Sabermetrics, My Take

[Stotle]

That's surprising to me. Thanks for all the info. Really appreciate it.

I should be clear, I'm only talking about the field level guys. I'm sure there are special assignment scouts/special assistants who do deeper on the analytics in their evaluations. Those are guys with voices at the decision-making table.

[CA-ORIOLE]

I should be clear, I'm only talking about the field level guys. I'm sure there are special assignment scouts/special assistants who do deeper on the analytics in their evaluations. Those are guys with voices at the decision-making table.

Ok, thanks.

[brianod]

I think it is a safe bet that neither of us fully understand the proprietary analytics teams use.

The "free versions" that we have is probably a garment cut from a baser cloth.

Fine, so if all those posting the free versions of sabermetrics as proof of how we should build the Orioles will admit that they might have holes in their collective theory, I will stand down and trust the more advanced metrics:)

[Can_of_corn]

Fine, so if all those posting the free versions of sabermetrics as proof of how we should build the Orioles will admit that they might have holes in their collective theory, I will stand down and trust the more advanced metrics:)

Once again no one on the stat side is actually claiming they are infallible. That is a false narrative propagated by others.

[Reboulet'sStache]

Are there really people out there that argue it's all or nothing?

[24fps]

Are there really people out there that argue it's all or nothing?

There are people who post here for whom that's implicit as they defend their bias.

[SteveA]

Once again no one on the stat side is actually claiming they are infallible. That is a false narrative propagated by others.

No but it is often implied, and I think that is what the non-stat oriented folks are sensitive to.

Say there is a discussion going on about a trade from a couple years ago, with people taking different sides, agreeing, disagreeing, going back and forth with opinions.

It's not uncommon for someone to step in and say something like "the players they got were worth over 7 WAR while the guys they gave up have been with just 4 WAR. Its obviously a good deal".

To the stat guy, he just feels like he has made a good, compelling argument that is backed by facts.

To the "non-stats" guy, it feels like he is being told: "There have been 20 posts in this thread discussing the pros and con's of the deal but I am coming down from the mountain and decreeing the issue settled".

[CA-ORIOLE]

No but it is often implied, and I think that is what the non-stat oriented folks are sensitive to.

Say there is a discussion going on about a trade from a couple years ago, with people taking different sides, agreeing, disagreeing, going back and forth with opinions.

It's not uncommon for someone to step in and say something like "the players they got were worth over 7 WAR while the guys they gave up have been with just 4 WAR. Its obviously a good deal".

To the stat guy, he just feels like he has made a good, compelling argument that is backed by facts.

To the "non-stats" guy, it feels like he is being told: "There have been 20 posts in this thread discussing the pros and con's of the deal but I am coming down from the mountain and decreeing the issue settled".

I consider myself well versed in statistics and argue with other "stat guys" all the time. That said, I understand your point that a lot of people are uncomfortable arguing statistics and it can be intimidating.

[Malike]

I find the "eye test" "I played the game" crowd to be as smug and rather more condescending then the "saber" crowd.

You won't find any "saber" guys saying that analytics is all you need but you will find plenty of examples of the "old school" crowd claiming that is what the "stat guys" say.

You are also a lot less likely to find the "saber" guy discounting overwhelming evidence.

Also, can you please use paragraph breaks in the future? I know the odd formatting isn't your fault but it, combined with the lack of breaks, makes following your post harder then it should be.

Have a pleasant night.

I don't disagree with any of this, but I tend to believe the real snobs and smugness come from message board posters, not people who are actually getting paid for a living by a MLB franchise.

[weams]

I respect that but I don't agree that it never happens. If we went by the sabermetrics for Paredes, we would have cut him.

Do you know that? I think his bat speed got him the chance. And his exit velocity.

[Can_of_corn]

No but it is often implied, and I think that is what the non-stat oriented folks are sensitive to.

Say there is a discussion going on about a trade from a couple years ago, with people taking different sides, agreeing, disagreeing, going back and forth with opinions.

It's not uncommon for someone to step in and say something like "the players they got were worth over 7 WAR while the guys they gave up have been with just 4 WAR. Its obviously a good deal".

To the stat guy, he just feels like he has made a good, compelling argument that is backed by facts.

To the "non-stats" guy, it feels like he is being told: "There have been 20 posts in this thread discussing the pros and con's of the deal but I am coming down from the mountain and decreeing the issue settled".

You bring up WAR, how can anyone think WAR is infallible when there are two distinct flavors of WAR? Which one is perfect? I was discussing De Aza yesterday, fangraphs rates him rather higher then bbref. How can anyone think they are perfect when they don't agree?

[Moose Milligan]

Honestly, I'm tired of sabermetrics. I don't mean that in a "I hate sabermetrics" way, because I don't, I understand their merits and why they're important. I'm just tired of them. Tired of learning whatever latest whippy dippy potato chippy metric comes down the pipe and why it's the new hot thing and why your argument is better than mine because you perceive your acronym to be better than mine (or vice versa). I understand that people get enjoyment out of them, and that's fine. Also, I spent the majority of my scholastic career avoiding anything/everything that has to do with math or calculations past what it takes to balance a checkbook. I understand the importance for a balanced education but I'm pretty sure I would have told you in the 5th grade I didn't want to be a scientist or a math whiz or have anything to do with math. Same in the 11th grade. Adding letters to math equations? Go screw yourself. No difference now. Only math I need is to calculate a commission check, which I can sometimes do in my head but would prefer to use a calculator for the trickier ones. Fat load of good that long division did for me in 4th grade, you complete #$%&@ Mrs. Lanpher. But I digress.

However, if my job depended on it like a front office type or a scout, I'd be more inclined to research them heavily and rely on them for what my eyes can't tell me. I understand their merits, but I just flat out don't enjoy them like others do. If that makes me less informed, so be it. I'll stick with the on base percentage and OPS. Keep your xFIPs and BABIPS, I'm tired of them.

[DrungoHazewood]

You bring up WAR, how can anyone think WAR is infallible when there are two distinct flavors of WAR? Which one is perfect? I was discussing De Aza yesterday, fangraphs rates him rather higher then bbref. How can anyone think they are perfect when they don't agree?

The whole idea that it's perfect or imperfect comes from a basic misunderstanding. All of these metrics have either an explicit or implicit uncertainty associated. If you're not taking into account that a projection or a WAR value is really +/- some amount you're completely off base. All of this is an approximation, our best look at the available data. All measurements have errors. When you get on the scale in the morning it might say you weigh 175, then tomorrow it says 177 and then five minutes later it says 176. The scale isn't broken, and it's still very valuable in assessing weight. But it's not going to get you repeatable 1/10th of a pound accuracy, and it probably doesn't need to for any real-life application.

[DrungoHazewood]

Also, I spent the majority of my scholastic career avoiding anything/everything that has to do with math or calculations past what it takes to balance a checkbook. I understand the importance for a balanced education but I'm pretty sure I would have told you in the 5th grade I didn't want to be a scientist or a math whiz or have anything to do with math. Same in the 11th grade. Adding letters to math equations? Go screw yourself. No difference now. Only math I need is to calculate a commission check, which I can sometimes do in my head but would prefer to use a calculator for the trickier ones. Fat load of good that long division did for me in 4th grade, you complete #$%&@ Mrs. Lanpher. But I digress.

As an engineer I... am... having... to... put a bunch of filters on my keyboard to keep from going over some line... I know you don't mean anything offensive here. Really. Not everyone is a math and science person. I get that. But much of the modern world wouldn't exist without advanced math and science. And I can't get enough of that stuff. My light reading in the evening is often articles from Slashdot, and the IEEE Spectrum magazine, and sometimes some pretty technical stuff about cars or airplanes or machines... that's the coolest stuff in the world. And it carries over to understanding baseball.

[mdbdotcom]

Honestly, I don't know a damn thing. I don't have the training to see what a scout sees or the knowledge to understand most advanced metrics. I find that I don't need either to enjoy the game and appreciate its intricacies at a level of detail that would make 80% of the general population that watches a couple of games a month to just shake their heads.

And yet, even at this level of profound ignorance, I dare to believe that my opinion is worth mentioning.

In recent years I've started trying to understand some of the advanced metrics a little bit. I consider the fairly simple ones I look at to be a snapshot of a player (or, of one aspect of the player). By snapshot, I mean an image that shows you what's there -- kind of like your old point and shoot mass market camera -- but not to the precision you'd get from a highly refined camera in the hands of a master. Most likely, the tools the teams use are closer to that.

So, I might mention that someone's rWAR over the past four years has been X. I would never assume that the same player's rWAR for the past three weeks means anything (I hope I got that right).

I do find it frustrating when mentioning someone's apparent defensive shortcomings and someone else posts, "But DRS says he is a better defender than Player Y," as if that's supposed to mean a lot. Is that for this season? For his entire career, including the two years he played with a bad leg? Is that the only defensive metric that matters (I suspect not).

For information to have value it must be put in context.

All these techniques for evaluating players are just tools that are available to us. It's good that they are all brought up. Ignoring any one of them and relying totally on another seems foolish to me. I can't believe good organizations do that today.

And here's the most (only) important thing I have to say: double-header today. Let's win two.