WAR: assembling the whole player
Bat, glove, legs, position, playing time — one number. Here's the assembly, and here's why two sites disagree about it.
Wins Above Replacement is not a new measurement — it's the sum of everything you've already learned: batting runs (wRAA, Level 3) + baserunning runs + fielding runs (OAA-flavored, Level 4) + positional adjustment + replacement-level credit, all divided by ~10 runs per win. A 6-WAR season means: this player added about six more wins than a AAA fill-in would have.
Why fWAR and bWAR disagree
FanGraphs and Baseball-Reference use the same skeleton with different organs — most notably pitching (FanGraphs builds off FIP, B-Ref off runs allowed) and different defensive inputs. Same player, 4.8 vs 5.6, both defensible. The disagreement is a feature: it shows you which assumptions matter. (StatLab teaches the framework; adding those sources for a side-by-side toggle is on the roadmap.)
How to use WAR like an adult: 0 = replacement, 2 = solid starter, 4 = All-Star-ish, 6+ = MVP conversation. Treat gaps under ~1 WAR as ties, check what's driving big defensive swings, and never end an argument with 'because WAR says so' — end it with the components.
The formula (optional — skippable)
WAR ≈ (Batting + Baserunning + Fielding + Positional + Replacement runs) ÷ Runs-per-Win(≈10)
How this stat lies to you
- Defensive inputs are the noisiest part — single-season WAR swings on them; trust 3-year WAR more.
- Different sites, different WAR — always cite which flavor.
- It's context-neutral by design: a 3-WAR season of garbage-time stats and a 3-WAR season of pennant-race heroics count the same (WPA, next lesson, is the other lens).
Check yourself
1. WAR's components are…
2. Player A: 5.2 fWAR. Player B: 4.9 fWAR. The honest take?