Semien

[Dick Allen]

Aren't eyes used for defensive metrics?

 

 

[witesoxfan]

RBI's and Wins have absolutely nothing to do with this conversation. Nice strawman, homie.

 

What are you talking about? It's a corollary. You stated that defense is judged with the eye test because they've been doing it for 100 years. Thus, if using your eyes for defensive quality is true because it's been used for 100 years, then using RBI and W must also be true because they've done it for 100 years. Please tell me why that is and why people don't adapt and use other mediums of information as they evolve and begin to better understand certain topics.

 

You can't say that "oh this is true because they've done it for 100 years but this is not true because they did it for 100 years."

[witesoxfan]

Aren't eyes used for defensive metrics?

 

To some extent, yes, they are.

[Dick Allen]

And if defensive metrics need 3 seasons to be accurate, doesn't that mean that WAR ultimately may not be accurate?

[TaylorStSox]

What are you talking about? It's a corollary. You stated that defense is judged with the eye test because they've been doing it for 100 years. Thus, if using your eyes for defensive quality is true because it's been used for 100 years, then using RBI and W must also be true because they've done it for 100 years. Please tell me why that is and why people don't adapt and use other mediums of information as they evolve and begin to better understand certain topics.

 

You can't say that "oh this is true because they've done it for 100 years but this is not true because they did it for 100 years."

You're using a false equivalency to try to justify your argument. I'm not talking about RBI's and W's. They have nothing to do with this conversation. You believe that it takes 3 years to judge defense. That's laughable. It takes 5 minutes with your eyes. You guys spend all this time trying to create perceived value for players and it's just stupid.

[NorthSideSox72]

People should not read anything into Beckham's rehab numbers, for a whole list of reasons. Same goes for Semien's MLB D. In both cases, sample size is waaaaaaaaaay to small.

 

[TaylorStSox]

People should not read anything into Beckham's rehab numbers, for a whole list of reasons. Same goes for Semien's MLB D. In both cases, sample size is waaaaaaaaaay to small.

Semien's defense is obvious, solid and average. He makes tough plays, but doesn't have much range and has an average arm. He's a lot like Beckham defensively.

[Balta1701]

Semien's defense is obvious, solid and average. He makes tough plays, but doesn't have much range and has an average arm. He's a lot like Beckham defensively.

Worth thinking about though...he hasn't been a full time 2b, so his ability to cover a lot of ground will be impacted by his ability to read the ball off the bat. With additional time at 2b, it's plausible he could improve at that. He is certainly a solid athlete and he has the speed to cover a good amount of range, so if he's not showing range yet that tells me he isn't getting the best jumps on ground balls.

 

With coaching and experience that could change.

[TaylorStSox]

Worth thinking about though...he hasn't been a full time 2b, so his ability to cover a lot of ground will be impacted by his ability to read the ball off the bat. With additional time at 2b, it's plausible he could improve at that. He is certainly a solid athlete and he has the speed to cover a good amount of range, so if he's not showing range yet that tells me he isn't getting the best jumps on ground balls.

 

With coaching and experience that could change.

His reactions are very good and it shows when he plays 3rd. With the right positioning, he can be a plus second baseman, but won't be spectacular. That's fine. It's Alexei's job to make the spectacular plays. Overall, I'm very happy with Semien. It's obvious that he just gets it. He's going to be a very good pro.

[witesoxfan]

You're using a false equivalency to try to justify your argument. I'm not talking about RBI's and W's. They have nothing to do with this conversation. You believe that it takes 3 years to judge defense. That's laughable. It takes 5 minutes with your eyes. You guys spend all this time trying to create perceived value for players and it's just stupid.

 

No, it takes 3 years to come up with some sort of predictive capabilities of UZR because it has a large variation, but with limited ability to actually view plays, many people believe that Derek Jeter is still a good defensive SS which is clearly not true. Statistics help us understand that further. You can tell how players are playing defensively right away, but sample size also has to come into effect as well.

 

You can say they are useless all you want, but they cover all their bases and help us understand how quickly and often players actually change defensively.

[TaylorStSox]

No, it takes 3 years to come up with some sort of predictive capabilities of UZR because it has a large variation, but with limited ability to actually view plays, many people believe that Derek Jeter is still a good defensive SS which is clearly not true. Statistics help us understand that further. You can tell how players are playing defensively right away, but sample size also has to come into effect as well.

 

You can say they are useless all you want, but they cover all their bases and help us understand how quickly and often players actually change defensively.

I'm surprised you actually believe this because it's absolute bulls***. Sorry.

[witesoxfan]

I'm surprised you actually believe this because it's absolute bulls***. Sorry.

 

Except that it's not, so I'm sorry you don't understand that.

[ILMOU]

And if defensive metrics need 3 seasons to be accurate, doesn't that mean that WAR ultimately may not be accurate?

To the extent that UZR factors into defensive WAR, this is true. The better metrics are coming and UZR will be on the pile with VHS and Beta.

[TaylorStSox]

Except that it's not, so I'm sorry you don't understand that.

Now the "you just don't get it" argument. You're on a roll today.

[lasttriptotulsa]

Except that it's not, so I'm sorry you don't understand that.

 

Got to love his attitude though. He doesn't like it or understand it, therefore it is wrong.

[witesoxfan]

Now the "you just don't get it" argument. You're on a roll today.

UZR uses Baseball Info Solutions (BIS) data in calculating its results. It’s important to note that this data is compiled by human scorers, which means that it likely includes some human error.

 

http://www.fangraphs.com/library/defense/uzr/

 

No, UZR doesn't use the eye test at all, other than the fact that it's determined by people who watch every single play. It's all based willy nilly on numbers.

 

So yeah, you're wrong on multiple levels.

[Jake]

I don't agree that they're "absolutely" meaningless, but like all stats they fail to take into account unquantifiable variables, like weather, field condition, the speed at which a ball is traveling, defensive positioning (good coaching) and minor injuries that probably contribute a lot more to defensive range than even a 3-year average statistic will ever be able to measure. Like you, I prefer the eye test when it comes to defense. But there must be some validity to UZR since it seems to be used more and more in professional evaluation.

 

Actually, they do try to take into account those unquantifiable variables. This is why...well, it uses the "eye test"

 

Aren't eyes used for defensive metrics?

 

Yes ... as wite so nicely tells us below

 

http://www.fangraphs.com/library/defense/uzr/

 

No, UZR doesn't use the eye test at all, other than the fact that it's determined by people who watch every single play. It's all based willy nilly on numbers.

 

So yeah, you're wrong on multiple levels.

 

[Dick Allen]

It will be interesting to see what effect shifts have on defensive metrics moving forward. Your infielders won't require as much range, I think sure-handedness will become a premium.

 

I was watching ESPN last night and the White Sox are 4th in baseball right now using shifts. Houston is far and away #1.

 

I do think, at least defensively, the eye test is much better than UZR. Jeter's name gets thrown out there, but the only people who have thought he was elite defensively are those voting for GG, and we all know offense has come into play for that award. 3 years for results? Do we not yet know if Viciedo is a good OF or not?

Edited April 21, 2014 by Dick Allen

[Eminor3rd]

Defensive metrics try to prove what your eyes are showing you. They're meaningless.

 

No, this is not true. Defensive metrics try to remove subjectivity from the equation and provide a common denominator for comparison between players. For example, between Elvis Andrus and Alexei Ramirez, who is the better defender? You have no idea, you just know they're both good. The metrics give us an unbiased, consistent account that serves as a frame of reference.

 

No, I won't see 1% of baseball plays because I actually watch the games. Errors are meaningless too. I can tell you that L Garcia is plus defender because he has insane range. Bobbling a couple balls doesn't change that.

 

Are you saying that you actually DO see enough baseball games to intelligently remark on who the best defenders are? No, you don't.

 

You could watch every inning of every single White Sox game this year and still only see 6% of the baseball that happened, and 85% of what you saw would be the centerfield view of the pitcher and hitter. No idea about positioning, no idea about first step, no idea even of a good portion of the route taken on the overwhelming majority of the plays made, excepted only by those accompanied by a really complete slow motion replay.

 

And you'd have NO idea how to compare White Sox players to almost every other team, considering you'd see at MOST 18 games of some Central teams.

 

Also, range is only one component of defense. Bobbling a couple balls doesn't make a guy a bad defender, but bobbling balls with some level of consistency DOES affect it. Arm strength, arm accuracy, instincts, and situational awareness are also massive factors.

 

Is there a stat that proves Alexei is the most underrated Sox player of my life?

 

Yes. fWAR has been giving Alexei TONS of credit for his defense for years. How? Defensive metrics!

[Eminor3rd]

And if defensive metrics need 3 seasons to be accurate, doesn't that mean that WAR ultimately may not be accurate?

 

No, it means WAR ultimately may not be predictive. Defensive metrics in low samples are accurate, but they aren't predictive.

[southsider2k5]

No, it means WAR ultimately may not be predictive. Defensive metrics in low samples are accurate, but they aren't predictive.

 

I think this is the biggest mistake/misrepresentation of stats. I don't think any are predictive in nature. I think they are all reactionary based on history. They are only predictive to the extent that history repeats itself in the same form that it has already happened.

[witesoxfan]

It will be interesting to see what effect shifts have on defensive metrics moving forward. Your infielders won't require as much range, I think sure-handedness will become a premium.

 

I was watching ESPN last night and the White Sox are 4th in baseball right now using shifts. Houston is far and away #1.

 

FanGraphs had an interesting article up the other day about the Astros apparently pitching "away" from a shift. It showed a infield shifted very far to the left (no 2B) with the pitcher working away from Jose Bautista. The theory behind it is Bautista is a rather extreme pull hitter, especially with offspeed stuff, so he's more likely to pull into that shift no matter where you're throwing the breaking balls.

 

http://www.fangraphs.com/blogs/why-would-a...inst-the-shift/

 

Shifting is really quite interesting to think and read about.

[Eminor3rd]

It will be interesting to see what effect shifts have on defensive metrics moving forward. Your infielders won't require as much range, I think sure-handedness will become a premium.

 

I was watching ESPN last night and the White Sox are 4th in baseball right now using shifts. Houston is far and away #1.

 

I do think, at least defensively, the eye test is much better than UZR. Jeter's name gets thrown out there, but the only people who have thought he was elite defensively are those voting for GG, and we all know offense has come into play for that award. 3 years for results? Do we not yet know if Viciedo is a good OF or not?

 

The shift is a huge issue. Right now, BIS simply ignores plays when shifts are implemented, because they don't know how to deal with them. This wasn't a huge deal even a few years ago, but shifts are becoming so common that large chunks of data are starting to be ignored, further reducing our already barely large enough samples.

 

Fortunately the MLBAM stuff they're implementing now will actually use analysis of recorded video rather than simply positioning points like UZR. Things are about to get much better there.

[Eminor3rd]

I think this is the biggest mistake/misrepresentation of stats. I don't think any are predictive in nature. I think they are all reactionary based on history. They are only predictive to the extent that history repeats itself in the same form that it has already happened.

 

This is an excellent point and is exactly what I was trying to get across to Marty a few days ago RE: "track record." The advanced stats we have now, particularly those that measure offense, are very good at accurately describing results. But very few are good at predicting the future with any level of precision. The best we can do now is identify the factors that will affect results going forward, but we can't get much past "it'll get worse" or "it'll get better." There are just too many factors involved. Players are changing, teams are changing, the environment is always changing.

[witesoxfan]

I think this is the biggest mistake/misrepresentation of stats. I don't think any are predictive in nature. I think they are all reactionary based on history. They are only predictive to the extent that history repeats itself in the same form that it has already happened.

 

This is why projections are made prior to the season. We use past history to help us determine what we can most likely expect, and then room for error is given within those projections (optimistic and pessimistic projections). Sometimes, players even [over/under] perform those numbers too. Trends are found and developed based on the average case - players will continually improve as they near the ages 29/30/31 with the growth tapering off, and beyond that point, they have a tendency of producing less and less with those numbers decreasing more and more as the player nears retirement.