Jim Caple on WAR

[filihok]

Jim Caple wrote an article today on WAR that has Twitter all a twitter.

Here it is.

http://espn.go.com/mlb/story/_/id/89...placement-stat

You should read it. He makes some good points. But also makes some poor points.




There are a lot of responses to Caple's piece.

This one, by Bill Baer of Crashburn Alley (A Phillies blog) is one of the best.

http://crashburnalley.com/2013/02/01...oughts-on-war/

You should read it as well.

[Bothrops Atrox]

WAR is a lot easier to calculate and understand than QB rating, but that hasn't stopped ESPN from using it for 20 years. There is a lot to it, but very little post-middle school math. Anybody who takes the time and has a middle-school education could understand it, if they wanted to.

The point regarding the difficulty in quantifying defense is very good, but also the main reason why there is so much difference between the two WARs. So really, Caple made one good point there, not two.

But when you consider that even estimations of defensive ability are better than ignoring defense all together, nothing Caple said about defense gets me too excited one way or another.

There are at least 4-5 other big issues with WAR that he didn't even touch on, such as: (batted ball assumptions, lack of comprehensive catcher evaluations, correct replacement level, league quality issues, etc.). So as it is with many posters here, he didn't even understand it enough to have the credibility to trash it.

[chicagowhitesox1173]

He does bring up a good point on how Brendan Ryan had a higher war than Derek Jeter on baseball reference and Jeter had a higher war on fan graphs. Same with Morris.

[Bothrops Atrox]

He does bring up a good point on how Brendan Ryan had a higher war than Derek Jeter on baseball reference and Jeter had a higher war on fan graphs. Same with Morris.

That is due to the fact that the two WARs use different defensive systems. There is virtually no universal agreement on which system between UZR or DRS is better. If one WAR went with the defensive metric that they don't use now, the only real difference between the two would be a lower replacement level/ higher WAR in Fangraphs and an AL vs. NL component is BBref that FG does not have. If they both adopted the same replacement value and defensive system, one WAR would become redundant for batters.

Pitchers is a different story. Most of Morris' WAR difference between the two is due to differences in replacement level. The difference between FIP and RA makes up the rest. That is a fundamental difference between the two that there is also not total agreement on.

[Ubiquitous]

A huge problem with WAR that never gets talked about is how the defensive side of the equation isn't consistent throughout history. The data to make it consistent simply doesn't exist but it almost seems like nobody out there really realizes this when they look at WAR. WAR spits out a number for Willie Mays and for Mike Trout but one is using detailed play by play data while the other is basically using stats and metrics that haven't changed in over a decade.

[Bothrops Atrox]

A huge problem with WAR that never gets talked about is how the defensive side of the equation isn't consistent throughout history. The data to make it consistent simply doesn't exist but it almost seems like nobody out there really realizes this when they look at WAR. WAR spits out a number for Willie Mays and for Mike Trout but one is using detailed play by play data while the other is basically using stats and metrics that haven't changed in over a decade.

This is 100% correct, and by far the biggest issue with WAR.
Somebody could use TZ or Michael Humphry's DRA throughout history, but I guess they went with (assumed) increase in accuracy from 2003-on in place of consistency.

[Jackaroo Dave]

A huge problem with WAR that never gets talked about is how the defensive side of the equation isn't consistent throughout history. The data to make it consistent simply doesn't exist but it almost seems like nobody out there really realizes this when they look at WAR. WAR spits out a number for Willie Mays and for Mike Trout but one is using detailed play by play data while the other is basically using stats and metrics that haven't changed in over a decade.

This should be obvious to anyone who looks at bbref's advanced fielding stats and compares what's there and what isn't. The dinosaurs really have a point when they resent someone waltzing in, "La la WAR says blah," and waltzing out.

[9RoyHobbsRF]

copied and pasted from another site

what is the answer to this?

I'll completely buy in to WAR when it accurately shadows team wins. According to Fangraph's WAR, the Orioles, were 20th in WAR but won 93 games. The A's won 93 games with a 14th ranked WAR. The Diamondbacks where 6th and only won 81 games. The Giants were only 12th and won it all

I suspect that it just looks at big stats and does formulas and ratings

A 3 run homer when you are already winning by 6 runs is virtually menaingless while a groundout to second with no outs and a runner on second to advance a runner to third who then scores the winning run is MUCH more meaningful, we know which one WAR rates as superior

you can have WAR, quote WAR, use WAR - I will take something more reliable

[ol' aches and pains]

Caple says in his blog: "Batting average, OPS, ERA, etc., are the products of such simple and certain mathematical calculations that even I can figure out armed with only a pencil and paper. WAR, however, is partially a product of fielding metrics that are by no means certainties..." "In other words, most baseball stats are based entirely on indisputable math calculations. WAR has an element of theory and assumption to it."

That pretty well sums up my objection to WAR. I want to know what the player actually did, not what he might do compared to some hypothetical replacement player. OBP, OPS, SLG, are absolute, no theory or assumption necessary.

Defensive play is difficult, if not impossible to quantify statiscally. To throw questionable defensive metrics into the calculation renders the data worthless, IMO. Was Alex Rodriguez, juiced to the gills, a better hitter than Rabbit Maranville? Of course he was. Was he a better defensive shortstop? Who the hell knows?

Of course my other problem with WAR is the multiple calculations; i.e. fWAR vs. bWAR. I'm not going to pretend that I understand how WAR is calculated, my brain hurts just thinking about it. But the people who are championing WAR as the sexy new evaluation tool can't even agree among themselves. I think it's rather presumptuous to ask us to accept WAR as the new way of doing things when there are different factions using different models to arrive at different data.

I know I'm probably going to get flamed six ways from Sunday for this post, but that wasn't my purpose in writing it. I have enough aggravation in my real life without looking for more on here. But I strongly believe the bugs haven't been worked out of WAR yet, and while I'm by no means a scientist, I can't believe that a metric with so many variables and intangibles would be accepted in any other field.

[9RoyHobbsRF]

GREAT post aches

Caple says in his blog: "Batting average, OPS, ERA, etc., are the products of such simple and certain mathematical calculations that even I can figure out armed with only a pencil and paper. WAR, however, is partially a product of fielding metrics that are by no means certainties..." "In other words, most baseball stats are based entirely on indisputable math calculations. WAR has an element of theory and assumption to it."

That pretty well sums up my objection to WAR. I want to know what the player actually did, not what he might do compared to some hypothetical replacement player. OBP, OPS, SLG, are absolute, no theory or assumption necessary.

Defensive play is difficult, if not impossible to quantify statiscally. To throw questionable defensive metrics into the calculation renders the data worthless, IMO. Was Alex Rodriguez, juiced to the gills, a better hitter than Rabbit Maranville? Of course he was. Was he a better defensive shortstop? Who the hell knows?

Of course my other problem with WAR is the multiple calculations; i.e. fWAR vs. bWAR. I'm not going to pretend that I understand how WAR is calculated, my brain hurts just thinking about it. But the people who are championing WAR as the sexy new evaluation tool can't even agree among themselves. I think it's rather presumptuous to ask us to accept WAR as the new way of doing things when there are different factions using different models to arrive at different data.

I know I'm probably going to get flamed six ways from Sunday for this post, but that wasn't my purpose in writing it. I have enough aggravation in my real life without looking for more on here. But I strongly believe the bugs haven't been worked out of WAR yet, and while I'm by no means a scientist, I can't believe that a metric with so many variables and intangibles would be accepted in any other field.

[Bothrops Atrox]

copied and pasted from another site


I'll completely buy in to WAR when it accurately shadows team wins. According to Fangraph's WAR, the Orioles, were 20th in WAR but won 93 games. The A's won 93 games with a 14th ranked WAR. The Diamondbacks where 6th and only won 81 games. The Giants were only 12th and won it all

WAR lines up with run differential very well. The issue is how much stock does one put into run differential vs. WL record.

[Bothrops Atrox]

Caple says in his blog: "Batting average, OPS, ERA, etc., are the products of such simple and certain mathematical calculations that even I can figure out armed with only a pencil and paper. WAR, however, is partially a product of fielding metrics that are by no means certainties..." "In other words, most baseball stats are based entirely on indisputable math calculations. WAR has an element of theory and assumption to it."

That pretty well sums up my objection to WAR. I want to know what the player actually did, not what he might do compared to some hypothetical replacement player. OBP, OPS, SLG, are absolute, no theory or assumption necessary.

Defensive play is difficult, if not impossible to quantify statiscally. To throw questionable defensive metrics into the calculation renders the data worthless, IMO. Was Alex Rodriguez, juiced to the gills, a better hitter than Rabbit Maranville? Of course he was. Was he a better defensive shortstop? Who the hell knows?

Of course my other problem with WAR is the multiple calculations; i.e. fWAR vs. bWAR. I'm not going to pretend that I understand how WAR is calculated, my brain hurts just thinking about it. But the people who are championing WAR as the sexy new evaluation tool can't even agree among themselves. I think it's rather presumptuous to ask us to accept WAR as the new way of doing things when there are different factions using different models to arrive at different data.

I know I'm probably going to get flamed six ways from Sunday for this post, but that wasn't my purpose in writing it. I have enough aggravation in my real life without looking for more on here. But I strongly believe the bugs haven't been worked out of WAR yet, and while I'm by no means a scientist, I can't believe that a metric with so many variables and intangibles would be accepted in any other field.

Again, a slightly different replacement level and the two different defensive components make up almost all of the differences in the two WARs for position players.

[Bothrops Atrox]

A 3 run homer when you are already winning by 6 runs is virtually menaingless while a groundout to second with no outs and a runner on second to advance a runner to third who then scores the winning run is MUCH more meaningful, we know which one WAR rates as superior

If you wanted to, you could take WAR as-is and add in a "clutch" or WPA component to handle what you are referring to. It is quite easy. Since "Clutch" (how a player performs in high leverage vs. low leverage) is already set at win above average, you can just add it in each year if you want.

[Bothrops Atrox]

Everybody...I am all for questioning and picking apart WAR. Just please don't talk about the lack of context in WAR and then trumpet a stat with even less context or no context whatsoever. That is all I ever want to see.

[9RoyHobbsRF]

run differential is not a good measure

see the 1997 giants 1987 twins etc

win games 3-2 and 3-1 and lose a game 7-2 and tell me how good run differential is