A Realistic View at 2019: Part II

[Kimmi]

This is the problem with the stat-driven "trained observers". If a ball is hit 50' to JBJ's left, he runs to the spot and waits for the ball, and another OF'er has a ball hit 50' to his left and he catches it on the run they both get credit for being able to catch up with a ball hit 50' to their left.

Statistically these two players are equal.

 

The problem isn't in the formula. It's in the stat driven "trained observers" who are more interested in accurately reporting the fact that the ball was caught than in the fact that JBJ is obviously the better player because he was waiting for the ball.

 

You are right, the defenders don't get any style points for the way they make the plays. However, if Jackie is really that much better than other defenders, he will get to balls that other defenders do not get to.

 

How do you know how well Jackie is doing, relative to the other centerfielders this season?

 

Also, and more importantly, don't read too much into a 1/2 season's worth of defensive metrics. One or two missed balls, which Jackie has had this season, can greatly sway the numbers in such a small sample size.

[Kimmi]

For defensive ratings that mean anything you only have two choices - WAR analytical stuff or eyeball stuff.

 

As flawed as WAR stuff may be, the idea that eyeball stuff is better is kind of a joke IMHO

 

This is it, in a nutshell.

 

And actually, as flawed as WAR may be, it is pretty doggone good in describing what has happened on the field.

[notin]

This is the problem with the stat-driven "trained observers". If a ball is hit 50' to JBJ's left, he runs to the spot and waits for the ball, and another OF'er has a ball hit 50' to his left and he catches it on the run they both get credit for being able to catch up with a ball hit 50' to their left.

Statistically these two players are equal.

 

The problem isn't in the formula. It's in the stat driven "trained observers" who are more interested in accurately reporting the fact that the ball was caught than in the fact that JBJ is obviously the better player because he was waiting for the ball.

 

 

So... you want Jackie to get more credit for waiting?

 

This type of discrepancy is handled when both fielders have a ball hit 70 feet to their left and Jackie gets to it on the run while CF X chases it to the wall.

 

The bottom line for Bradley’s rating:

 

3rd in innings in CF

8 in plays made in zone

14th in plays out of zone

 

The plays in zone and innings is the key. While he’s out there in CF more than nearly everyone else, 7 other center fielders, many with fewer innings, are making more routine plays. His balls in zone/inning within the zone are very low, probably among the worst in the league. This means Jackie is simply less involved defensively. If he is less involved defensively, the CFs making plays are getting (and deserve) more credit.

 

Jackie’s Out Of Zone plays - the ones he has to run 50 or whatever feet to make - are 14th in the league. Now as we have already seen Bradley to be less involved within the zone, it’s not hard to grasp that the OOZ opportunities/inning are probably also low. And since Betts (4th in OOZ) is probably taking a few himself - and can only take them from Bradley - that diminishes his opportunities even more.

 

So basically Bradley is not seeing the chances he needs to get the better UZR. Not his fault, but also not a flaw...

Edited July 16, 2019 by notin

[dgalehouse]

WAR is the invention of a sportswriter. It uses mostly basic stats , mixed together in a complicated formula , with some opinion added . When run through the blender , it spits out a numerical rating for a given player . Different websites use somewhat different formulas , but the result is supposed to tell you how many wins a player is worth as opposed to some jabroni . It gives a nice , convenient number . It does have some value . But we really don't have to swear by it . I know I don't. It does little to enhance my enjoyment of the game .

[Bellhorn04]

I think JBJ is a perfect example of why WAR is a valuable thing. If you only looked at the back of his baseball card, you'd wonder why he's been our everyday CFer so long.

[OH FOY]

"Two thirds of the earth is covered by water, the other one-third, by Garry Maddox". Edited July 16, 2019 by OH FOY!

[notin]

WAR is the invention of a sportswriter. It uses mostly basic stats , mixed together in a complicated formula , with some opinion added . When run through the blender , it spits out a numerical rating for a given player . Different websites use somewhat different formulas , but the result is supposed to tell you how many wins a player is worth as opposed to some jabroni . It gives a nice , convenient number . It does have some value . But we really don't have to swear by it . I know I don't. It does little to enhance my enjoyment of the game .

 

Almost all stats were invented by sportswriters, and all of them were created to try and separate the abilities and performances of players to some degree.

 

Beach Rickey concocted slugging percentage because batting average didn’t distinguish between a 3 for 10 hitter with 3 singles from a 3 for 10 hitter with 2 home runs and a double. Many were skeptical and it took the BBWAA about 40 years to make it an official stat. This is all nothing new...

[notin]

"Two thirds of the earth was covered by water, the other one-third, by Garry Maddox".

 

One of my favorites.

 

Or Davey Lopes upon learning Gary Pettis and Rickey Henderson we’re minor league teammates. “They didn’t need a third outfielder.”

[S5Dewey]

We're getting away from my point though - that according to dWAR JBJ is a worse defensive player than a AAAA guy. Can't we say that any formula that says that one of the top defensive outfielders in MLB is worse defensively than a AAAA player has a very significant flaw someplace?

 

As I said, JBJ is just the local poster child for demonstrating that flaw.

[Bellhorn04]

We're getting away from my point though - that according to dWAR JBJ is a worse defensive player than a AAAA guy.

 

dWAR currently has him at 0.0 for the season.

[S5Dewey]

Almost all stats were invented by sportswriters, and all of them were created to try and separate the abilities and performances of players to some degree.

 

Beach Rickey concocted slugging percentage because batting average didn’t distinguish between a 3 for 10 hitter with 3 singles from a 3 for 10 hitter with 2 home runs and a double. Many were skeptical and it took the BBWAA about 40 years to make it an official stat. This is all nothing new...

 

The difference is that SLG, OPS, etc. are more objective stats. They're driven by numbers we can hang out hats on. Sure, there are some minor issues with those stats but defensive stats are more subjective. I've worked with people all my life and I can tell you for sure that people are going to see and do things differently regardless of how much "training" they get. And the farther removed (in time) from the training they get the more subjective they get.

[S5Dewey]

dWAR currently has him at 0.0 for the season.

 

fangraphs had him @-0.4 this morning, but I'll settle for your saying that JBJ is no better than a AAAA player. I don't think either of us believes that.

[a700hitter]

For defensive ratings that mean anything you only have two choices - WAR analytical stuff or eyeball stuff.

 

As flawed as WAR stuff may be, the idea that eyeball stuff is better is kind of a joke IMHO

It depends up whose eyeballs are being used. A good set of eyeballs that understands what they are seeing is better at evaluating than any stat. The one shortcoming of the eyeballs is that even with Extra Innings Directv they can’t watch 800 MLB players and thousands of minor leaguers. If someone doesn’t have the eyeballs to process what they are seeing and evaluate it, they probably can’t use stats properly either.

[Bellhorn04]

fangraphs had him @-0.4 this morning, but I'll settle for your saying that JBJ is no better than a AAAA player. I don't think either of us believes that.

 

I don't believe that, no. But I do believe it's quite possible his play hasn't been anything special this year.

 

It's like a lot of other stats. You look at the current year but you also look at the career numbers. His career numbers clearly show he's a well-above average CFer. But performances can vary a lot.

[Bellhorn04]

It depends up whose eyeballs are being used. A good set of eyeballs that understands what they are seeing is better at evaluating than any stat. The one shortcoming of the eyeballs is that even with Extra Innings Directv they can’t watch 800 MLB players and thousands of minor leaguers. If someone doesn’t have the eyeballs to process what they are seeing and evaluate it, they probably can’t use stats properly either.

 

I don't watch that many games any more so without the stats I'd be pretty lost as to what was going on.

[harmony]

This is the problem with the stat-driven "trained observers". If a ball is hit 50' to JBJ's left, he runs to the spot and waits for the ball, and another OF'er has a ball hit 50' to his left and he catches it on the run they both get credit for being able to catch up with a ball hit 50' to their left.

Statistically these two players are equal.

 

The problem isn't in the formula. It's in the stat driven "trained observers" who are more interested in accurately reporting the fact that the ball was caught than in the fact that JBJ is obviously the better player because he was waiting for the ball.

In the "old days" -- perhaps eight years ago -- the trained observers made the defensive assessments on each play.

 

Today those calls are made by Statcast and similar devices that precisely track, among other things, the angle and velocity of the ball as well as the position of the player.

 

For better or worse, technology has taken over the game. I recommend the new book The MVP Machine: How Baseball's New Nonconformists Are Using Data to Build Better Players by Ben Lindbergh and Travis Sawchik:

 

https://www.theatlantic.com/magazine/archive/2019/07/building-the-next-babe-ruth/590641/

Edited July 16, 2019 by harmony

[notin]

The difference is that SLG, OPS, etc. are more objective stats. They're driven by numbers we can hang out hats on. Sure, there are some minor issues with those stats but defensive stats are more subjective. I've worked with people all my life and I can tell you for sure that people are going to see and do things differently regardless of how much "training" they get. And the farther removed (in time) from the training they get the more subjective they get.

 

The point was about the invention and acceptance of “new stats.” All conventional stats were new stats at one point.

 

But even hits are somewhat subjective. A certain percentage do fall into the opinion of an official scorer.

 

Defense in baseball has always been difficult to measure. And this system tries to standardize it as much as possible. Nothing is more subjective than the Eyeball Test. Nothing. Too often the Eyeball Test turns into “I saw that guy make an error once.”

 

The defensive metrics try to mitigate a lot of the subjectivity by measuring balls hit into defined zones aka “the part of the field a player should cover.” Some things are difficult to account for, such as height of a flyball, etc. After all, the zone system was created for infielders, where some of these types of flaws are inherently accounted for. But not all.

 

Is it perfect? No. Is it better than heavily weighted eyeball testing? Yes, in the same way that a car is better than a horse...

[dgalehouse]

The Sox have a very realistic shot at a seven game winning streak. If you think Toronto is bad , wait until you see the Orioles. This is like men against boys . However , at some point, the Sox are going to have to show they can beat the good teams . Halfway through July , they really haven't shown that . Next week will begin the critical stretch . Time to step up .

[notin]

We're getting away from my point though - that according to dWAR JBJ is a worse defensive player than a AAAA guy. Can't we say that any formula that says that one of the top defensive outfielders in MLB is worse defensively than a AAAA player has a very significant flaw someplace?

 

As I said, JBJ is just the local poster child for demonstrating that flaw.

 

Is that the point? I don’t think it is.

 

The point is actually that the Red Sox wouldn’t be any different with an AAAA centerfielder playing defense than they are with Bradley. And the reason is - stay with me here - hardly anything is being hit out there. If the Sox promoted, say, Cole Sturgeon to play CF, it would make very little difference defensively, because Bradley is one of the least involved centerfielders in the league. Not because he can’t get to anything, but because there just isn’t all that much to get to compared to his counterparts around the league...

[S5Dewey]

I don't believe that, no. But I do believe it's quite possible his play hasn't been anything special this year.

 

It's like a lot of other stats. You look at the current year but you also look at the career numbers. His career numbers clearly show he's a well-above average CFer. But performances can vary a lot.

 

You talk about career dWAR numbers and I did look them up. Here's what they are year to year and I find it to be really ODD.

 

2013......-1.7

2014......13.0

2015....... 2.9

2016......10.5

2017.......5.8

2018......10.1

2019......-0.4 (so far)

 

I find this odd because it's hard to believe that a player has that kind of swings in defense year-to-year. Usually a player is what he is but in alternate years JBJ has vacillated between pretty good and OMG good. Strange. (Or maybe it's a product of those "trained observers

[moonslav59]

Lol. Who says they are unbiased?

Who says they are well trained?

A replacement player is worse then every major league player. If you think JBj’s defense is worse then that guy I don’t know what to say....

 

They are less biases than I.

 

They are more trained than I.

 

A AAAA player can be an outstanding fielder with no hit tool.

[moonslav59]

You talk about career dWAR numbers and I did look them up. Here's what they are year to year and I find it to be really ODD.

 

2013......-1.7

2014......13.0

2015....... 2.9

2016......10.5

2017.......5.8

2018......10.1

2019......-0.4 (so far)

 

Looks about as up and down as his OPS, and that ain't odd.

 

They say look at 2-3 years combined sample sizes anyway.

 

[notin]

You talk about career dWAR numbers and I did look them up. Here's what they are year to year and I find it to be really ODD.

 

2013......-1.7

2014......13.0

2015....... 2.9

2016......10.5

2017.......5.8

2018......10.1

2019......-0.4 (so far)

 

I find this odd because it's hard to believe that a player has that kind of swings in defense year-to-year. Usually a player is what he is but in alternate years JBJ has vacillated between pretty good and OMG good. Strange. (Or maybe it's a product of those "trained observers

 

I think fans think a player’s defense doesn’t fluctuate from year to year, but since for most our baseball-watching lives no one measured it, it was really just an assumption.

 

Overall, his bad years listed above were also the years he made the fewest plays. Now they are not all distinctly different from the good ones, but they might be on a per inning basis relative to the rest of the league...

[a700hitter]

I don't watch that many games any more so without the stats I'd be pretty lost as to what was going on.

I need them for some of the West Coast teams that I don't watch much, but I don't need to look at stats to know that Bradley is a tremendous defensive talent. If he makes a few errors, that would not change my mind. He's one of the best. Kiermeyer is right there with him. The kid Laureano in Oakland is pretty good to as is Buxton on MN. Pillar is a slight notch below, but also excellent. No stats needed.

[a700hitter]

Defense in baseball has always been difficult to measure. And this system tries to standardize it as much as possible. Nothing is more subjective than the Eyeball Test. Nothing. Too often the Eyeball Test turns into “I saw that guy make an error once.”

 

That's a bad set of eyeballs and someone like that would probably misuse stats too.

[Bellhorn04]

What I like about WAR is that it tries to address questions like this one: does JBJ's defense offset his bad offense enough to make him a valuable player?

[OH FOY]

What I like about WAR is that it tries to address questions like this one: does JBJ's defense offset his bad offense enough to make him a valuable player?

 

Your joking right? You must be.

[S5Dewey]

Is that the point? I don’t think it is.

 

Bradley is one of the least involved centerfielders in the league.

 

AL East PO's by CF'ers

 

Boston.......194

NYY...........197

TBay.........172

Tor............211

Balt...........288

 

I don't see a huge difference there. If you throw out Baltimore, whose pitching staff is going to guarantee a lot of balls hit a long way, the average of the remaining four is 193.5, exactly where Boston is. Since JBJ is essentially the only guy who plays CF for the Sox it's safe to say that he's getting about the same number of PO's as the remaining four CF's in the AL East. (which surprises me too, given that he's playing next to Mookie Freakin' Betts!)

[harmony]

You talk about career dWAR numbers and I did look them up. Here's what they are year to year and I find it to be really ODD.

 

2013......-1.7

2014......13.0

2015....... 2.9

2016......10.5

2017.......5.8

2018......10.1

2019......-0.4 (so far)

 

I find this odd because it's hard to believe that a player has that kind of swings in defense year-to-year. Usually a player is what he is but in alternate years JBJ has vacillated between pretty good and OMG good. Strange. (Or maybe it's a product of those "trained observers

Jackie Bradley Jr.'s OPS+ went from 49 in 423 plate appearances in 2014 to 119 in 255 plate appearances in 2015. Mookie Betts' OPS+ went from 108 in 2017 to 185 in 2018 to 124 this year.

 

Stats sometimes fluctuate year to year.

[Slasher9]

Dwar can suck a D. So can war. If you want to just use stats try WAA. At least its comparing the MLB player to the MLB average. But defense war can really suck stinky feet.

Funny how the stat folk ignore WAA but cling to WAR as if it were their crying child.