2B in 2010.

[manbearcub]

ignored what? the graphic? i don't see how that says anything about how accurate uzr is for specific single players. i'm not sure i even understand what the point of that graphic is. i never said uzr is always wrong or that there's no truth in it. i said that i don't put any stock into it because of how unreliable it is when you're trying rate players.

if one were to put some thought into it, one could extrapolate that since team UZR does, in fact, have a positive relationship with lowering ERA, that individual UZR could, in fact, have a similar relationship.

[A-Cal]

brian roberts?

[17 Seconds]

ignored what? the graphic? i don't see how that says anything about how accurate uzr is for specific single players. i'm not sure i even understand what the point of that graphic is. i never said uzr is always wrong or that there's no truth in it. i said that i don't put any stock into it because of how unreliable it is when you're trying rate players.

if one were to put some thought into it, one could extrapolate that since team UZR does, in fact, have a positive relationship with lowering ERA, that individual UZR could, in fact, have a similar relationship.

 

like i said, i understand that there is a lot of truth in uzr. i never said there wasn't. i'm saying it's unreliable and shouldn't be used to judge performance by a player and then act like it is definitive at all.

 

according to uzr, derek jeter has become one of the best defensive shortstops in the league the last 2 seasons after being horrible for 3 straight years prior to that. so are the numbers inaccurate, or has jeter suddenly become a great shortstop? i think the former seems more likely.

[Guest]

wait, so because i think he's a poor defender, that means i'm jaundiced? okay.

 

seriously, this is not even some kind of heretic opinion. i thought it was pretty common knowledge that he is a poor defender.

 

oh but some defensive metrics that will probably have him as horrible and miguel tejada as the best shortstop next season say he's good, so that's that.

 

Are you ever right about anything?

I agree with his opinion about theriot's defense if that means anything.

[Geech]

like i said, i understand that there is a lot of truth in uzr. i never said there wasn't. i'm saying it's unreliable and shouldn't be used to judge performance by a player and then act like it is definitive at all.

 

The point is that eyeballing a player's defense is an even worse idea. At least something like UZR or Plus/Minus is attempting to add some objectivity to the process of defensive metrics with their advanced analysis of batted ball data. You're right that they aren't perfect, but they're better than nothing, which is what you seem to be using.

[Geech]

I agree with his opinion about theriot's defense if that means anything.

 

Why is that?

[Guest]

I don't like stats that don't have an auto-correlation.

[17 Seconds]

like i said, i understand that there is a lot of truth in uzr. i never said there wasn't. i'm saying it's unreliable and shouldn't be used to judge performance by a player and then act like it is definitive at all.

 

The point is that eyeballing a player's defense is an even worse idea. At least something like UZR or Plus/Minus is attempting to add some objectivity to the process of defensive metrics with their advanced analysis of batted ball data. You're right that they aren't perfect, but they're better than nothing, which is what you seem to be using.

 

i'd hardly call watching 95% of a guy's plays at shortstop and then making an opinion "nothing". is it absolutely accurate? no. but calling it nothing and saying some unreliable statistic is superior is kind of ridiculous i think. it's not like i'm basing my opinion of his defense on a handful of plays or anything.

Edited September 6, 2009 by 17 Seconds

[Guest]

To be more verbose about the subject...

 

I don't think we should ignore any information about a player, whether that be scouting or stats. While defensive stats that I do not yet trust tell me that Theriot is an average or better SS, my eyes tell me that he is not. Given what I feel about defensive stats and their ability to trend from year to year versus my faith in my scouting ability (misplaced or not), I'll trust the latter to this point.

 

I hope that defensive stats evolve to the point where I no longer have to trust my vision.

[Guest]

Everyone feel free to rip on me equally, btw.

[A-Cal]

instead of making a one-liner joke, i will go ahead and jump into this fun

 

i also believe theriot is not a great defender, and i don't place much trust in defensive metrics either

 

there is nothing wrong with using your eyes, if it looks like a duck and quacks like a duck...well you know

[Geech]

i'd hardly call watching 95% of a guy's plays at shortstop and then making an opinion "nothing". is it absolutely accurate? no. but calling it nothing and saying some unreliable statistic is superior is kind of ridiculous i think. it's not like i'm basing my opinion of his defense on a handful of plays or anything.

 

What do you do for a living that allows you the luxury of watching 95% of Theriot's plays? I doubt even scouts see that much.

 

Anyway, Plus/Minus is video scouting in which someone watches all of the player's plays. The field is divided into a series of zones and each zone is analyzed to determine how often plays are made on different velocities and types of batted balls. Players are then penalized or given credit for making plays based on how often a similar play is made.

 

The results of Plus/Minus are derived from no less observation than you perform, and much more analysis. It's fine to be skeptical of the method when there are clear deficiencies or when similar systems produce widely different results, but it's hubris to completely reject them simply because they don't mesh with your observations. Assuming for the sake of argument that you're a rational human, even if you literally watched 95% of Theriot's plays your observations would still be fraught with bias and inaccuracy. That's just the nature of human observation and memory, and that's why having an objective approach is so important.

 

So, yes. Compared to plus/minus or UZR, you are basing on your opinion on virtually nothing.

 

I'm specifically responding to dexter, but this post can obviously apply to all three of you, BTW.

[A-Cal]

hmmm...i thought all opinions had a base, no matter how offset

[Geech]

I don't know what that is supposed to mean, but I have a strong suspicion that I don't care.

[manbearcub]

ignored what? the graphic? i don't see how that says anything about how accurate uzr is for specific single players. i'm not sure i even understand what the point of that graphic is. i never said uzr is always wrong or that there's no truth in it. i said that i don't put any stock into it because of how unreliable it is when you're trying rate players.

if one were to put some thought into it, one could extrapolate that since team UZR does, in fact, have a positive relationship with lowering ERA, that individual UZR could, in fact, have a similar relationship.

 

like i said, i understand that there is a lot of truth in uzr. i never said there wasn't. i'm saying it's unreliable and shouldn't be used to judge performance by a player and then act like it is definitive at all.

 

according to uzr, derek jeter has become one of the best defensive shortstops in the league the last 2 seasons after being horrible for 3 straight years prior to that. so are the numbers inaccurate, or has jeter suddenly become a great shortstop? i think the former seems more likely.

who knows? is there a reason Geovany Soto all of a sudden became a terrible hitter? he is seeing the same pitches as last year. or is defense something that doesnt change for a player year to year?

 

in any case the fact that you jump on UZR any time it is mentioned on this board undermines any pleas you attempt to ralionalize your position.

[TruffleShuffle]

uzr isn't the only defensive metric that has theriot right around average, and not just this year, but pretty much since he's become a starting shortstop.

[Shilzzz]

Hudson is very streaky and we all know how Cubs fans get with streaky players. He's only OPSing .782 this year.

 

Adam Kennedy has been decent this year for Oakland. I could probably stand a Kennedy Baker platoon.

[Guest]

i'd hardly call watching 95% of a guy's plays at shortstop and then making an opinion "nothing". is it absolutely accurate? no. but calling it nothing and saying some unreliable statistic is superior is kind of ridiculous i think. it's not like i'm basing my opinion of his defense on a handful of plays or anything.

 

What do you do for a living that allows you the luxury of watching 95% of Theriot's plays? I doubt even scouts see that much.

 

Anyway, Plus/Minus is video scouting in which someone watches all of the player's plays. The field is divided into a series of zones and each zone is analyzed to determine how often plays are made on different velocities and types of batted balls. Players are then penalized or given credit for making plays based on how often a similar play is made.

 

The results of Plus/Minus are derived from no less observation than you perform, and much more analysis. It's fine to be skeptical of the method when there are clear deficiencies or when similar systems produce widely different results, but it's hubris to completely reject them simply because they don't mesh with your observations. Assuming for the sake of argument that you're a rational human, even if you literally watched 95% of Theriot's plays your observations would still be fraught with bias and inaccuracy. That's just the nature of human observation and memory, and that's why having an objective approach is so important.

 

So, yes. Compared to plus/minus or UZR, you are basing on your opinion on virtually nothing.

 

I'm specifically responding to dexter, but this post can obviously apply to all three of you, BTW.

Plus/Minus rates how well a defender is initially positioned as much as it measures his range. Same thing with any zone rating scheme.

 

If you are watching the results of the plays for a player and trying to remember them all to form a judgment, you aren't scouting. You have to watch the process by which the player goes about playing to form the opinion. I will grant you that this method of determining defensive worth is fraught with difficulty. However, it doesn't have the problem of rating players good-awful-great-bad in consecutive years.

[rocket]

http://en.wikipedia.org/wiki/Linear_regression

 

i understand what the graph is.

 

i've seen you talk in the past about how bad jeter is defensively. so then why has jeter been in the top half of the league in uzr rankings the past 2 seasons?

 

because his defense has gotten better?

[Guest]

I have to say that the defensive metrics are pretty unison and consistent in their opinion of Theriot, aren't they?

[Guest]

http://en.wikipedia.org/wiki/Linear_regression

 

i understand what the graph is.

 

i've seen you talk in the past about how bad jeter is defensively. so then why has jeter been in the top half of the league in uzr rankings the past 2 seasons?

 

because his defense has gotten better?

That's possible, but it would be pretty unusual for a shortstop in his 30's to actually improve his defense. I don't have to tell you how much defense at that position typically decays with age.

 

Besides, unlike Theriot, the systems are all over the place on Jeter's defense.

[Geech]

Plus/Minus rates how well a defender is initially positioned as much as it measures his range. Same thing with any zone rating scheme.

 

If you are watching the results of the plays for a player and trying to remember them all to form a judgment, you aren't scouting. You have to watch the process by which the player goes about playing to form the opinion. I will grant you that this method of determining defensive worth is fraught with difficulty. However, it doesn't have the problem of rating players good-awful-great-bad in consecutive years.

 

It's not fraught with difficulty, it's fraught with impossibility. Here's the thing, even if we assume for the sake of argument that you can form an accurate judgment of a given players fielding, that's not enough. In order to be useful, you have to compare it with something. In order to rate Theriot's defense at shortstop as average, above average or below average, then by definition you must watch 95% of every shortstop's plays, form a judgment for each player and then make meaningful comparisons between them. Attempting to do this on your own, inside your own head, is ridiculous.

 

And while that process may lend itself to consistent evaluations of players from year to year, that is because those evaluations are consistently meaningless.

[radarluv1]

the mets radio guys were saying how much theriot looks like eckstein, especially with his throwing. i think that comparison has been made around here a few times, and is pretty accurate. he doesnt put up enough O to justify a move to 2B.

[Guest]

Plus/Minus rates how well a defender is initially positioned as much as it measures his range. Same thing with any zone rating scheme.

 

If you are watching the results of the plays for a player and trying to remember them all to form a judgment, you aren't scouting. You have to watch the process by which the player goes about playing to form the opinion. I will grant you that this method of determining defensive worth is fraught with difficulty. However, it doesn't have the problem of rating players good-awful-great-bad in consecutive years.

 

It's not fraught with difficulty, it's fraught with impossibility. Here's the thing, even if we assume for the sake of argument that you can form an accurate judgment of a given players fielding, that's not enough. In order to be useful, you have to compare it with something. In order to rate Theriot's defense at shortstop as average, above average or below average, then by definition you must watch 95% of every shortstop's plays, form a judgment for each player and then make meaningful comparisons between them. Attempting to do this on your own, inside your own head, is ridiculous.

 

And while that process may lend itself to consistent evaluations of players from year to year, that is because those evaluations are consistently meaningless.

Yes, you do have to do it as a comparison. No...you don't have to watch 95% of every shortstop's plays to be able to make judgments about a particular player's ability relative to the "typical" shortstop.

 

Honest question for you: Do you believe that scouting is completely worthless and that all evaluation in baseball should be done with statistics?

[Geech]

Honest question for you: Do you believe that scouting is completely worthless and that all evaluation in baseball should be done with statistics?

 

No, I don't. I think there are some areas of performance that are not measured well enough to derive useful statistics. Wherever possible, however, objectivity is desirable.

 

My turn: Why do you feel your haphazard observations and analysis are superior to UZR and Plus/Minus? Is it simply because you don't like the results or can you point to serious flaws in the methodology?