WAR is a JOKE

[Uli2001]

I see that some of you do not understand rhetoric.

[Can_of_corn]

3 minutes ago, Uli2001 said:

I see that some of you do not understand rhetoric.

That is what that was supposed to be?

I thought rhetoric was supposed to be persuasive?

[BohKnowsBmore]

On 6/8/2019 at 10:01 AM, DrungoHazewood said:

I don't need to rehash old arguments, but defense slumps, too.  And players could have great defensive seasons.  No one bats an eye when George Brett hit .329-.390-.314 in consecutive years.  Or .255 the year after he won a batting title at .329. Nobody says that the flaws in batting average need to be addressed when players have wild swings.

Because BA is based on easily-verifiable outcomes (i.e. hit or not hit) that aren't nearly as simple for the defensive metrics.  Are you seriously suggesting that currently available defensive metrics have the same level of reliability as BA?

One can question whether BA reflects true talent in a given year, but one cannot question whether it adequately reflects results.  That same level of certainty does not exist for defensive metrics.

[DrungoHazewood]

10 hours ago, BohKnowsBmore said:

Because BA is based on easily-verifiable outcomes (i.e. hit or not hit) that aren't nearly as simple for the defensive metrics.  Are you seriously suggesting that currently available defensive metrics have the same level of reliability as BA?

One can question whether BA reflects true talent in a given year, but one cannot question whether it adequately reflects results.  That same level of certainty does not exist for defensive metrics.

No. What I'm saying is that because we haven't had solid metrics it has been assumed that fielding is much more consistent that offensive performance.  I think fielding is just as inconsistent, or at least subject to slumps and hot streaks.  

Turn it around the other way... what if we only had primitive metrics for hitting, and mainly judged hitters with the eye test and ESPN highlights.  We might give the silver slugger to the same guy 16 years in a row.

You can't assume that because defensive metrics sometimes have swings in individual performances year-to-year that they're wrong.

[DrungoHazewood]

11 hours ago, Can_of_corn said:

That is what that was supposed to be?

I thought rhetoric was supposed to be persuasive?

Rhetoric can be defined as language designed to have a persuasive or impressive effect on its audience, but often regarded as lacking in sincerity or meaningful content.

[weams]

On 6/8/2019 at 1:43 PM, Can_of_corn said:

Fangraphs calculates one of the two more frequently cited flavors of WAR.  This is their definition.

https://library.fangraphs.com/misc/war/

https://sabr.org/research/measuring-defense-entering-zones-fielding-statistics

[weams]

Just now, weams said:

https://sabr.org/research/measuring-defense-entering-zones-fielding-statistics

Base Running (BsR) is FanGraphs' all encompassing base running statistic that turns stolen bases, caught stealings, and other base running plays (taking extra bases, being thrown out on the bases, etc) into runs above and below average.

[weams]

Ultimate Zone Rating (UZR) is the defensive metric used in our WAR calculations to measuring fielding runs above average relative to the average player at that position. If you wanted to use Defensive Runs Saved (DRS) or any other related metric, you could do so manually, but our Def calculations use UZR for all years since 2003. Prior to that we use Total Zone (TZ) because UZR data does not exist prior to 2003 and TZ is the best available measure of fielding runs for earlier seasons.

The positional adjustment is used to account for the fact that defensive positions vary in their difficulty and importance. In other words, a league average shortstop prevents substantially more runs than a league average first baseman and we need to find a way to account for this when determining overall value. Adding the positional adjustment to the player’s fielding runs does exactly that. In the “Value” section of the player pages and leaderboards, you can find each player’s fielding runs and positional runs. Adding them together results in Def.

[weams]

12 hours ago, Can_of_corn said:

That is what that was supposed to be?

I thought rhetoric was supposed to be persuasive?

https://library.fangraphs.com/misc/war/

[weams]

9 minutes ago, DrungoHazewood said:

No. What I'm saying is that because we haven't had solid metrics it has been assumed that fielding is much more consistent that offensive performance.  I think fielding is just as inconsistent, or at least subject to slumps and hot streaks.  

Turn it around the other way... what if we only had primitive metrics for hitting, and mainly judged hitters with the eye test and ESPN highlights.  We might give the silver slugger to the same guy 16 years in a row.

You can't assume that because defensive metrics sometimes have swings in individual performances year-to-year that they're wrong.

I can assume that they don't have the proper components to be right. I can. 

[Can_of_corn]

14 minutes ago, DrungoHazewood said:

Rhetoric can be defined as language designed to have a persuasive or impressive effect on its audience, but often regarded as lacking in sincerity or meaningful content.

True, but do you think he would describe his own comments that way?

[DrungoHazewood]

28 minutes ago, weams said:

I can assume that they don't have the proper components to be right. I can. 

Sure.  My take is that in the aggregate they're pretty good, but there can be individual cases where strange things are afoot.

That year Gerardo Parra was a +35 outfielder.  I assume he was really a +10-15 outfielder who had a combination of teammates, attem balls, park peculiarities, and shifts that all broke in his favor that season.  That's why you talk about stuff instead of throwing MVP votes at a metric.  But outliers are not a reason to dismiss useful tools.  WAR is open source.  We can break down the components and the calculations, and talk about what might be problems.

[DrungoHazewood]

41 minutes ago, Can_of_corn said:

True, but do you think he would describe his own comments that way?

You can convince people of the worthiness of your cause by speaking authoritatively on subjects of which you have no knowledge.  Not everybody, but sometimes 30% is enough.  Especially if the 30% was already leaning in your direction.

[weams]

14 minutes ago, DrungoHazewood said:

Sure.  My take is that in the aggregate they're pretty good, but there can be individual cases where strange things are afoot.

That year Gerardo Parra was a +35 outfielder.  I assume he was really a +10-15 outfielder who had a combination of teammates, attem balls, park peculiarities, and shifts that all broke in his favor that season.  That's why you talk about stuff instead of throwing MVP votes at a metric.  But outliers are not a reason to dismiss useful tools.  WAR is open source.  We can break down the components and the calculations, and talk about what might be problems.

I just think the volatility is a false positive that is inherent in the formula. Make a better one. I'll buy in. I think the baserunning components and the positional modifications are weirdly pushing certain players up or down from what a real value might well be. 

[DrungoHazewood]

3 minutes ago, weams said:

I just think the volatility is a false positive that is inherent in the formula. Make a better one. I'll buy in. I think the baserunning components and the positional modifications are weirdly pushing certain players up or down from what a real value might well be. 

I think there are cases where teams have different value associated with a player at a position than WAR would indicate.  Manny at short, for example.  Both the O's and Padres decided that it was better for them to play Manny at short, even though he's not a plus defender there, than use a random body.  Manny's individual value might go down becasue of his defense at short, but it's better for the team than playing a Ryan Flaherty for weeks or months.

So there are cases where trade value or value to another team isn't accurately reflected by WAR in the context of current team.  And from there you can have a discussion about how that should impact things like awards, or All Star selections, or if someone is earning their contract.