Izturis a Gold Glove snub???

[Lucky Jim]

I believe there were many threads early in the year about both of these guys not looking as good in the field as we are used to seeing.

It's just silly. Around the same time that I mentioned that I was disappointed with Nick and AJ's defense, I said that paying attention to Izturis's UZR was wrong-headed because it was non-representative due to the small sample.

If Shack's theory is true, why would I have not been down on Izturis, too?

[Lucky Jim]

Oh, I agree with that. But that's way different than saying they were below-average OF'ers.

Who's an average RF?

[DrungoHazewood]

As for how many games, I watched most of them, probably 90% of them, and in each and every one of those games I saw a minimum of 3 opposing OF'ers. As for how do you tell, well, you tell by watching lots of baseball.

What channel do you watch to get a read on each outfielder's first step quickness? Or reaction time? Or positioning? Or how the wind effects the ball? Unless you're sitting at the game, concentrating on a particular player, I don't see how you can get much of the information you need to subjectively judge how well someone plays defense.

[RShack]

I'll trust the goofy mirrors that at least try to be systematic and objective over the others that don't.

When it comes to stats, especially the still-rapidly-evolving D-stats, it is unwise to either trust them or dismiss them just because you support the use of stats to reveal things. Admiring the efforts of stat people and supporting their work is way different than arbitrarily trusting the results of their current prototypes.

[Lucky Jim]

When it comes to stats, especially the still-rapidly-evolving D-stats, it is unwise to either trust them or dismiss them just because you support the use of stats to reveal things. Admiring the efforts of stat people and supporting their work is way different than arbitrarily trusting the results of their current prototypes.

And I've sat here and said that I like the metrics because they track what I thought I was seeing. Your response was either "I don't believe you" or "you see things incorrectly."

That really should be the end of the argument, I guess.

[RShack]

Highly self-serving. Mackus and I have no reason to lie. But thanks.

Oh, I don't think you're lying. I think you're just downplaying the influence the stats had on your judgment. In general, people are not good at knowing the determinants of their own opinions. People are demonstrably lousy at that.

Are you telling me that, absent stat-data, you would have concluded that the 2009 version of Nick was a below-average OF'er? Is that what you're saying?

[Mackus]

Oh, I agree with that. But that's way different than saying they were below-average OF'ers.

Its a matter of degree. How much worse were they than we're used to, and exactly how good are they compared to average usually.

I feel that both were easily identifiable as being worse than we're used to right from the start of the season. Jones was to a much higher degree than Markakis. So how much worse than average they were is certainly debatable. I don't believe its debatable that Jones was even an average CF last year. Just watching him compared to everyone else it was easy to tell he wasn't getting to balls over his head that most guys get to.

Markakis had some obvious struggles early on, but was pretty much back to the level we're all accustomed to by midseason, so he probably ended the year around average cumulatively, maybe a bit above, but I think he was also obviously below average early in the year. Not just below what he usually gives us, but below what you expect from any RF.

[Lucky Jim]

What channel do you watch to get a read on each outfielder's first step quickness? Or reaction time? Or positioning? Or how the wind effects the ball? Unless you're sitting at the game, concentrating on a particular player, I don't see how you can get much of the information you need to subjectively judge how well someone plays defense.

Exactly. At best you catch the last couple of steps toward a ball, and that's about it.

Oh, I don't think you're lying. I think you're just downplaying the influence the stats had on your judgment. In general, people are not good at knowing the determinants of their own opinions. People are demonstrably lousy at that.

Are you telling me that, absent stat-data, you would have concluded that the 2009 version of Nick was a below-average OF'er? Is that what you're saying?

I would have thought that he was a slightly below average RF, who showed marked improvement over the course of the year. I also don't think -5.5 in UZR is written in stone, over a full year, which means there's some wiggle. So UZR reflects my take, which was that Nick got better, but that he was still slightly below average.

[BaltimoreTerp]

But that's a different issue. The month-to-month view can reflect how they did throughout the season, in a way that compares the August-version of Nick to the April-version of Nick. But that says nothing about whether the 2009-version of Nick was a below-average OF'er. I can't see how anybody who watched the games could possibly conclude that the 2009-version of Nick was a below-average OF'er. I think that's a goofy conclusion, one that demonstrates that you shouldn't blindly trust a stat that says he was.

A stat that tells you that Nick was a below-average OF'er is reflecting reality like a fun-house mirror.

Is it a reflection? Yes, it is.

Is it based on reality? Yes, it is.

Is it an accurate reflection that gives a true representation of reality? Nope, it's not. It's a flawed reflection that bends and twists reality somehow, to the point of meaningful distortion.

This is why we have multiple modern metrics that disagree with each other. Different flawed reflections of the same reality.

Doesn't mean we don't need mirrors, it just means we shouldn't trust what they show us until we get them right.

We'll know when we get them right because, when we do, they will show the same thing and most everybody will agree that what they show is right.

We're not there yet. It's foolish to pretend we are when we aren't...

Now, just watch... somebody is gonna claim that I'm against using mirrors ;-)

Drungo already got it pretty good, but here is the biggest issue with what you are saying.

In this metaphor, there is no such thing as a "normal" mirror. Every mirror is distorting things in some fashion, so you really have very little idea of how you actually look. You can use a couple mirrors to make your best guess. You can try and create new mirrors that provide what is supposed to be a more accurate view. You can choose to ignore all mirrors and try to make you best guess out of the information you can see on your own. But no matter what you do, your view will be distorted.

Your best bet is to use as many different mirrors as possible to try and figure things out, both the ones you have and new ones that are created.

[Lucky Jim]

Drungo already got it pretty good, but here is the biggest issue with what you are saying.

In this metaphor, there is no such thing as a "normal" mirror. Every mirror is distorting things in some fashion, so you really have very little idea of how you actually look. You can use a couple mirrors to make your best guess. You can try and create new mirrors that provide what is supposed to be a more accurate view. You can choose to ignore all mirrors and try to make you best guess out of the information you can see on your own. But no matter what you do, your view will be distorted.

Your best bet is to use as many different mirrors as possible to try and figure things out, both the ones you have and new ones that are created.

Nicely put, DJ.

[TommyD4207]

Drungo already got it pretty good, but here is the biggest issue with what you are saying.

In this metaphor, there is no such thing as a "normal" mirror. Every mirror is distorting things in some fashion, so you really have very little idea of how you actually look. You can use a couple mirrors to make your best guess. You can try and create new mirrors that provide what is supposed to be a more accurate view. You can choose to ignore all mirrors and try to make you best guess out of the information you can see on your own. But no matter what you do, your view will be distorted.

Your best bet is to use as many different mirrors as possible to try and figure things out, both the ones you have and new ones that are created.

Uhh yeah, you basically just twisted around what rshack said and said the same thing.

That was the whole point of the fun house mirror exercise. He was showing that none of them show the real thing and that you come to conclusions based off of a combo of things. The whole damn crux of the mirror point was that there is no normal mirror, as you essentially repeated.

That was his point. Sorry that you didn't pick up on that.

[Lucky Jim]

Uhh yeah, you basically just twisted around what rshack said and said the same thing.

That was the whole point of the fun house mirror exercise. He was showing that none of them show the real thing and that you come to conclusions based off of a combo of things. The whole damn crux of the mirror point was that there is no normal mirror, as you essentially repeated.

That was his point. Sorry that you didn't pick up on that.

Except for that whole part where he disregards metrics when they clash with what he thinks he sees and accuses those who saw something else with being manipulated by statistics.

Throughout his entire conversation Shack has put a primacy on what he thinks he's seen, even though his viewing was limited to about 90% of Orioles games, and therefore saw only roughly 6% of all games played, during which the camera was on each position player a very small percentage of the time.

Clearly, Shack thinks that watching the game - and by that I mean following a single team, and therefore a single set of players, throughout a single season - is a more accurate "mirror" than the metrics. And he's expressed as much throughout this thread.

Sorry you didn't pick up on that.

Of the two sides here there is one which judged what it saw by their own observation, realized that this is a severely circumscribed way to make a judgment, and then went to the metrics to confirm it. And then there's the side that watched 90% of the Orioles' games, came to a conclusion which differered with the metrics and thinks, based on that, the metrics must be severely flawed.

I can't see how anybody who watched the games could possibly conclude that the 2009-version of Nick was a below-average OF'er. I think that's a goofy conclusion, one that demonstrates that you shouldn't blindly trust a stat that says he was.

Mind you, no one here has said that the statistics are enough, on their own. No one.

[RShack]

Except for that whole part where he disregards metrics when they clash with what he thinks he sees and accuses those who saw something else with being manipulated by statistics.

Throughout his entire conversation Shack has put a primacy on what he thinks he's seen, even though his viewing was limited to about 90% of Orioles games, and therefore saw only roughly 6% of all games played, during which the camera was on each position player a very small percentage of the time.

Clearly, Shack thinks that watching the game - and by that I mean following a single team, and therefore a single set of players, throughout a single season - is a more accurate "mirror" than the metrics. And he's expressed as much throughout this thread.

Sorry you didn't pick up on that.

Of the two sides here there is one which judged what it saw by their own observation, realized that this is a severely circumscribed way to make a judgment, and then went to the metrics to confirm it. And then there's the side that watched 90% of the Orioles' games, came to a conclusion which differered with the metrics and thinks, based on that, the metrics must be severely flawed.

Mind you, no one here has said that the statistics are enough, on their own. No one.

Nah... Tommy got it exactly right...

As for the specific thing about whether or not Nick is a below-average OF'er, I bet there's only a minuscule number of people on the entire planet who think that, and almost all of them are here in this thread. Apart from this thread, I bet there's very, very few people here who believe it. If you made that claim to people anywhere else (except maybe the far reaches of ideological stat-consumers), they'd laugh at you.

Too bad we didn't have iffy and evolving prototype D-stats when Brooks Robinson was around, or I bet somebody would claim that in Year-X of Brooks' prime, he was a below-average 3B-man too. Of course, I bet in 10 years, people will look back on any methodology that says Nick is a below average OF-er and shake their heads at the notion that people would have put faith in a methodology that did that. Just because stats can be good. that doesn't mean you don't need some reality testing. This is especially true of D-stats, which are based on a fundamentally different model than either P-stats or hitting-stats. The model is so new that it's not even fleshed out yet, we don't even know what that model is yet, except that it's way different than what both P-stats and O-stats are based on. It's entirely new territory. When the D-stats get right, they will agree with each other and with what people see. How anybody can really believe we're close to that now is beyond me.

[Lucky Jim]

Nah... Tommy got it exactly right...

I know this will come as an enormous surprise to you, but you're not the final arbiter of what's right or wrong. You've clearly placed primacy in a small sample of viewed games over two different metrics which both align with what the rest of us have seen with our eyes.

As 1970 noted, with great clarity, here:

Someone mentioned that we don't see enough of other players (or really even of an Orioles player - in terms of what he does on the entire play). It just happens that there is a group that sees every game of every player - the BIS Video Scouts. They felt like there were 10 RF better than Markakis in 2009. They ranked him as the best defensive RF in 2008 and second best in 2007, so it isn't as if they have some bias against him. It should also be noted that they aren't making this determination via stats. Last year every voter had him in the top 9. This year one had him 2nd, another 4th, another 5th, and another 10th. No, this doesn't prove he was below average on the whole for last season, but they do verify what many of us on here have said since at least May - he wasn't as good this year as he was last year.

Also somone mentioned these results as if they refuted TFB and/or UZR, but do they?

1B - Pujols (9th UZR / 1st TFB)

2B - Hill (12th / 3rd)

3B - Zimmerman (2nd / 2nd)

SS - Wilson (1st / 1st)

LF - Crawford (1st / 1st)

CF - Gutierrez (1st / 1st)

RF - Ichiro (3rd / 1st)

With the exception of the left side of the IF (for UZR, at least) the above seems to validate UZR and TFB. It is also plainly clear by the overall vote totals that those involved in this vote didn't blindly go by stats (despite the accusation).

Considering how well TFB fared in this vote what does this say about their -10 (runs) rating of him for this year?

TBF had him going from +10 on shallow balls in 08 down to -4 this year, from +4 to -4 in medium depth balls, and from -4 to -9 on deep balls. I don't remember exactly what LJ wrote, but I believe this backs him up to some extent.

No one is saying that Nick is a "below average RF". All anyone said - and Mackus has been most emphatic about this over the last month - was that he was below average in 2009, but above in 2007 and 2008...and not unlikely to rise to that level again.

Of course, Shack knows more from watching the O's than the entire BIS Video Scouting group.

Hilarious, btw, that you think the metrics don't "agree with what people see" when BIS is comprised of people watching every play. Did you mean, align with what you see?

[RShack]

I know this will come as an enormous surprise to you, but you're not the final arbiter of what's right or wrong. You've clearly placed primacy in a small sample of viewed games over two different metrics which both align with what the rest of us have seen with our eyes.

As 1970 noted, with great clarity, here:

No one is saying that Nick is a "below average RF". All anyone said - and Mackus has been most emphatic about this over the last month - was that he was below average in 2009, but above in 2007 and 2008...and not unlikely to rise to that level again.

Of course, Shack knows more from watching the O's than the entire BIS Video Scouting group.

Hilarious, btw, that you think the metrics don't "agree with what people see" when BIS is comprised of people watching every play. Did you mean, align with what you see?

Well, everybody else gets to be a final arbiter of their opinion, I don't see why I shouldn't get the same leeway.

As for the assertion that Nick-2009 is a below-average OF-er, I am quite sure that's a goofy thing to say. I don't think it would have even crossed anybody's mind, had it not been for what some are calling "evidence".

It is quite true that some methodologies rely on human observers, but I hope you're not implying that those observers concluded that Nick was a "below-average OF'er"... because they didn't conclude any such thing.