: Basic Math, Basic Statistical Analysis, and Argumentation are possessed by about 90% of the folks on the internet. Suddenly, all of us are baseball experts, despite the fact that only the tiniest percentage of us have the baseball knowledge/experience to add meaningful insight on Major League caliber baseball skills/strategy/tactics/play.

To make matters worse, because our sole claim to competence comes from statistical analysis and comparison, we look with disdain on any factor that cannot be reduced to an objective statistic. On this and just about any baseball board, there are countless posts ranging from sarcasm to full blown disgust in reaction to any mention of intangible factors such as: clubhouse leadership, lineup stability, the litany of non statistical catcher skills, mentoring, confidence, the eye test, and so on.

Because we cannot attack or defend these various factors with statistics, we give them little, if any, weight in the arguments. I mean, why would you ever choose a player with a career/current OPS that is 30 points lower? Clubhouse presence? Balderdash. Or, don't tell me that a guy who can't OPS over .720 is an asset to the team just because he is a "leader."

But the fact is, 30 major league ball teams all pay scouts to provide more insight about players than what the statistics show. Why?

Should we accept that Scott Rolen had a greater impact on the Reds in 2010 than we can justify with statistics or not? (I use this merely as a Reds example of the issue... please do not turn this into a Rolen debate).

There is one more dark side to using statistics as our primary means to evaluate players... they cause us to have very little patience with all but the most proven of players.

If Votto goes 2 for 30 over a stretch of games, we will just wonder if something is wrong. If anyone else does, we want them benched, traded or sent down. Why? because we can use that statistic to justify our fears that X player: lost it, never had it, was overrated, was rushed to the majors, and so on. If we can convince ourselves of this, then we spend our time trying to convince others.

If the player turns it around for a while... we put our fears on the back burner, but the second they go into a statistical slide again, we pull all the old arguments back out. We ask "Why Can't the guys who are paying millions to make these decisions see things that are so obvious to us?"

...and then we conclude that they must all be idiots.

Ahhh... the glorious power of statistics. I can jump from ignoramous to expert with nothing more than a calculator. Fear me!

[westofyou]

Stats are the result of action; contained and distributed in a data format. Simply put they are a recording of an event. They do not replace the event or the human recording of the event.

The assumption that people eschew the beauty of the game over pure data is true in some cases, but can not be used as a blanket statement that covers the whole world of folks who follow the game with a foot planted firmly in both worlds.

MLB teams have both feet in both realms of the game, don't kid yourself that a scout isn't using stats and don't kid yourself that stat heads aren't examining the pitchers form.

[WrongVerb]

Stats are the result of action; contained and distributed in a data format. Simply put they are a recording of an event. They do not replace the event or the human recording of the event.

The assumption that people eschew the beauty of the game over pure data is true in some cases, but can not be used as a blanket statement that covers the whole world of folks who follow the game with a foot planted firmly in both worlds.

MLB teams have both feet in both realms of the game, don't kid yourself that a scout isn't using stats and don't kid yourself that stat heads aren't examining the pitchers form.

To augment your idea a bit...the "eyes" get to see the humanity of the individual player: his skills, his talent (which is the ability to apply skill in a given situation), his make-up, where he might be physically in 5-10 years, how athletic he is, etc. The stats are a recording of how well that player actually puts it all together.

[Johnny Footstool]

Brisco, every point you make about the hubris of stats geeks can be flipped around to reflect the hubris of scouts.

"Who cares if a player is only hitting .220? He's fast, and that *must* create havoc on the basepaths."

"So what if your veteran third baseman only hit 3 home runs last season. He brings presence to the dugout."

"I don't care about Juan Castro's range factor; he looks smooth when he fields a ground ball, so he must be a great defender. And I know, because I played two years of NCAA D-1A ball."

The problem doesn't stem from one side or the other. The problem you're describing is the result of everyone having an opinion and insisting that their opinion is *right*.

*IN MY OPINION*

[RedsManRick]

I agree with your point that many people are much too eager to jump to conclusions based on the available statistical record. However, as Johnny pointed out, the exact same criticism can be legitimately applied the other way.

The wrong conclusion to take from your observation is that we should dial back our use of stats or mistrust those who use them. Rather, it's that we always be humble and recognize the limits of our knowledge while at the same time constantly pushing ourselves and others, testing our assumptions about what we really know and how we know it.

Ultimately it comes down to this, when we're talking about baseball in the public sphere, I refuse to put my faith in people who claim special knowledge but refuse to engage in any critical examination of it. Show your work, keep an open mind, and always be open to new evidence. If you aren't willing to engage in that conversation, if you're only willing to assert a position but not examine it critically together, then I don't really see much point in having a conversation.

[AtomicDumpling]

I think most people would agree that you need to have a solid foundation in both stats-based and "eyes"-based baseball knowledge to have a thorough understanding of the game.

Personally I enjoy both sides. I enjoyed playing baseball through high school and have made a big effort to learn about the sciences of hitting and pitching in recent years. I now know a lot more than I ever did when I was playing. I have definitely learned enough to know that my coaches didn't have the knowledge to properly teach me and my teammates during those years. Pretty much everything we were taught was wrong. I appreciate all the hard work and dedication my coaches did for us and I know they were doing what they thought was best. I just wish I could go back and start over knowing what I know now. How does the song go?: "I wish... that... I knew what I know now... when I was younger".

I am definitely a proponent of the sabermetric or analytic approach to studying the game as well. The computer age has enabled number-crunching that has dramatically changed the game of major league baseball. Many age-old adages and strategies that were once taken as gospel have been proven wrong. Many new ideas have been developed and have improved the game.

The only real problem I have with sabermetric folks is that some of them study the science not so they can understand the game better, but rather because they want to convince other people that they know more than they do. Some people see sabermetrics as a way to win an argument and want to shove their numbers in the face of other people. That has turned a lot of people against sabermetrics. Just because sabermetrics spits out a definite number does not mean that number is the end of the argument. A players WAR score is not the only thing you need to know about a player, in fact it is only a tiny fraction of the real story.

Baseball is a game with many layers. There are many different ways to enjoy this game.

[Kc61]

Happy to dust off the debate, but what exactly are we debating?

How to measure past performance?
How to predict future success in baseball?
How to select the MVP or an All Star?
How to determine if an older player still has it?

IMO, the use of stats and "eyes" depend on the question being asked. Clarity in the question will help determine the answer.

Measuring past performance, for example, I think stats are far more significant than "eyes" as long as the sample is reasonably large.

Predicting the future of a raw rookie, I think minor league stats are relevant but clearly the eye test has a role.

And at this point in time, I doubt there is any baseball GM who denies the importance of advanced stats. They provide much information, but how much emphasis they get likely depends on the question being asked.

[Brisco]

Brisco, every point you make about the hubris of stats geeks can be flipped around to reflect the hubris of scouts.

"Who cares if a player is only hitting .220? He's fast, and that *must* create havoc on the basepaths."

"So what if your veteran third baseman only hit 3 home runs last season. He brings presence to the dugout."

"I don't care about Juan Castro's range factor; he looks smooth when he fields a ground ball, so he must be a great defender. And I know, because I played two years of NCAA D-1A ball."

The problem doesn't stem from one side or the other. The problem you're describing is the result of everyone having an opinion and insisting that their opinion is *right*.

*IN MY OPINION*

Johnny has hit the nail on the head for the stats vs eyes/intangibles conflict, albeit possibly unintentionally.

His presented the example of: "So what if your veteran third baseman only hit 3 home runs last season. He brings presence to the dugout." to show how folks should not use the intangible argument to totally ignore stats.

Of course we have to use both stats and intangibles. However, this begs the question, how do we have a reasonable discussion/disagreement on a given player when some of the factors are objective and others are not?

let's assume the third baseman IS a tremendous clubhouse leader, how much slack do we cut him in the stat arena? So 3 home runs is not enough... is thirteen? Why or why not? How many "wins", "runs", or OPS points is great leadership worth?

What about a true clubhouse cancer? (Imagine the Parkman character in Major League II - not a great movie, but the a great example for discussion of this point) How high would Parkman's OPS have to be to outweigh the negative effect on esprit de corps and team morale?

[jojo]

I reject the characterization that because of stats, "suddenly all of us are baseball experts".

If anything sabermetrics have dramatically increased the knowledge of common but passionate baseball fans-thats a great thing. But rewind the time machine dial to before James. What was the typical baseball debate like? Not much different...and most certainly no one was prefacing their strongly expressed opinion with "based upon my experience in little league thirty years ago"... The guy on the less compelling end of things who would be prone to call people names on the internet was still prone to call the other guy names if he thought he could get away with it.

Sabermetrics have just allowed strawmen to enter the argument but sabermetrics have also allowed a much deeper understanding of the game and frankly have been a boon to anyone passionate about the game but who "did not play the game".

Really this isn't about stats versus eyes. If a person can't recognize the difference between a slider and four seam fastball and that same person argues that Votto has lost it after a week of hitting under .100, i'd argue that their opinion is neither a referendum on sabermetrics nor the eyes. It's a referendum on argumentation.

Really this isn't about stats versus eyes-it's about critically thinking and formulating a rational argument neither of which is dependent upon whether a person favors a statistical or a traditional approach to trying to better understand the game. And in the internet age, it's mostly about being bothered to take the time to do it.

[M2]

Coaching the mechanics of the game ideally ought to work hand-in-glove with the use of statistical data. Stats can show you that a player is struggling to hit inside fastballs, but they can't tell you what he needs to change mechanically in order to catch up to that pitch.

I use the same approach with youth soccer. I'm coaching technique and tactics all the time, but I have someone tape our games and I stat them so I know (rather than have to guess) how many scoring chances we're creating and which players are holding possession better than others. No matter how good your eye for the game is, there are things you can know as an absolute if you use stats ... and it's just part of good coaching to know those absolutes.

[RollyInRaleigh]

Coaching the mechanics of the game ideally ought to work hand-in-glove with the use of statistical data. Stats can show you that a player is struggling to hit inside fastballs, but they can't tell you what he needs to change mechanically in order to catch up to that pitch.

No matter how good your eye for the game is, there are things you can know as an absolute if you use stats ... and it's just part of good coaching to know those absolutes.

Been taking that approach for years.