BasicOverall
WAR
Wins Above Replacement
An estimate of how many fewer games a team would have won with a readily available fill-in, a replacement player, in this player's place.
Simple explanation
Hitting, baserunning, defense and the difficulty of the position all get converted into wins and added together. That's what makes it possible to put a shortstop and a designated hitter, or a starting pitcher and a closer, on the same list.
The baseline isn't an average player. A replacement player is the kind a team can pick up cheaply from the minors or the bench. A WAR of 0 doesn't mean average. It means anybody could fill the spot. The usual feel is that an everyday regular should be good for about 2 a season.
Benchmarks
- 0
- Replacement level
- 2
- Regular
- 4
- Core regular
- 5
- All-Star
- 6
- MVP candidate
Based on a full season of playing time.
Detailed explanation
The general shape: take the runs from hitting, baserunning and defense, add adjustments for position, league and replacement level, and convert those runs into wins. Roughly 10 runs make a win, and that ratio shifts somewhat with each year's run environment.
WAR is closer to the name of a method than to one formula. FanGraphs' fWAR values pitchers by FIP, and Baseball-Reference's bWAR uses the runs they actually allowed. As a result, a pitcher who had luck on his side often has the higher bWAR, and one who didn't often has the higher fWAR.
Defensive metrics take a long time to build up a reliable sample. In a single season's WAR, the defensive share carries an especially wide margin of error.
Common misconception
“A half-win gap in WAR settles which of two players was better.”
It's an estimate, so it comes with error. Over one season, players within about a win of each other are safer to treat as roughly equal.
Across leagues
KBOKBO WAR also differs depending on who calculates it, so the same player can come out with different values. When you quote a number, note whose it is.