WAR is really just an idea of how to value players. Once you have the idea in place, there are dozens of implementation details that have multiple reasonable approaches to take. We've summarized the major frameworks below for easy comparison.
FanGraphs also has an explanation of their WAR and our previous WAR framework. Baseball Prospectus has their framework, WARP. I'm happy to add more WAR frameworks like Baseball Gauge's WAR if folks want to help me fill in the details below.
| Name | Previous B-R WAR | FanGraphs WAR | Updated B-R WAR | Prospectus PBP WARP | Prospectus Historical WARP |
|---|---|---|---|---|---|
| Abbreviation | rWAR (for Rally WAR, "RallyMonkey", Sean Smith created this version) | fWAR | rWAR, maybe bWAR? | WARP | WARP |
| Offensive Batting Metric Used | BaseRuns based offensive measure, so sum of players batting runs equals team total. | wRAA | wRAA with additional tweaks | Empirc LWTS, plus situational runs (DP, SH, SF) | BaseRuns-derived LWTS |
| Fielding #'s Used | Total Zone Ratings using PBP when possible. | TZR pre-2002, Ultimate Zone Rating 2002 and after | TZR pre-2003, Baseball Info Solutions Defensive Runs Saved with Batted Ball Timer 2003 and after | FRAA | FRAA (non pbp) |
| Starting Point for Pitching WAR | Runs Allowed | Fielding Independent Pitching | Runs Allowed | Runs Allowed | |
| Batting Details | |||||
| Adjusted for park | Yes | 5-year regressed park factors | 3-year park factors | Regressed component factors converted into run-based PFs (up to 5 years) | |
| Park Factor Applied to Batter or League Average | Batter | Batter | Batter | League | League |
| Computes Batting Lg Avgs using entire MLB or individual leagues | Ind. Leagues | All of MLB | Ind. Leagues | Ind. Leagues | |
| Includes P's in Lg Avg | No | Yes | No | No | |
| Includes ROE | Yes | No | Yes | Yes | |
| Estimates CS for seasons it is unknown | Yes | No | Yes | -- | |
| Includes SB/CS in Batting or Baserunning | In Baserunning | In Batting | In Baserunning | In Baserunning | |
| Includes Positional Adjustments | Yes | Yes | Yes | Yes | |
| Varies Replacement Level by Quality of Competition | Yes | No | Yes | Yes | |
| Caps total contribution by Replacement Level Runs to prevent overvaluing of leadoff hitters | Yes | No | No | No | |
| Counts SO and non-SO outs separately | Yes | No | Yes | Yes | |
| Directly uses the PythagenPat WL% estimator to compute Wins Above Avg | No | No | Yes | No | |
| Runs to win calculation includes the league run environment | Yes | Yes, but not described for batters | Yes | Yes | |
| Runs to wins calculation includes the runs the player adds or subtracts | No | No | Yes | No | |
| Differentiates between infield and outfield singles from 2003 on. | No | No | Yes | All Seasons | |
| Adjusts for ability to avoid double plays | Yes | No | Yes | Yes | |
| Adjusts for Non-basestealing baserunning | Yes | Yes | Yes | ||
| Available in a form with fielding metrics excluded | Yes, oWAR is WAR with defensive runs assumed to be zero or average | No | Yes, oWAR for offensive WAR | VORP | |
| Fielding Details | |||||
| Uses measures for player range | Yes, TZR | Yes, UZR or TZR | Yes, DRS or TZR | Yes, FRAA | |
| Uses measures for OF arms and DP ability | Yes | Yes | Yes | Yes, BRR Against | |
| Uses measures for catcher throwing | Yes | Yes, DRS | Yes, DRS | Not yet | |
| Uses measures for good plays and misplays like relay throws or missing the cutoff man | No | No | Yes, DRS | No | |
| Uses measures for catcher defense including blocking pitches and framing | No | Yes on blocking, but no on framing | Yes, DRS | No | |
| Pitching Details | |||||
| Years Available | 1871-present | 1974-present | 1871-present | 1950-Present | |
| Uses Park Factors | Yes, 3-year | Yes, 5-year, regressed | Yes, 3-year | Regressed component factors converted to run-based PFs (up to 5 years) | |
| Adjusts Park Factors to parks actually pitched in | No | No | Yes | Yes | |
| Park Factor applied to pitcher or run support? | Lg Avg Pitcher | Pitcher | Lg Avg Pitcher | Run Support | Run Support |
| Adjusts league average by teams pitched against | Yes | No | Yes | No | |
| Adjusts league average for interleague play | No | No | Yes | Yes | |
| Adjusts replacement level by quality of league | Yes | No | Yes | Yes | |
| Recent replacement level | .320 (52-110) | .265 (43-119) (link) | .320 (52-110) | ||
| Directly uses the PythagenPat WL% estimator to compute Wins Above Avg | No | No | Yes | No | |
| Runs to win calculation includes the league run environment | Yes | Yes, but not described for batters | Yes | Yes | |
| Runs to wins calculation includes the runs the player adds or subtracts | No | Yes | Yes | Yes | |
| Adjusts runs to win calculation by length of pitcher's outings | No | Unclear, I believe no | Yes | Yes | Yes |
| Adjusts runs above replacement level for when pitcher is starter or reliever | Yes | Yes, as part of replacement level calculation | Yes, but it is applied to avg rather than replacement | Yes | Yes |
| Adjusts performance by team defense | Yes | Not applicable since FIP-based | Yes | Yes | Yes |
| Adjusts Reliever WAR by Leverage Index | Yes | Yes, I believe it is applied to the WAR value | Yes, Applied to WAA and then recentered to not affect league total WAA | No | No |
| Credits starting pitcher for effect on saving/taxing bullpen | No | No | No | Yes | |
| Other Details | |||||
| Converts WAR into FA Salary Values | No | Yes | No | No | |
| Downloadable | No | Yes, annual and custom reports dowloadable to excel | Yes, full historical csv updated daily and usable by anyone with citation and link | Yes, via sortables | No (but eventually yes) |