What is WAR, anyway? (Split from Lucas Duda)

[Frayed Knot]

Benjamin Grimm wrote: That might have inspired me to actually learn what exactly a WAR is!

A new-fangled stat designed to estimate Wins Above Replacement
Joe Morgan, of course, doesn't like it and he and other curmudgeons have, in protest, adopted the slogan; "War, what is it good for? ... Absolutely nothin'!!"

[Ceetar]

Edgy DC wrote: I was thinking I should have asked people to guess the number of WARs each guy would contribute.

Duda? .2

[Benjamin Grimm]

Frayed Knot wrote: Benjamin Grimm wrote: That might have inspired me to actually learn what exactly a WAR is! A new-fangled stat designed to estimate Wins Above Replacement Joe Morgan, of course, doesn't like it and he and other curmudgeons have, in protest, adopted the slogan; "War, what is it good for? ... Absolutely nothin'!!"

My ignorance comes not from contempt, as with Morgan, but from laziness. I did know what the acronym stands for, but I have no idea how it's determined, or what would be considered a good number to have.

Some day, perhaps, I'll take a closer look at this stuff. If it had come along earlier, like in the 1980's, I would have been all over it. At that time, a new baseball stat would have been as exciting to me as a new planet would be to an astronomer.

[Frayed Knot]

I really just answered in order to get the joke in.

How to calculate? -- probably only about 5 people on the planet really know and I'm sure there are several variations, but it's supposed to take in the total contribution from a player both offensively and defensively and convert that into the number of extra wins your team would see above what you'd get if you filled his position with the level of player usually available for little cost.
As far as what's a good number to have? -- A [u:1569z8ty]GREAT[/u:1569z8ty] player might net you an extra 10 wins in a season above what your team would get if you replaced said superstar with whatever scrub was available off the waiver wire. Most will be a whole lot less than that; really bad ones could end up in the negative.

A lot of detailed stat sites now list some version of that stat for viewing pleasure.
Wright's best season (just to choose an example) was 7.8 in 2007. He dropped to 2.4 in 2009 then bounced back to 3.9 this year.

Edited January 14, 2011 by Guest

[Guest John Cougar Lunchbucket]

Benjamin Grimm wrote: Benjamin Grimm wrote: That might have inspired me to actually learn what exactly a WAR is! A new-fangled stat designed to estimate Wins Above Replacement Joe Morgan, of course, doesn't like it and he and other curmudgeons have, in protest, adopted the slogan; "War, what is it good for? ... Absolutely nothin'!!" My ignorance comes not from contempt, as with Morgan, but from laziness. I did know what the acronym stands for, but I have no idea how it's determined, or what would be considered a good number to have. Some day, perhaps, I'll take a closer look at this stuff. If it had come along earlier, like in the 1980's, I would have been all over it. At that time, a new baseball stat would have been as exciting to me as a new planet would be to an astronomer.

I harbor at least a little suspicion of all stats, but yeah, "what would be considered a good number to have" is what trips me up with WAR. I mean, I don;t need it explained to me, I'm just not very conversant in it, it's a pretty compact field.

[metirish]

I'm in the same boat as Grim.....maybe even more lazy though. I can't even pretend I know enough about it to drop it in conversations.

[Guest Edgy DC]

There are more than one way of calculating so it's fuzzy, and it depends on how much you play, but 5.0 is a surefire All Star. Seven is a potential MVP.

David Wright was a 3.9 last season (accooriding to BB-R) and Angel Pagan a 4.8 (2.2 of that with the glove). Ike Davis was a 2.5 and Resisting Analysis Dickey went for a 4.7.

[Benjamin Grimm]

So what was the Mets best WAR season? I'll guess Carlos Beltran in 2006 for offense and Tom Seaver in 1971 for pitchers.*


*These are wild-assed guesses, by the way.

[Guest Edgy DC]

I don't know, but I can say that Seaver '71 is wrong.

[Guest attgig]

Starts in Buffalo. He'll showcase his power, but not much else, but enough that it draws interest of our front office and many others.

at the end of june, gets called up to DH vs rightines, conveniently before the trading deadline. However, he ultimately flops during that callup, showing little patience, but hits a towering home run that is the farthest hr of the year by a met - his only hit of the road trip.
after that AL roadtrip, gets sent back down, and plays in the AAA all star game, hitting the game winning home run and becoming the all star game mvp.
Once beltran gets traded, he gets called up to play RF and occasionally spell Bay & Davis when they need a day off. It soon becomes apparent that his righty/lefty splits are horrible, and Nick Evans soon joins to become a platoon partner.

Final stats:
46 games
7 HR
45 SO
.209 Ave
.460 Slg

[Ceetar]

Benjamin Grimm wrote: So what was the Mets best WAR season? I'll guess Carlos Beltran in 2006 for offense and Tom Seaver in 1971 for pitchers.* *These are wild-assed guesses, by the way.

http://www.baseball-reference.com/teams/NYM/
sort by top player

Gooden in '85, 12.5 WAR

Seaver was 9.7 in '73 and 9.6 in '71

Top offense, Olerud in '98 ties with Gilkey in '96 with 8.1 (Beltran 8.0 in '06)

Pete Falcone let the '81 Mets with 2 WAR

[Guest attgig]

Benjamin Grimm wrote: So what was the Mets best WAR season? I'll guess Carlos Beltran in 2006 for offense and Tom Seaver in 1971 for pitchers.* *These are wild-assed guesses, by the way.

depends on who's war calculations you go by. the main difference is in calculating the defense.
according to baseball-reference, it's 85 gooden (12.5) and a tie of olerud's 98 and gilkey's 86 (both at 8.1)

but if you look at fangraphs,
olerud's 98 season is an 8.4 WAR and gilkey's 96 is a 7.6
meanwhile wright's '07 is a 8.6 WAR (7.8 according to baseball-reference).
there's no good way to filter by team on fangraphs, but i'm going to guess that it's wright's 07 according to fangraphs.

that make it any clearer? :-P

on edit:
there is a way to check on fangraphs for non pitchers:
http://www.fangraphs.com/graphswd.aspx?teamid=25&pos=All&season=2010&season1=1871&grid=25
it orders by total, but you can see that wright's season is the highest. nobody else as a 9 or higher anywhere.

Edited January 14, 2011 by Guest

[Guest Edgy DC]

Ceetar wrote: Gooden in '85, 12.5 WAR

That's 11.7 with the arm, 0.8 with the bat, and 0.0 with the glove.

To think, that's four seasons removed from Pete Falcone's 2.0 pacing the team.

You don't really need the numbers to tell you, but that season was a volcanic eruption of goodness.

[Ceetar]

Edgy DC wrote: Ceetar wrote: Gooden in '85, 12.5 WAR That's 11.7 with the arm, 0.8 with the bat, and 0.0 with the glove. To think, that's four seasons removed from Pete Falcone's 2.0 pacing the team. You don't really need the numbers to tell you, but that season was a volcanic eruption of goodness.

Well, i kinda do, as I was 3 at the time, but yeah.

.8 with the bat? heh, he probably could've batted cleanup on that '81 team.

[Benjamin Grimm]

Edgy DC wrote: You don't really need the numbers to tell you, but that season was a volcanic eruption of goodness.

I've never seen anything like it. Every start Doc made that year was an Event with a capital E. Actually, a capital V, E, N, and T as well.

[Guest attgig]

and I just come to find out fangraphs does their pitching different as well....

gooden's 85 is a 9.0WAR
and they don't have war for seaver for before 1980...

[Guest attgig]

Well, i kinda do, as I was 3 at the time, but yeah.

heh. I was 6, but was the only person in my family that watched baseball, and being the youngest, didn't really get to watch much baseball...

so, yeah, I'm with you too.

[Guest attgig]

attgig wrote: Starts in Buffalo. He'll showcase his power, but not much else, but enough that it draws interest of our front office and many others. at the end of june, gets called up to DH vs rightines, conveniently before the trading deadline. However, he ultimately flops during that callup, showing little patience, but hits a towering home run that is the farthest hr of the year by a met - his only hit of the road trip. after that AL roadtrip, gets sent back down, and plays in the AAA all star game, hitting the game winning home run and becoming the all star game mvp. Once beltran gets traded, he gets called up to play RF and occasionally spell Bay & Davis when they need a day off. It soon becomes apparent that his righty/lefty splits are horrible, and Nick Evans soon joins to become a platoon partner. Final stats: 46 games 7 HR 45 SO .209 Ave .460 Slg

I think my lucas duda projection got split into the war thread....

[TransMonk]

[Guest LeiterWagnerFasterStrongr]

WAR primer here: Nutritious... and tasty, too!

Positional adjustments are key, as well-- Chase Utley's stats at second make him an MVP candidate, but if he moves to the outfield, he's Josh Willingham or Corey Hart with an on-base bonus.

[Benjamin Grimm]

attgig wrote: I think my lucas duda projection got split into the war thread....

Oops... sorry. You may want to paste it back into the Duda thread, for the sake of posterity.

[metirish]

This thread if anything has complicated WAR, or I should say shows how complicated it is.

The great Mets infield of Ordonez, Olerud, Ventura and Alfonzo must have had a great collective WAR?, is that even possible?

[Ceetar]

metirish wrote: This thread if anything has complicated WAR, or I should say shows how complicated it is. The great Mets infield of Ordonez, Olerud, Ventura and Alfonzo must have had a great collective WAR?, is that even possible?

They had a combined (b-r) 8 defensive WAR.

20.4 overall (although Rey was negative offensively) The team had a 25.4 WAR via batting. McRae and Bonilla combing for a -4.2 WAR without even a full season of AB. Although how McRae managed a -2.2 Defensive WAR probably speaks to the credibility of the defensive stats.. I feel like I'd put up better than that.

[Willets Point]

Okay, which one of you just posted this at MetaFilter?

[MFS62]

I look at WAR like I look at a Calculus equation.
It holds up in theory. But in Calculus, you can't get an answer until you replace the symbols with real numbers.
In the case of WAR, it isn't always a good way to calculate how much losing a player will mean to your team. Or how valuable the player who you lost was, compared to the real player who replaced him.
For example, Wally Pipp got hurt. And the guy who replaced him was Lou Gehrig.
I'm all for new baseball measures. But I just can't get my head around this one.

Later

[metsmarathon]

the hangup that i think is common is the idea of the replacement player. we're tempted to thik of who exactly that replacement player is, or could be.

the truth of the matter is, that the replacement player is a theoretical construct. a baseline against which to measure.

the replacement player is imminently replaceable and interchangeable. he's what you'd expect to get out of any random AAA callup, not your top prospect or your first round draft pick. but the guy you call on after your other three backup shortstops get hurt. he's your yardstick.

but why do you need him? well, we're rolling things up into wins, and how much a given player contributes to winning. you hear about it all the time in the mvp discussions - how good would the yankees be without jeter? how good would cincy be without votto, or st. louis without pujols.

well, if you pulled derek jeter off the field, you'd leave a gaping hole at shortstop wouldn't you? well, ok. maybe it wouldn't be too noticeable at first. so you need a warm body to filll the void. that warm body is the replacement player. and instead of having an 8-man lineup, that warm body also gets to stand in the batters box and hit soft pop flys into the friendly confines of mfys3.

think of him as your bare minimum expectations for a ball player. roster fodder. if you got him, you'd want to throw him back.

last year's analagous replacement player, (using fangraphs) was jeff francoeur, with a fancy 0.1 WAR. the year before, was alex cora, who sported a solid 0.0 WAR. for a pitching analogue, think 2010's raul valdes, or 2009's tim redding, both of the 0.0 WAR.

...

i'd look at it like you would temperature. how hot is something? water freezes at 32�F, room temperature is about 72�F, while body temperature is 99�F. in fahrenheit, that baseline is fairly arbitrary. but we all know what we're talking about.

the replacement player is the same baseline.

when we ask 'how hot is it outside', we compare against the baseline, and get a number. we don't then ask, 'well, what is next tuesday going to be like,' before we decide to put on a hat.

when we ask 'how good is david wright' we compare against the baseline, and get a number. we don't instead ask, 'well, how good is mike hessman?'

[metirish]

MM just explained it in terms I understand......thank you.

Edited January 15, 2011 by Guest

[Frayed Knot]

MFS62 wrote: I look at WAR like I look at a Calculus equation. It holds up in theory. But in Calculus, you can't get an answer until you replace the symbols with real numbers. In the case of WAR, it isn't always a good way to calculate how much losing a player will mean to your team. Or how valuable the player who you lost was, compared to the real player who replaced him. For example, Wally Pipp got hurt. And the guy who replaced him was Lou Gehrig. I'm all for new baseball measures. But I just can't get my head around this one.

WAR is nothing more than an estimate so it's always going to be theoretical.
That's also the reason you compare it not to who actually is replacing the player you're losing but against an agreed-upon norm (the concept of the replacement player), so that different players are compared evenly.

You can't say, for instance, that Pujols would rate as less valuable than Ryan Howard simply because when the Cards lose Pujols next year they manage to replace him with Teixeira while the Phils used an untested rookie to fill their hole which made their drop-off much larger.

[MFS62]

Thanks, folks.
'Thon gave me an idea.
Why not, after calculating the "replacement player" numbers, (I'm supposing they change every year) you give him a name?
Or, if they calculate WAR by position, names. It would be the current player whose numbers came closest to the WAR baseline calculation.
It would make it a lot easier to picture if I knew this guy is 1.7 wins better than Elio Chacon. Or 2.9 wins above Dave Schneck. Well, you know.

Later

[Guest LeiterWagnerFasterStrongr]

Think Alex Cora (career Baseball-Ref WAR of 0.9 or Fangraphs WAR of 4.5 over 1182 games).