Much, perhaps too much, is made of the multiple models existent that attempt to characterize a player's value relative to replacement level. "A man with one watch always knows the time, while a man with two watches can never be sure," is an expression. Most of the time, though, WAR, WARP, and WAR are close allies. How good is Joey Votto? you might ask. And the answers you get: 

  • At 5.0 wins, he was excellent! – Baseball Prospectus
  • At 5.6 wins, he was tremendous! – Baseball-Reference
  • At 5.9 wins, he was a handsome contributor! – FanGraphs

will all lead you to the same general picnic spot. You are a busy person and if you don't want to spend most of your time weighing competing assessments of major-league players, it it understandable, and you will rarely pay for narrowing your vision to just one, whichever one it is.

But sometimes these systems stray from each other by such a magnitude that they tell very different stories about players and seasons. In such a case, which system should you follow? It is a personal decision, one without necessarily a correct answer, but one that we are prepared to help you answer. What follow are a dozen timely disagreements between the models. In each case, simply pick the statement that seems *most* right to *you*; keep track of your answers. By the end, you will know something about yourself, namely which WAR(P) is right for you.  

1. This is a picture of B.J. Upton, who signed a five-year contract worth $75 million this offseason. 

a. B.J. Upton is a consistent All-Star-level performer, producing 13.8 wins in the past four seasons. 
b. B.J. Upton is better than average but not by much, producing 9.8 wins in the past four seasons. 
c. B.J. Upton is slightly worse than average, producing 7.2 wins in the past four seasons.

2. This is a picture of Zack Greinke, who became the highest-paid right-handed starter in history this offseason. 

a. Zack Greinke is a legit ace, producing 14.2 wins — seventh-most in baseball — over the past three seasons.
b. Zack Greinke is very good! He has produced 10.3 wins over the past three seasons.
c. Zack Greinke is above-average. He has produced 7.9 wins over the past three seasons.

3. This is a picture of James Loney, who signed a one-year, $2 million deal with the Rays. 

a. Sure. Loney has produced 3.7 wins over the past three years, so he has some value. 
b. LOL. Loney has produced exactly 0.0 wins over the past three years, so he has no value. 
c. Ehhh. Loney has produced 1.6 wins over the past three years, so he has done that.

4. This is a picture of Jeremy Hellickson. Hellickson was part of trade rumors this winter, but ultimately it was James Shields, not Hellickson, who was traded. 

a. He's an enigma! He has produced 2.9 wins so far in his career. 
b. The Royals probably didn't want him. In more than two full seasons, he has produced just 0.8 wins. 
c. The Rays probably wanted to keep him. He has produced 6.9 wins already in just over two full seasons in his career. 

5. This is a picture of James Shields. The Royals traded for him because they needed somebody who could start Opening Day. How good is Shields?

a. Real good! At 11.1 wins over the past three years, and 4.3 wins in 2012, he's among the 20 best pitchers in baseball. 
b. Not great! At 5.1 wins over the past three years, he's below average — including in 2012, when he produced 0.8 wins. 
c. A little better than okay. At 5.1 wins over the past three years, he's below average — but in 2012 he was solidly above-average, at 2.2 wins. 

6. This is a picture of Evan Longoria, whom the Rays extended for a long, long time. All the systems agree that Longoria is elite: 15.3 wins over the past three seasons on the low end, 17.3 on the high. But evaluating his extension means figuring out what the going rate is for third-base extension candidates. Longoria's deal is similar to Ryan Zimmerman's, signed a year ago. What sort of player is Ryan Zimmerman? 

a. Just about as good as Longoria: 14.2 wins over the past three seasons. 
b. Good, but probably two tiers worse than Longoria: 9.6 wins over the past three seasons. 
c. Reliably at All-Star level, but a tier below Longoria: 11.4 wins over the past three seasons. 

7. This is a picture of Ichiro Suzuki, who will reportedly return to the Yankees on a two-year deal. What sort of player is Ichiro these days? 

a. After a lousy 2011 season, he bounced back to be above-average in 2012. In total, he produced 3 wins in those two seasons. 
b. After a lousy 2011 season, he was barely better than replacement in 2012. In total, he produced -0.7 wins in those two seasons. 
c. After a lousy 2011 season, he was a little below average in 2012. In total, 1.9 wins in those two seasons.

8. This is a picture of Brandon League, who signed a three-year, $22.5 million. Most would agree it was an overpay, but by how much? 

a. League is at least a solid reliever, producing 2.5 wins over the past three seasons. 
b. League is totally replaceable, producing 0.4 wins over the past three seasons. 
c. League is pretty fungible but has value, producing 1.4 wins over the past three seasons. 

9. This is a picture of Ben Zobrist. The addition of Yunel Escobar allows the Rays to move Zobrist off shortstop next year. Zobrist is great! How great? 

a. A regular downballot MVP candidate: 16.4 wins over the past three seasons. 
b. A regular All-Star contender: 11.3 wins over the past three seasons. 
c. One of the half-dozen elite players in baseball: 18.2 wins over the past three seasons.

10. This is a picture of Brandon McCarthy. McCarthy signed for two years and $15.5 million with the Diamondbacks, about the same contract Joe Blanton got. 

a. An outright theft by Arizona! McCarthy has produced 6.6 wins over the past two years. 
b. A defensible move! McCarthy has produced 3.4 wins over the past two years. 
c. A relative bargain! McCarthy has produced 4.9 wins over the past two years.

11. This is a picture of Hanley Ramirez. The Dodgers have spent unthinkable sums in the past six months, but their chances in 2013 rest heavily on Hanley Ramirez. How good is Hanley Ramirez? 

a. Solidly above average: 3.0 wins in 2012. 
b. Slightly above average as a Dodger: 1.2 wins after the trade, and 1.4 overall in 2012. 
c. Solidly below average: 1.1 wins in 2012, and just 0.6 after the trade. 

12. And, finally, this is a picture of R.A. Dickey, who won the Cy Young award but can't get an extension worked out with the Mets. No wonder, given how much disagreement there is over Dickey among the advanced metrics. R.A. Dickey is either: 

a. Very good, if a bit overrated by Cy Young voters: 4.6 wins this year, 9.9 over the past three. 
b. Pretty good, if plenty overrated by Cy Young voters: 3.7 wins this year, 8.4 over the past three. 
c. Extremely good, a deserving Cy Young: 5.7 wins this year, 12.1 over the past three. 

Scoring is ridiculously simple. Count up your answers. What's the most frequent answer? 

If it's A, you're an fWAR Man/Woman.
If it's B, you're a WARP Man/Woman. Huzzah!
If it's C, you're a bWAR/rWAR Man/Woman. 

And if you find yourself with a three-way tie, congratulations. You are an intellectually curious sort who appreciates knowledge and insight whencesoever it comes.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
The whole article is great, but the pictures are definitely the best part.
Those pictures are frigging hilarious. I think Grienke or Ramirez are my favorite ones.
So much pathos in the Hanley shot.
You know what the Hanley picture reminds me of?

"Once again, the L.A.P.D. is asking Los Angelenos not to fire their guns at the visitor spacecraft. You may inadvertently trigger an interstellar war." - Independence Day

As far as pathos go, I can imagine Hanley being one of the leisurely marksmen.
Hanley has really let himself go.
So, this begs the question "why?"

What are the systemic differences in the three WAR systems that lead to these divergent results. Presumable it's things like:

- Different definitions of replacement
- Different ways of measuring defense
- Different ways of measuring pitcher performance (mostly related to the way in which they do or don't isolate defense and "luck")

Is it possible to identify the strengths weaknesses of each system as it relates to a certain type of player and pick which of the three is most likely the "best" measure?
Baseball Reference has a fanastic tabular breakdown of the differences between each metric:
This seems a bit backwards. Each implementation of the WAR framework makes slightly different assumptions and thus answers slightly different questions. One should start with figuring what question(s) one wants to answer and what assumptions one is comfortable with, and that can dictate which implementation to use. Starting with the answer and then searching for what question produces it works better in Jeopardy than sabermetrics.
I love the Hanley picture most of all.
Subjective-objective metrics. I like it.
Why is only B has Huzzah!? This must mean something.
Because WARP is a BP metric and the other metrics aren't BP.
The biggest career difference I've noticed is Tommy John, who I looked up after reading Adam Sobsey's recent article that mentioned him. He had a lousy FRAA, so BP has him at 12.7 career WARP, while B-R has him at 56.7 and Fangraphs 78.7.
Those will make excellent desktop wallpapers... tiled even!
Shouldn't this be a zero-sum game?

If one system assigns a player more wins than the otther system, it must assign another player(s) a corresponding amount less, so that all should come out with exactly 2430 wins per season (30 teams x an average of 81 wins each).

Has anyone checked to see if the arithmetic balances correctly in all three systems this way?
Sorry, I screwed up the math.

It should be 31 wins per team (above the replacement level of 50 wins) times 30 teams, equals 930 total WAR each season.
That assumes they all use the same replacement level, which isn't true, and further that the replacement level is adjusted empirically each year, which, e.g., at least per the chart posted above, fWAR does not do.
I am not a good enough mathematician to ever figure this out, so I will just ask it as a question:

However they are calculated, is there enough negative fWAR, WARP, and bWAR, in any given year, to keep the average from going over 81 wins per team? Because just from my cursory observation, that does not seem to be the case.
I found that was was a WARP feller for batters and a bWAR one for pitchers.

Note that the "evidence" is to some extent tainted, since one's answers will depend to a great extent on which sites/analysts one reads.
bWar wins for me and I visit that site the least! Interesting.