2012 Brett Lawrie WAR.... How?

[RockCoCougars]

Perhaps this discussion should be in the statistical analysis forum, but since the Jays are in town, and it's Brewer-related in a roundabout way (and because a lot of discussions turn to 'advanced' stats) I'm posting it here.

 

I know that different sources have different ways of computing WAR, but how the heck is Brett Lawrie leading the A.L. in WAR on baseball reference... higher than Hamilton, higher than Adam Jones, higher than everybody. He's been absolutely pedestrian at the plate, which is reflected in his offensive WAR, but the guy has a 3.0 defensive WAR. Correct me if I'm wrong, but is WAR not a counting stat? Does this mean that his defense has been worth 3 'wins'? Perhaps I've missed something watching Blue Jay games this season.... is Lawrie covering the entire left side of the infield at once and acting as a roamer in left-center or something? I know that many don't like using errors as a stat, but the guy already has 9, which isn't that good for a third baseman at this point in the year.

 

Seriously, this causes me to once again question the validity of WAR as a meaningful stat. Perhaps the best regarded defensive 3B of all time in Brooks Robinson only had two seasons with a DWAR exceeding 3.0. The best defensive 3B that I've seen was Buddy Bell... he also only had two seasons with DWAR exceeding 3.0. Since we are not even halfway through the season is it really possible for Lawrie to have a DWAR in the 6.0 neighborhood? What gives?

[DuWayne Steurer]

Defensive war is extremely susceptible to weird fluctuations over small samples, and anything less than a few years is a small sample, especially when it comes to defensive numbers.

[Funketown]

Im not sure what d metroc they use over there but uzr only has him at +7 runs. So far bref has him at +40 in 110 or so games. Thats downright screwy.

[Dollarbill13]

I read somewhere that the way the Blue Jays use shifts makes Lawrie's defensive stats look really out of whack since he ends up fielding ground balls on the right field side of 2nd base. When shifting, most teams essentially move each guy over while the Blue Jays just put Lawrie in shallow right field and leave the rest of the infielders in place.

[rluzinski]

 

Seriously, this causes me to once again question the validity of WAR as a meaningful stat.

 

What it should do is cause you to question the validity of defensive metrics based on small samples. WAR is trying to estimate the value of a player's performance over a certain period of time. It is NOT trying to estimate the true level of talent of a player. For batter stats, sample size doesn't matter. A single is a single, no matter if it was a line drive one hopper to the CFer or a 8 hopper groundball between 3rd and short. The batter is given the run value of an average single in an average base/out situation.

 

For defense, it's different. We don't know if a defender saved a hit or botched a sure out, so we make an educated guess based on the trajectory and speed of the hit ball. That may work reasonably well over thousands of fielding opportunities but there is just too much noise after even half a season.

 

Long story short, in-season WAR numbers are very suspect when they are heavily influenced by the defensive run component. I'm even suspicious of single year WAR values when they include double-digit defensive run values. It seems to me like the defensive component should be regressed.

[rluzinski]

 

Seriously, this causes me to once again question the validity of WAR as a meaningful stat.

 

What it should do is cause you to question the validity of defensive metrics based on small samples. WAR is trying to estimate the value of a player's performance over a certain period of time. It is NOT trying to estimate the true level of talent of a player. For batter stats, sample size doesn't matter. A single is a single, no matter if it was a line drive one hopper to the CFer or a 8 hopper groundball between 3rd and short. The batter is given the run value of an average single in an average base/out situation.

 

For defense, it's different. We don't know if a defender saved a hit or botched a sure out, so we make an educated guess based on the trajectory and speed of the hit ball. That may work reasonably well over thousands of fielding opportunities but there is just too much noise after even half a season.

 

Long story short, in-season WAR numbers are very suspect when they are heavily influenced by the defensive run component. I'm even suspicious of single year WAR values when they include double-digit defensive run values. It seems to me like the defensive component should be regressed.

 

I read somewhere that the way the Blue Jays use shifts makes Lawrie's defensive stats look really out of whack since he ends up fielding ground balls on the right field side of 2nd base. When shifting, most teams essentially move each guy over while the Blue Jays just put Lawrie in shallow right field and leave the rest of the infielders in place.

 

With more and more teams employing extreme shifts, I believe that is a huge issue these days. I've heard the creator of UZR downplay that concern by stating that defensive positioning is a player skill, just like any other defensive skill. I agree with that in principle but often, the team is dictating a player's position, not the player himself.

[RockCoCougars]

Long story short, in-season WAR numbers are very suspect when they are heavily influenced by the defensive run component. I'm even suspicious of single year WAR values when they include double-digit defensive run values. It seems to me like the defensive component should be regressed.

 

This sounds logical to me.

 

I'll be the first to admit that I don't know very much about how WAR is computed, but the major problem I have with it is that people will toss a number like this out there in an argument, when it's obviously flawed. No way, no how would anyone's defense be worth roughly 10 percent of a teams wins.... especially a guy who is leading the league in errors at his position. I know the term 'win' in this case is not literal, but still.

[logan82]

With more and more teams employing extreme shifts, I believe that is a huge issue these days. I've heard the creator of UZR downplay that concern by stating that defensive positioning is a player skill, just like any other defensive skill. I agree with that in principle but often, the team is dictating a player's position, not the player himself.

I was under the impression that plays with defensive shifts were thrown out.

 

 

especially a guy who is leading the league in errors at his position.

Error are not equal to bad defense. Range can make up for errors.

[Funketown]

I think UZR's method throws out shifts, but I'd have to think BBRef's doesn't throw out shifts or something goofy. Has anybody actually watched enough of his defense in the majors to comment? Every defensive metric grades him out as a very good defender, and I expected him to be a bit of a butcher to be honest.

[logan82]

He is a good athlete so I don't really know what to expect. I never got to see him in Appleton. He has graded out well in both UZR and DRS so far in his career. Well above average in fact. Small sample though. I don't remember the discussion bout his defense in the minors.

[RockCoCougars]

Error are not equal to bad defense. Range can make up for errors.

 

I'll give you that, but what about throwing errors? I haven't watched the Jays play a ton this season, but I've seen Lawrie make two throwing errors. One was in a crucial spot, and basically cost them the game. Obviously a small sample, but his defense really hasn't stood out as great to me, and I would have taken note had he looked great out there.

[jjfanec]

Makes no sense. An OPS of .729. Only 8 of 16 on steals. Sure his defense is solid but it hasnt been amazing

[3and2Fastball]

I read somewhere that the way the Blue Jays use shifts makes Lawrie's defensive stats look really out of whack since he ends up fielding ground balls on the right field side of 2nd base. When shifting, most teams essentially move each guy over while the Blue Jays just put Lawrie in shallow right field and leave the rest of the infielders in place.

 

That is it essentially. If Aramis Ramirez could move from 3rd base position quick enough to field grounders on the right field side of 2B he'd have a WAR of 3 too... the computers think Lawrie is the Cool Papa Bell of 3rd base....

[rluzinski]

I think UZR's method throws out shifts, but I'd have to think BBRef's doesn't throw out shifts or something goofy. Has anybody actually watched enough of his defense in the majors to comment? Every defensive metric grades him out as a very good defender, and I expected him to be a bit of a butcher to be honest.

Here is a whole thread about Lawrie, his inflated DRS numbers this year (the defensive metric BF.net uses) and defensive shifts:

 

http://www.insidethebook.com/ee/index.php/site/comments/has_brett_lawrie_already_saved_14_runs_so_far_this_year/

 

MGL, who created UZR said this, "For the record, when there is a shift and it “affected” the play (whatever that means), according to the data provider, UZR ignores that play."

 

That is news to me. I guess DRS does not do this.

[Funketown]

Thanks rluz.

 

So DRS does not throw out those plays from what I understand. Looks like much of his inflated DRS could be coming from that shift then; the article from the Ottawa paper (silly Canadians talking defensive metrics!) says 5 of his 12 runs saved up to a point were from right of 2B. So it seems that the shift is causing at least some part of the disparity, no?

[Thurston Fluff]

Do I sense more allies in my war against WAR? It is not a good stat because it uses incomplete stats. It uses some form of UZR. UZR needs three years for every one of offensive stats to be as accurate. Yet it uses them in a one to one ratio. It is useless...almost. I'll give the likes of RLuz their due.

 

What it should do is cause you to question the validity of defensive metrics based on small samples. WAR is trying to estimate the value of a player's performance over a certain period of time. It is NOT trying to estimate the true level of talent of a player.

 

You've just summed up both of my primary arguments in my war on WAR. I question why anyone should believe a stat that incorporates a questionable stat is somehow accurate? I fully understand what they are trying to do and applaud the effort. But at this point using one season of defensive stats is like using leaches to cure a fever.

The second reason for my war on WAR is how it is used. If everyone used it like you do fine. You tend to say so and so had a WAR of such and such over this period of time therefore was/n't a value. I'm ok with that. I think there are better measures but I can see it useful as part of an overall argument. But it often gets used as a predictive, stand alone, stat when it should not be used that way. It is particularly useless when looking at young players. Overall if the questionable numbers being put in aren't bad enough the way it is used makes it essentially worthless.

 

Be a sabre hippie. Join the stop the WAR movement today.

[HiAndTight]

It's funny, I said that Lawrie had a chance to be as good as Ryan Braun and match him if not surpass him in his peak years given his ability to play defense and his obvious ridiculous offense(again, at this age Braun was playing in A+ and AA).

 

I was mocked for making that statement.

 

And now that Lawrie's bat's heating up, and he's playing great defense there is every excuse to diminish him.

 

Dude's just a stud. The Marcum-Lawrie trade will probably be talked about in 15 years, if not by many people as a terrible trade, at least Brewers fans.

 

For the record, I do believe Braun will always be the better hitter. But I think plus defense and VERY good offense from Lawrie could easily lead to 7-8 WAR seasons a couple times in his career.

[Raw Biz]

It's funny, I said that Lawrie had a chance to be as good as Ryan Braun and match him if not surpass him in his peak years given his ability to play defense and his obvious ridiculous offense(again, at this age Braun was playing in A+ and AA).

 

I was mocked for making that statement.

 

And now that Lawrie's bat's heating up, and he's playing great defense there is every excuse to diminish him.

 

Dude's just a stud. The Marcum-Lawrie trade will probably be talked about in 15 years, if not by many people as a terrible trade, at least Brewers fans.

 

For the record, I do believe Braun will always be the better hitter. But I think plus defense and VERY good offense from Lawrie could easily lead to 7-8 WAR seasons a couple times in his career.

 

Again, the Brewers most likely don't make the playoffs with Marcum. Sure Lawrie could end up being a stud for many years to come. But that doesn't mean it was a terrible trade. It helped the Brewers. It may just be an example of one of those win-win trades. Now you might argue that Lawrie will ultimately have more value. However, that doesn't mean it was a bad trade.

[HiAndTight]

Do I sense more allies in my war against WAR? It is not a good stat because it uses incomplete stats. It uses some form of UZR. UZR needs three years for every one of offensive stats to be as accurate. Yet it uses them in a one to one ratio. It is useless...almost. I'll give the likes of RLuz their due.

 

 

You have an allie here because people use WAR to try and determine how many games you'll literally win with one player.

 

Ie, "it doesn't make sense to trade for K-Rod, at best he'll give you a half a win over the rest of the season."

It's absurd to look at it that way.

 

I think though comparing players it's a great metric. It's not perfect, and I think it's most useful at the end of the year(obviously).

 

But I just hear excuses regarding Lawrie. We shift out guys all over the place. Lots of teams use shifts. Not all the sudden it costs Lawrie?

 

Who by the way, I would say has been more than "pedestrian," at the plate. .291/.338/.417 line from a guy who won't turn 23 until next Jan? Pretty respectable line. But I think it softens the blow to try and argue why he's not as good as the numbers show or most of the experts are arguing.

[jjfanec]

It's funny, I said that Lawrie had a chance to be as good as Ryan Braun and match him if not surpass him in his peak years given his ability to play defense and his obvious ridiculous offense(again, at this age Braun was playing in A+ and AA).

 

I was mocked for making that statement.

 

And now that Lawrie's bat's heating up, and he's playing great defense there is every excuse to diminish him.

 

Dude's just a stud. The Marcum-Lawrie trade will probably be talked about in 15 years, if not by many people as a terrible trade, at least Brewers fans.

 

For the record, I do believe Braun will always be the better hitter. But I think plus defense and VERY good offense from Lawrie could easily lead to 7-8 WAR seasons a couple times in his career.

 

His bat had to heat up since his OPS was under .700 for a while. His bat is nowhere near Braun's even after his hot streak. Plus defense doesnt make for a .300 point different in OPS. His bat could turn into something special but last year was the first year he put up out-of-this-world numbers. People are not making excuses to diminish him. They are looking at reality. His batting numbers are average. His base-running is poor, and he has made a few errors so the WAR seems out of line. He is going to be a very good player but he is nowhere what his WAR says.

[RockCoCougars]

I called him pedestrian because he has been at the plate. I should know because I overpaid for him in fantasy baseball. He started fast, then disappeared for about 6 weeks outside of the helmet toss. He's come around a bit lately, but he certainly didn't pick off where he left off last year. I say this knowing full well that he's still very young, but so far this year he hasn't been any great shakes.

[logan82]

he has made a few errors so the WAR seems out of line.

Possibly. Is he making errors on plays that most 3B wouldn't even get to or on easy plays? If he is making a lot of plays on balls most 3B wouldn't get to he is more than making up for the errors. Seems to me that it is possible he doesn't know when to hold onto the ball and when he should throw it. Much like Bill Hall a few years back, he may think he can make every play.

 

He definitely has a lot of raw talent. When Braun came up he was older and more polished as a hitter. I don't think we should be making any definitive judgements yet.

[monty57]

 

Seriously, this causes me to once again question the validity of WAR as a meaningful stat.

 

What it should do is cause you to question the validity of defensive metrics based on small samples. WAR is trying to estimate the value of a player's performance over a certain period of time. It is NOT trying to estimate the true level of talent of a player. For batter stats, sample size doesn't matter. A single is a single, no matter if it was a line drive one hopper to the CFer or a 8 hopper groundball between 3rd and short. The batter is given the run value of an average single in an average base/out situation.

 

For defense, it's different. We don't know if a defender saved a hit or botched a sure out, so we make an educated guess based on the trajectory and speed of the hit ball. That may work reasonably well over thousands of fielding opportunities but there is just too much noise after even half a season.

 

Long story short, in-season WAR numbers are very suspect when they are heavily influenced by the defensive run component. I'm even suspicious of single year WAR values when they include double-digit defensive run values. It seems to me like the defensive component should be regressed.

 

I read somewhere that the way the Blue Jays use shifts makes Lawrie's defensive stats look really out of whack since he ends up fielding ground balls on the right field side of 2nd base. When shifting, most teams essentially move each guy over while the Blue Jays just put Lawrie in shallow right field and leave the rest of the infielders in place.

 

With more and more teams employing extreme shifts, I believe that is a huge issue these days. I've heard the creator of UZR downplay that concern by stating that defensive positioning is a player skill, just like any other defensive skill. I agree with that in principle but often, the team is dictating a player's position, not the player himself.

 

Since they have all this data, why don't they just peg the starting point from where the player actually started the play instead of where a 3B "should have" started the play?

[Baldkin]

Because, for the most part, the public (which is what these websites are) doesn't have access to that data. Camera's don't show where position players are positioned before the pitch (or as the pitch is thrown).

[Thurston Fluff]

He is going to be a very good player but he is nowhere what his WAR says.

 

WAR isn't supposed to tell you that. Which is one of my major problems with it. It's supposed to tell you how well he did over a certain period of time not how good he is as a player. I don't think it does a particularly good job of even that but I'll let that be for now. I suppose a number of WARs over a long period of time should get you closer to actual ability simply because a player's numbers over a long period is going to be close to his ability. But that is not the case here. A single season of WAR will never tell you much about any players true ability though. It was not designed to.