PDA

View Full Version : Why should I care about WAR? Are there any faults in how it's calculated?



baseballguru3
07-17-2010, 05:36 PM
:confused:

TheGiantYankee
07-17-2010, 05:38 PM
:burn:

In reality though, I wouldn't mind learning more

chicagofan71
07-17-2010, 05:40 PM
Then look at the sabremetrics forum.

nymetsrule
07-17-2010, 05:41 PM
If you don't care about it, people will attack you and call you names.

Zaunnie
07-17-2010, 05:43 PM
War, huh, yeah
What is it good for
Absolutely nothing

stupidmop
07-17-2010, 06:02 PM
War is a combative state between two entities, generally spoken of in terms of violent conflict between nations. There are many faults in how it's calculated. For reference, see American Foreign Policy in the 21st Century.

Mudvayne91
07-17-2010, 06:06 PM
Any real answers.....?

Gigantes4Life
07-17-2010, 06:08 PM
You should care because it is the best known way to evaluate players on an individual and complete level. It combines offensive value, defensive value and baserunning, or it values pitching performance.

I wouldn't say there are any notable faults, but more in the methods used to calculate offense, defense, etc. (as you use other metrics for those).

It's simply one metric which combines the best of offense, defense and pitching.

WickedBadMan
07-17-2010, 06:15 PM
Biggest fault that comes to mind is various places can use different methods to calculate it, which can lead to discrepancies.

baseballguru3
07-17-2010, 06:16 PM
How on earth does Liriano have a higher WAR than Josh Johnson? That makes me believe this WAR stat is flawed in a noticeable way.

Gigantes4Life
07-17-2010, 06:20 PM
How on earth does Liriano have a higher WAR than Josh Johnson? That makes me believe this WAR stat is flawed in a noticeable way.

Well he strikes out more, walks a bit more, and is having a slightly unlucky season (high BABIP), and the AL is the superior league so any player there receives a slight bonus.

Obviously, if JJ was in the AL he'd face the DH rather than another pitcher every time through the lineup and that would weaken his stats a bit (not a lot, but significant enough to make a difference).

baseballguru3
07-17-2010, 06:26 PM
That still doesn't make any sense.

Liriano is not having the type of season JJ is having - nobody would say otherwise. They have over a 2 ERA difference..... yet I'm suppose to believe Liriano is having a better year?

Pffft, come on now.

Yendil
07-17-2010, 06:29 PM
WAR is a good stat, sure it has its minimal miscalculations, but generally only for pitchers since it uses FIP instead of tERA which is a better statistic to use than FIP. It doesn't even use xFIP just plain FIP.


Offensive players - Take wRAA and UZR and add them together. Add in a positional adjustment, since some positions are tougher to play than others. And then convert the numbers so that they’re based not league average, but on replacement level (which again, is the value a team would lose if they had to replace that player with a minor leaguer or someone from the waiver wire). Convert the run value you have to wins (10 runs = 1 win)


Pitchers – Where offensive WAR used wRAA and UZR, pitching WAR uses FIP. Based on how many innings a pitcher threw, FIP is turned into runs form, converted to represent value above replacement level, and is then converted from runs to wins.

Why tERA is better than FIP


say a batter with the bases empty hits a ball with x velocity and y trajectory and it lands at location z, tERA will have enough data (40 years worth) to determine how that play turns out with a completely average fielding team. Let’s say the aforementioned ball would drop in for a single 20% of the time, get past the defender for a double 45% of the time, and be caught for an out 35% of the time. This play would then be worth (.29)(.20)+(.45)(.49)-(.35)(.20)=.2085 runs against. Unlike statistics before it such as FIP that paid zero attention to the situation of the game, tERA will be able to level upon itself because each play will measure a pitchers true ability. Complicated linear weights will be used to calculate the tERA for the 2nd, 3rd, 4th, etc batter of each inning because tERA is a completely neutral statistic. In other words, the pitcher can be held 100% accountable for his actions as expressed by tERA. Let’s run through an example. If the first batter of an inning hits a ball in the exact same location with the exact same trajectory as the previous example, we obviously have a situation where there is a base runner on first 20% of the time, a base runner on second 45% of the time, and no base runner 35 percent of the time. To make the math a little simpler, let’s say that the next batter hits home run. That is a play worth 1.00 runs with bases empty, 1.74 with a runner on first, and 1.60 with a runner on second. We then have to take the result of the first batter and use leveling to get the true value of this second outcome. The math looks like this: (.20)(1.74)+(.45)(1.60)(.35)(1.00). This equals .348+.72+.35=1.418 runs against. In this situation, the pitcher has given up two runs, but because a truly average defensive team would have made the first play 35% of the time, the pitcher is only held accountable for 1.6265 runs.

And therein lies the beauty of tERA. Every pitcher will be judged against the exact same baseline, with regards to game situation (read: “clutch”), and with a highly perfected way to measure truly earned runs. tERA truly revolutionized the way baseball was looked at from an analytic perspective, and judging the value of pitchers has never been easier.

C1Bman88
07-17-2010, 06:34 PM
WAR is good because it's a total value metric. It combines offensive performance, defensive performance, a positional adjustment and an adjustment for playing time. There are variations in how it is calculated, of course.

Fangraphs, for example, is only offense + defense + positional adjustment.

Baseball-Reference uses offense (which includes reaching on error, staying out of the double play) + defense + baserunning + positional adjustment.

I calculate it differently- I incorporate situational hitting and park factors based on batter handedness, in addition to overall hitting, baserunning, defense, and the positional adjustment.

If there are faults, it's in the components. The concept of it is sound.

Gigantes4Life
07-17-2010, 06:53 PM
That still doesn't make any sense.

Liriano is not having the type of season JJ is having - nobody would say otherwise. They have over a 2 ERA difference..... yet I'm suppose to believe Liriano is having a better year?

Pffft, come on now.

That has nothing to do with WAR, simply the metrics that people choose to use.

Liriano isn't necessarily having a better year. There is a margin of error, and when players are that close in performance, you can give the edge to JJ.

Going forward, I wouldn't be surprised if they have similar numbers. The things pitchers can control the most (K and BB) are almost identical between the two pitchers.

baseballguru3
07-17-2010, 08:45 PM
So Joe Mauer if you go by advanced stats is pretty much the best player in the league, right? Or right there with Pujols? Since WAR doesn't factor in his gold glove defense and all.....

C1Bman88
07-17-2010, 08:56 PM
So Joe Mauer if you go by advanced stats is pretty much the best player in the league, right? Or right there with Pujols? Since WAR doesn't factor in his gold glove defense and all.....

WAR factors in Mauer's defense; at least part of it.

Gigantes4Life
07-17-2010, 08:58 PM
So Joe Mauer if you go by advanced stats is pretty much the best player in the league, right? Or right there with Pujols? Since WAR doesn't factor in his gold glove defense and all.....

It factors in his defense.

He's continually near the top each year however.

baseballguru3
07-17-2010, 09:26 PM
I thought WAR gave a defensive value on all catchers as just "average"....myabe that's another stat.

C1Bman88
07-17-2010, 09:49 PM
I thought WAR gave a defensive value on all catchers as just "average"....myabe that's another stat.

Fangraphs' WAR *used* to consider all catchers average defensively, but that's since changed. Unfortunately, the only thing they account for are SB/CS. The WAR on Baseball-reference, though, uses PB/WP, etc.

fvthreeee
07-18-2010, 12:06 AM
Liriano is the MOST unlucky, .361 BABIP, he doesnt walk anyone, strikes out a ton, and doesn't give up home runs. So look at it this way, His K's, BB's, and HR allowed FIP and xFIP , are saying he is a bona-fide stud, while his ERA say it doesnt. Which is wrong?

C1Bman88
07-18-2010, 01:36 AM
Liriano is the MOST unlucky, .361 BABIP, he doesnt walk anyone, strikes out a ton, and doesn't give up home runs. So look at it this way, His K's, BB's, and HR allowed FIP and xFIP , are saying he is a bona-fide stud, while his ERA say it doesnt. Which is wrong?

FIP and xFIP are going to predict his future line better than his current ERA. As you've pointed out, his BABIP is higher than usual, and chances are it'll regress closer to his mean. That means he'll most likely shave off a bit of that ERA, unless he doesn't perform as well in the second half.

Gigantes4Life
07-18-2010, 02:20 AM
FIP and xFIP are going to predict his future line better than his current ERA. As you've pointed out, his BABIP is higher than usual, and chances are it'll regress closer to his mean. That means he'll most likely shave off a bit of that ERA, unless he doesn't perform as well in the second half.

For some reason I think he understands that and was asking a rhetorical question.

I could be wrong though.

C1Bman88
07-18-2010, 04:26 AM
For some reason I think he understands that and was asking a rhetorical question.

I could be wrong though.

Sometimes I hate the internet. It's not always easy to tell when someone's being facetious or asking a rhetorical question.

Jeffy25
07-18-2010, 04:56 AM
That still doesn't make any sense.

Liriano is not having the type of season JJ is having - nobody would say otherwise. They have over a 2 ERA difference..... yet I'm suppose to believe Liriano is having a better year?

Pffft, come on now.

Personally, I don't like WAR for pitchers

I love it for position players

Gigantes4Life
07-18-2010, 05:37 AM
Sometimes I hate the internet. It's not always easy to tell when someone's being facetious or asking a rhetorical question.

Sometimes. But the internet has everything! Can't stay mad at it.

I just figured he understood when he said


Liriano is the MOST unlucky, .361 BABIP, he doesnt walk anyone, strikes out a ton, and doesn't give up home runs.

Seems like he gets it. :)

baseballguru3
07-18-2010, 11:00 AM
Can't you turn BABIP around and say his is so high because he gets hit hard a lot? And that good pitchers don't allow batters to get good wood on the ball?

Toirtap
07-18-2010, 11:13 AM
The most important thing to know about WAR is what goes into it--how offense is measured, what the positional adjustments are, what is used for the replacement level, etc.--not any one site or person's specific implementation. Allowing one specific implementation of WAR to discredit other implementations is like declaring that all pizza is bad because you don't like anchovy pizza.

Truth be told, I don't care much for Fangraphs pitcher WAR either. Using FIP makes it something of a forward-looking metric rather than a retrospective valuation, but it is the only element of their WAR done that way. While pitcher BABIP is being de facto regressed, UZR is used with no regression at all, leading one of their writers to aggressively state that Nyjer Morgan was an equally valuable player than Adam Dunn last year (I'm not saying that this *couldn't* have been true, just that the degree of confidence with which the statement was made was excessive).

The worst thing that could happen is for one version of WAR to be universally accepted and taken as an absolute number, sort of like BA/HR/RBI in the traditional mindset. Uncertainty and continual attempt to improve metrics are two of the most important sabermetric values.

C1Bman88
07-18-2010, 01:34 PM
Can't you turn BABIP around and say his is so high because he gets hit hard a lot? And that good pitchers don't allow batters to get good wood on the ball?

If the pitcher is allowing an inordinate amount of line drives, then sure. But Liriano's inducing a ton of ground balls and...really, with his stuff, does it seem like he could ever really be hit hard?