Christian Vazquez

[notin]

I do. cover to cover. Win or lose. Rain or snow. Spring training or regular season. And they should forget about Swihart as a catcher. He's not a catcher. He was awful. He played well as in LF last year until he was hurt. It was not necessary to catch every game to see that. A .285 switch hitter in LF? Not bad, I'd say.

 

So you're saying you watch every game played by every team? All 2,430 regular season games every year?

[notin]

And they should forget about Swihart as a catcher. He's not a catcher. He was awful. He played well as in LF last year until he was hurt. It was not necessary to catch every game to see that. A .285 switch hitter in LF? Not bad, I'd say.

 

 

Last spring, you were citing Vazquez as being among the best defensive catcher you ever and how he needed to be handed the starting job immediately. I asked one simple question - "what if he hits .240?" You're answer wasn't "That's OK. We can live with that," Your answer was "He will never hit .240," which as an odd comment because the last time we saw him in MLB, he hit exactly .240. Not to mention, any player is capable of hitting .240, so I'm not sure why anyone would think a .240 hitter would be immune to it.

 

With your obvious Vazquez favoritisms and denials in mind (which isn't a sin), why am I supposed to buy into Swihart "never being able to play catcher" campaign?

 

So,

[moonslav59]

I won't disagree with that.. it really all depends on the return for each player

I always say Swihart because it's obvious he could probably get the biggest return

 

He gets the biggest return, because he has the biggest upside, but the upside value has value to us as well...assuming we trade Leon or Vaz instead.

[S5Dewey]

Moon said: Plus who among us watches every single mlb game?

 

The people who put together the UZR stat. Other than them, not really sure...

 

Not according to what I read here. I don't personally know anyone who develops the UZR stat but this is the first time I've read that the same person who puts together the UZR stat watches all the games. I would find it impossible that one person watches all (averaging 3+ hours pers game) and still has time to enter all that info into a computer, and then post the results on line.

 

What I've read in the past is that there is a cadre of people who watch the games and make these decisions. Since no two people see things exactly the same way that means that there is some subjectivity involved in it. I made half a career of training people to do things that could be quantified objectively and even then there were nuances in their testing procedures that made them less than 100% perfect.

 

If someone wants to see UZR as the be-all, end-all of range then have at it! I'm going to continue to roll my eyes every time I see it knowing that there is an element of subjectivity in it. I'm not saying that there's no value in UZR, but it ain't all it's cracked up to be by the stat geeks.

[notin]

Moon said: Plus who among us watches every single mlb game?

 

 

 

Not according to what I read here. I don't personally know anyone who develops the UZR stat but this is the first time I've read that the same person who puts together the UZR stat watches all the games. I would find it impossible that one person watches all (averaging 3+ hours pers game) and still has time to enter all that info into a computer, and then post the results on line.

 

What I've read in the past is that there is a cadre of people who watch the games and make these decisions. Since no two people see things exactly the same way that means that there is some subjectivity involved in it. I made half a career of training people to do things that could be quantified objectively and even then there were nuances in their testing procedures that made them less than 100% perfect.

 

If someone wants to see UZR as the be-all, end-all of range then have at it! I'm going to continue to roll my eyes every time I see it knowing that there is an element of subjectivity in it. I'm not saying that there's no value in UZR, but it ain't all it's cracked up to be by the stat geeks.

 

That has got to be the WORST criticism of UZR ever. It's subjective because multiple people are involved? Really?

 

As opposed to the "Eye Test" where dome fan who thinks he knows everything about defense but in reality has less knowledge than he realizes and no established baseline instead watches one team play 162 games and all the other teams 6 or 18 times, decides that is equal, and concludes "Hey Player X made more great plays by my proprietary definition of 'great' than any other shortstop so therefore he id the best defensive one in MLB!"

 

No one ever said UZR was the end-all, be-all. But I will positively state it is far more accurate than any individual "eye test"...

[S5Dewey]

That has got to be the WORST criticism of UZR ever. It's subjective because multiple people are involved? Really?

 

As opposed to the "Eye Test" where dome fan who thinks he knows everything about defense but in reality has less knowledge than he realizes and no established baseline instead watches one team play 162 games and all the other teams 6 or 18 times, decides that is equal, and concludes "Hey Player X made more great plays by my proprietary definition of 'great' than any other shortstop so therefore he id the best defensive one in MLB!"

 

No one ever said UZR was the end-all, be-all. But I will positively state it is far more accurate than any individual "eye test"...

 

And that has to be the worst criticism of a post, ever.

 

The highlighted part has to be one of the most egregious examples of a straw man I ever saw. I did not say that, so please don't try to hold me to it. My point was and remains that the more people who are assessing a player's range the more subjective it is, and the more subjective it is the less reliable it is.

[Bellhorn04]

And that has to be the worst criticism of a post, ever.

 

The highlighted part has to be one of the most egregious examples of a straw man I ever saw. I did not say that, so please don't try to hold me to it. My point was and remains that the more people who are assessing a player's range the more subjective it is, and the more subjective it is the less reliable it is.

 

Wouldn't it be less subjective the more people are involved? Nothing can be more subjective than the opinion of one person.

[notin]

And that has to be the worst criticism of a post, ever.

 

The highlighted part has to be one of the most egregious examples of a straw man I ever saw. I did not say that, so please don't try to hold me to it. My point was and remains that the more people who are assessing a player's range the more subjective it is, and the more subjective it is the less reliable it is.

 

That's not a straw man. That's a paraphrase of the pro-"eye test" argument, with some intentional sarcasm thrown in. Really it is how many use the eye test, myself included. To say UZR is even remotely equally subjective to the eye test is an error of unfathomable ignorance and arrogance. It simply lacks the data content and any sort of baseline, both of which are the focal points of UZR. Just because UZR isn't perfect isn't the condemnation many want it to be, and its imperfections do not necessarily level the playing field with other defensive metrics, especially the eye test. Sort of like how while air travel isn't perfect - flights get delayed, baggage lost, etc. - that doesn't mean it becomes equal to walking...

[notin]

Wouldn't it be less subjective the more people are involved? Nothing can be more subjective than the opinion of one person.

 

This is also true, especially when all weigh their opinion against a standard or established baseline. ..

[Kimmi]

Moon said: Plus who among us watches every single mlb game?

 

 

 

Not according to what I read here. I don't personally know anyone who develops the UZR stat but this is the first time I've read that the same person who puts together the UZR stat watches all the games. I would find it impossible that one person watches all (averaging 3+ hours pers game) and still has time to enter all that info into a computer, and then post the results on line.

 

What I've read in the past is that there is a cadre of people who watch the games and make these decisions. Since no two people see things exactly the same way that means that there is some subjectivity involved in it. I made half a career of training people to do things that could be quantified objectively and even then there were nuances in their testing procedures that made them less than 100% perfect.

 

If someone wants to see UZR as the be-all, end-all of range then have at it! I'm going to continue to roll my eyes every time I see it knowing that there is an element of subjectivity in it. I'm not saying that there's no value in UZR, but it ain't all it's cracked up to be by the stat geeks.

 

The scouts that are used go through some pretty intense training. They are then tested before they're allowed to become a scout. I'm not sure exactly what is involved in the testing, but I've read that many trainees do not make it past the testing phase. I'm guessing that part of the testing is to make sure the scouts 'see' what actually takes place during a play. Many of the scouts used are former baseball players, coaches, etc., people who already have trained eyes.

 

Each play is scouted by at least 2 different scouts independently and they must come to a reasonable agreement on how to 'grade' the play. If I recall correctly, if the reports from the scouts differ on any play, then it is viewed by more people who then come to a consensus. Also, the scouts are rotated regularly so that one scout is not watching the same player or the same team over and over again.

 

All of these measures eliminate the subjectivity to a good extent.

 

I don't think anyone sees UZR as the be all, end all, but it is a valuable tool that I believe has much more merit than you give it credit for.

[moonslav59]

uZR is a very valuable tool. It's much better than any single fan watching mostly one team's games and fielding % and RF9.

[S5Dewey]

That's not a straw man. That's a paraphrase of the pro-"eye test" argument, with some intentional sarcasm thrown in. Really it is how many use the eye test, myself included. To say UZR is even remotely equally subjective to the eye test is an error of unfathomable ignorance and arrogance. It simply lacks the data content and any sort of baseline, both of which are the focal points of UZR. Just because UZR isn't perfect isn't the condemnation many want it to be, and its imperfections do not necessarily level the playing field with other defensive metrics, especially the eye test. Sort of like how while air travel isn't perfect - flights get delayed, baggage lost, etc. - that doesn't mean it becomes equal to walking...

 

I will defer to the poster who knows more about arrogance than anyone here.

[moonslav59]

I will defer to the poster who knows more about arrogance than anyone here.

 

I don't recall 700 weighing in on this issue.

 

 

 

(Did I just say that out loud? ......Oooops!)

 

[Kimmi]

uZR is a very valuable tool. It's much better than any single fan watching mostly one team's games and fielding % and RF9.

 

Agree 100%.

 

The best way to assess any player will always be to use as many tools available as you can, from both the stats side and the scouting side. However, if I had to choose only one tool, I would pick UZR. And I would feel pretty good about what it tells me.

[Kimmi]

I will defer to the poster who knows more about arrogance than anyone here.

 

S5, here is a timely article written about a week ago on the very topic. You are far from being alone in your opinion of advanced defensive metrics. I disagree with much of what the players are saying, but many people feel the same way that you do.

 

On Players’ and Coaches’ Skepticism of Defensive Metrics

[S5Dewey]

S5, here is a timely article written about a week ago on the very topic. You are far from being alone in your opinion of advanced defensive metrics. I disagree with much of what the players are saying, but many people feel the same way that you do.

 

On Players’ and Coaches’ Skepticism of Defensive Metrics

 

Thanks, Kimmy. It's an interesting read for me and I now don't feel quite so alone in the wilderness.

 

And BTW, your posting it in spite of much of it being in conflict with what you seem to believe was a...er... classy act.

[Beantowne]

it's fairly common knowledge that "defensive metrics" UZR in particular has many flaws and that the process of evaualitng a player or teams defensive proficiency are still "works in progress". As stated above defensice metric all clearly, have value but not unlike war for example the formula is way to complicted for the average fan to wrap thier brains around (present company included).

 

As for the eye test (My prefered method), evaluating a players defensive ability is fairly simple. A good defensive player "makes the routine play, look routine". A gold glover makes the tough play look routine. Over simplification, yes..

 

End of the day it's not rocket science although the sabre boys try thier best to make it appear to be...

[Bellhorn04]

A perfect example of why the advanced defensive metrics matter was Jeter. Jeter was always a sure-handed fielder who made the plays on the balls he got to. But as he got older the number of balls he got to were less and less. If you only went by errors his numbers still looked great, but in truth he was a below-average defender because of all the balls getting past him untouched.

[moonslav59]

Agree 100%.

 

The best way to assess any player will always be to use as many tools available as you can, from both the stats side and the scouting side. However, if I had to choose only one tool, I would pick UZR. And I would feel pretty good about what it tells me.

 

Agreed. I think UZR 150 is better than any single person's opinion based on the eye test and old school use of FLG% and RF/9.

 

Personally, I like UZR/150 a lot, but only in large sample sizes that often need to be more than a full season. I also consult the Fielding Bible, DRS and Inside Edge data. The GG vote is a joke.

 

FLDG% is not even accurate as home team scorers often make very questionable choices-- probably more so than the UZR data collectors.

 

RF/9 can be easily influenced by how often a player's pitching staff allows GB, Fly balls or Ks.

[moonslav59]

A perfect example of why the advanced defensive metrics matter was Jeter. Jeter was always a sure-handed fielder who made the plays on the balls he got to. But as he got older the number of balls he got to were less and less. If you only went by errors his numbers still looked great, but in truth he was a below-average defender because of all the balls getting past him untouched.

 

Jeter went from being near average to being one of the worst defensive SSs in his time.

[Kimmi]

Thanks, Kimmy. It's an interesting read for me and I now don't feel quite so alone in the wilderness.

 

And BTW, your posting it in spite of much of it being in conflict with what you seem to believe was a...er... classy act.

 

I do try to be fair in what I post. I am not the least bit surprised to read that many players have no use for advanced stats.

[Kimmi]

it's fairly common knowledge that "defensive metrics" UZR in particular has many flaws and that the process of evaualitng a player or teams defensive proficiency are still "works in progress". As stated above defensice metric all clearly, have value but not unlike war for example the formula is way to complicted for the average fan to wrap thier brains around (present company included).

 

As for the eye test (My prefered method), evaluating a players defensive ability is fairly simple. A good defensive player "makes the routine play, look routine". A gold glover makes the tough play look routine. Over simplification, yes..

 

End of the day it's not rocket science although the sabre boys try thier best to make it appear to be...

 

The math behind UZR and WAR may be complicated, but I don't think the underlying idea behind what they are trying to measure is. The concept of UZR is fairly straightforward. I don't think the fact that the formula is complicated should take away from its usefulness or validity.

[Kimmi]

Agreed. I think UZR 150 is better than any single person's opinion based on the eye test and old school use of FLG% and RF/9.

 

Personally, I like UZR/150 a lot, but only in large sample sizes that often need to be more than a full season. I also consult the Fielding Bible, DRS and Inside Edge data. The GG vote is a joke.

 

FLDG% is not even accurate as home team scorers often make very questionable choices-- probably more so than the UZR data collectors.

 

RF/9 can be easily influenced by how often a player's pitching staff allows GB, Fly balls or Ks.

 

Agreed. I would guess that home team scorers are a lot more biased than the UZR data collectors are.

 

I believe that UZR does a good job with its single season data in describing what happened on the field that season. However, you do need more than a full season's worth of data to get a good feel for how good a defender really is.

 

It's similar to a batter who goes 0 for 20. His batting average over that span would be .000, which is an accurate account of what happened over that span. However, it is not an accurate assessment of the batter's true ability. For that, you need more data.

[moonslav59]

Agreed. I would guess that home team scorers are a lot more biased than the UZR data collectors are.

 

I believe that UZR does a good job with its single season data in describing what happened on the field that season. However, you do need more than a full season's worth of data to get a good feel for how good a defender really is.

 

It's similar to a batter who goes 0 for 20. His batting average over that span would be .000, which is an accurate account of what happened over that span. However, it is not an accurate assessment of the batter's true ability. For that, you need more data.

Some positions, like RF, get way less action than a position like SS. So, a season sample size for a SS is more telling than just a season from an OF'er.

[S5Dewey]

 

It's similar to a batter who goes 0 for 20. His batting average over that span would be .000, which is an accurate account of what happened over that span. However, it is not an accurate assessment of the batter's true ability. For that, you need more data.

 

This is the sort of discussion I used to get into with our old pal Pumpsie. Sometimes a recent sample size is more valuable than a larger one. If a hitter is hitting .333 for the season but is in an 0-20 slump IMO his chances of getting a hit in his next at bat are less than 1 in 3.

[moonslav59]

This is the sort of discussion I used to get into with our old pal Pumpsie. Sometimes a recent sample size is more valuable than a larger one. If a hitter is hitting .333 for the season but is in an 0-20 slump IMO his chances of getting a hit in his next at bat are less than 1 in 3.

 

I disagree.

 

What's the odds on this example player's projected next 100 or 1000 PAs?

 

.000

 

.333

 

.167

 

.310?

 

I'd project closer to .333 than anything else, unless the player is old or the original sample size was not large enough, or a long trend was under way before the 0-20 slump.

[FredLynn]

I disagree.

 

What's the odds on this example player's projected next 100 or 1000 PAs?

 

.000

 

.333

 

.167

 

.310?

 

I'd project closer to .333 than anything else, unless the player is old or the original sample size was not large enough, or a long trend was under way before the 0-20 slump.

 

You answered a different question that the one that was asked. I agree with you that the odds of him hitting close to his average in the next 100 PAs is the best guess, but for a single AB, for a player in an 0 for 20 hole, I think its far less. That player is not going well, for whatever reason. Maybe his mechanics are screwed up. He certainly would lack his usual confidence. I don't know that the chances of him getting a hit are, but I would not bet its .333.

[Hugh2]

I disagree.

 

What's the odds on this example player's projected next 100 or 1000 PAs?

 

.000

 

.333

 

.167

 

.310?

 

I'd project closer to .333 than anything else, unless the player is old or the original sample size was not large enough, or a long trend was under way before the 0-20 slump.

 

They did a study once where they took players in a slump and measured the probability of them getting a hit in their next at bat. Surprisingly the results were eerily close to a players career average.

 

I'm not disagreeing with the mental side of the game but at least to some extent slumps are just randomness.

 

Now this seems to disagree with what we see when you can really tell a guy is pressing at the plate but how often do we see guys who seem to be hitting rockets right people and others making weak contact that seems to find a hole?

 

But again even considering the mental aspect (which I do consider to be real) I'd say that the study proves a player is primed to break out of his slump at any moment. You shouldn't bet red at the roulette table just because it hit red 5 times in a row. The odds are exactly the same every time.

[Bellhorn04]

You answered a different question that the one that was asked. I agree with you that the odds of him hitting close to his average in the next 100 PAs is the best guess, but for a single AB, for a player in an 0 for 20 hole, I think its far less. That player is not going well, for whatever reason. Maybe his mechanics are screwed up. He certainly would lack his usual confidence. I don't know that the chances of him getting a hit are, but I would not bet its .333.

 

I think that's probably right. When a really good hitter finally does break out of a slump like that, though, they tend to go on a tear. This is what happened with Anthony Rizzo in the playoffs. He got off to a horrible start, in the first 7 games he hit 077/200/077 (2 for 26). In the last 10 games he hit 410/489/769.

[S5Dewey]

I disagree.

 

What's the odds on this example player's projected next 100 or 1000 PAs?

 

.000

 

.333

 

.167

 

.310?

 

I'd project closer to .333 than anything else, unless the player is old or the original sample size was not large enough, or a long trend was under way before the 0-20 slump.

 

I'm of the opinion that a player's previous 20 AB's are a better indicator of what he will do at his next AB than are his AB's two months ago. I always have to chuckle when someone says "He's due". All that means is that he hasn't had a hit in a while but that doesn't necessarily mean he's "due" for a hit now. It doesn't work that way.