How is WAR defined as a baseball statistic?

Question

WAR (wins above replacement) is a statistic that is often used to describe the value of a baseball player's performance in a given year.

How is this statistic determined?
What are the differences between WAR as defined by the two major providers (fangraphs and baseball-referenced)?

Answer 1

WAR is a context used by leagues and ballparks independently to measure a baseball player's performance and relative value. As such, WAR can be used to compare players across different years and teams.

WAR is derived from weighted runs above average, or the number of runs a player contributes through both his offense and defense above and beyond your typical "replacement level" player (i.e. waiver-wire pickup or the next guy in AAA). You divide the total offensive and defensive runs above average by 10 to arrive at wins. Note that these numbers are also adjusted by position since playing the catcher position is harder to play than playing right field.

Both Fangraphs (fWAR) and Baseball Reference (rWAR) use many of the same statistics to measure a player's offensive and defensive contributions, with the following differences:

1. rWAR and fWAR both use FIP to measure a pitcher's performance, but rWAR also includes a pitcher's defensive runs saved as well.
2. To measure a player's fielding capability, fWAR uses UZR, while rWAR uses TotalZone, two very similar defensive metrics. However, UZR has only been calculated since 2002.
3. rWAR uses baserunning metrics, fWAR does not.

By the way, I've paraphrased the above from Fangraph's excellent intro to WAR, which I recommend you take a look at: http://www.fangraphs.com/library/index.php/misc/war/