It’s impossible to sum up a Major League Baseball player’s value added into a single number, but WAR sure does a good job of trying.
WAR!
What is it good for?
Comparing Major League Baseball players from different positions or completely different eras by combining dozens of different data points into a formula that spits out a singular number.
That number is Wins Above Replacement (WAR), and it has in the past 10-15 years become the primary statistic/sabermetric cited in discussions about player value.
What all goes into that calculation, though? And why do the different sites that calculate WAR sometimes wildly disagree on a player’s value added?
We’ll do our best to explain a stat that a whole lot of people use, but only a handful have even the slightest clue how to calculate.
What Does WAR Mean in Baseball?
WAR is short for Wins Above Replacement. And the general goal of WAR is to quantify how much a player is worth by comparing everything that he does — at the plate, on the basepaths and in the field for position players; on the mound for pitchers — to what the expected contributions would have been if his playing time had been replaced by, essentially, a lifetime Triple-A player.
It’s an entirely theoretical number that is neither rooted in nor necessarily translates into actual team wins and losses. In fact, in 2024, the 92-win Cleveland Guardians ended up with a slightly lower teamwide fWAR than the 81-win Boston Red Sox. And goodness knows Mike Trout and Shohei Ohtani had some remarkably high WAR seasons for Los Angeles Angels teams that still lost a ton of games.
But, in theory at least, Trout’s 2016 AL MVP season for the 74-88 Angels was worth either 8.7 or 10.5 wins above replacement, depending on whether you prefer FanGraphs or Baseball Reference for WAR calculations.
Not only did that WAR play a key part in Trout winning the MVP, but that WAR allows us to compare his 2016 season to, say, Barry Bonds in 2004 posting a 10.6 Baseball Reference WAR (typically abbreviated bWAR or rWAR) or Babe Ruth in 1928 posting a 10.5 FanGraphs WAR (fWAR) when discussing the greatest seasons of all-time.
What Is a Replacement Player?
While the number of wins is what we all fixate on, we should probably start out with the concept of a replacement player.
Whether you prefer Baseball Reference or FanGraphs as your source of WAR, the good news is they agree on this part of the calculating process — at least since they unified this baseline in 2013.
The TL;DR definition of a replacement player is simply a theoretical person who would be called up from Triple-A (or still available to sign as a midseason free agent) should the team suddenly need a replacement player due to injury, suspension, etc.
It should be noted that this theoretical replacement player has nothing to do with the team’s finances, nor its current options either already on the roster or in its farm system. Whether it’s the Dodgers who can shell out money to replace an injured player, a team with the top prospect in all of baseball at that particular position or a team with neither of those luxuries is irrelevant when calculating WAR.
Rather, the assumption is that the team would be forced to call up a player who neither adds to nor detracts from the team’s ability to win. And WAR is, in a nutshell, how much better (or worse) than that theoretical player the player in question has been.
What B-R and FG agreed upon in 2013 was that an entire roster of these theoretical replacement players would be expected to win 47.7 games in a season, or a .294 winning percentage. (Comically, the 2024 Chicago White Sox only won 41 games.)
Moreover, those numbers were derived from the agreed upon standard that there should be 1,000 WAR per 2,430-game season at an MLB-wide level.
However, before we continue, it’s worth noting that Baseball Reference divides that 1,000 into 590 WAR for batters and 410 WAR for pitchers while FanGraphs goes with a 570/430 split.
There are plenty of other differences in their formulae that result in peculiarities, such as B-R saying Ronel Blanco was worth 4.5 bWAR in 2024 while FG puts his value added at 2.1 fWAR. But it might be useful to know that bWAR collectively shows a little more love to hitters and less love to pitchers than fWAR does.
How to Calculate WAR
If you’re sitting there with a paper and pencil hoping to calculate WAR on your own, we’ve got some bad news for you:
It is effectively impossible.
Oh, the respective formulae look simple enough, especially the Baseball Reference ones. For position players, the bWAR formula is:
bWAR = (Player Runs - Average Player Runs) + (Average Player Runs - Replacement Level Runs)
For a pitcher’s bWAR, it’s:
bWAR = Replacement Level + Wins Above Average + (Wins Above Average * (Leverage Multiplier + 1) / 2)
FanGraphs has a slightly longer formula for each, the “simple equation” as it words it in its library for position players is:
fWAR = (Batting Runs + Baserunning Runs +Fielding Runs + Positional Adjustment + League Adjustment +Replacement Runs) / (Runs Per Win)
And then for pitchers, it’s:
fWAR = [[([(League “FIP” – “FIP”) / Pitcher Specific Runs Per Win] + Replacement Level) * (IP/9)] * Leverage Multiplier for Relievers] + League Correction
However, that’s just a bunch of nonsense, right? We might as well have just told you that the ratio of Stanley Nickels to Schrute Bucks is the same as the ratio of unicorns to leprechauns, because those are ‘simple’ formulas in which each variable is the furthest thing from simple.
For example, in that FanGraphs position player formula, Batting Runs alone is: [((wOBA – lgwOBA)/wOBA Scale) * PA] + (lgR/PA – (PF*lgR/PA))*PA + (lgR/PA – (AL or NL non-pitcher wRC/PA))*PA. Which is yet another formula of formulas, wherein you need Park Factors (PF), the MLB-wide rate of runs per plate appearance for the timeframe in question, as well as the calculations for weighted on base-average (wOBA) for both the player and the league and weighted runs created (wRC).
And, again, that’s just for Batting Runs, which is one of the seven variables in the position player fWAR formula.
So, you know, good luck with your paper and pencil. It ain’t happening.
If you haven’t gone completely cross-eyed, though, the one noteworthy part to point out is the “Positional Adjustment” in the FanGraphs formula, which is also a key component of Baseball Reference’s Player Runs.
In fact, Baseball Reference has a handy-dandy list in its position player WAR explanation where it shows the current value assigned to each position per 1,350 innings played. (Each team will play roughly 1,450 innings over the course of a 162-game season.) That adjustment is +9 runs for catchers, +7 for shortstops, +3 for second basemen, +2.5 for center fielders, +2 for third basemen, -7 for corner outfielders, -9.5 for first basemen and -15 for designated hitters. [FanGraphs has similar adjustments, ranging from +12.5 to -17.5.]
Basically, if a full-time catcher and a full-time designated hitter put up the exact same batting stats in a given season, the catcher is going to have a considerably better WAR, provided he isn’t the most error-prone, no-value-added-on-defense catcher of all-time. This makes sense, since replacing a DH in the middle of the season is a whole heck of a lot easier than replacing a catcher.
What Is a Good WAR Rating?
This is where it starts to get a lot easier to comprehend.
About a million things go into the formula that spits out a singular number for WAR, but then we can take that number and say with some level of confidence how good the player’s season actually was compared to his peers.
A 0.0 WAR means the player provided no value, and either was a league-minimum salary player or might as well have been replaced by one. (A negative WAR suggests that the player should have been replaced.)
An average position player or starting pitcher who logs enough plate appearances or innings pitched to qualify for a batting crown or ERA title will have a WAR of around 2.0, which brings us to a brief aside about a similar metric: Wins Above Average (WAA).
Since an average everyday player / rotational cog is expected to have a WAR around 2.0, it stands to reason that WAA tends to be about 2.0 per season less than a player’s WAR. Case in point: Barry Bonds had a 22-year career with a bWAR of 162.8 and a WAA of 123.9, his WAR coming in somewhere between 1.4 and 2.2 greater than his WAA in each season with at least 450 plate appearances. Long story short, WAA is basically WAR minus 2. Now, let’s return to your regularly scheduled programming, already in progress.
Anything above a 4.0 WAR for a single season is an All-Star caliber campaign. However, most of the actual All-Stars in any given season will be below 4.0 at the time of the All-Star Game and may well fall short of 4.0 WAR by season’s end, so the correlation between the actual All-Stars and WAR All-Stars isn’t as strong as you might think. But there are typically around 40 position players who end the year with a 4.0 WAR or greater.
At 6.0 and above, now you’re talking MVP / Cy Young candidates.
One big thing to note is that WAR isn’t great for evaluating relief pitchers, as the best closers in a given season tend to end up somewhere in the 3.0 WAR range. Even when Eric Gagné won the NL Cy Young in 2003 during his streak of 84 consecutive successful saves, he was only worth 3.7 bWAR. Also, Mariano Rivera was evidently worth 39.1 fWAR for his career, which is slightly worse than Ryan Zimmerman (39.5).
Sure. OK.
See also: Baseball Statistics Explained
How Does WAR Impact Fantasy Baseball Strategy?
WAR is a great stat. It’s most certainly not the one-stop shop for assigning individual value that some in the MLB media tend to treat it as. But, for what it is, it’s a great tool that generally does a fine job of letting you know how good a player was.
Emphasis on that last word, though.
“Was.”
Past tense.
WAR is not a predictive metric. It can tell you how much value a player provided over a specific interval in the past, but it’s not much of a harbinger of what’s to come.
Now, there are some components of the WAR formula that are predictive in nature. Fielding Independent Pitching (FIP) is a major part of the pitching fWAR calculation, and that’s more predictive than ERA. Likewise, position player WAR uses Weighted On Base Average (wOBA) as opposed to batting average. Though, in both cases, using Expected FIP (xFIP) and Expected wOBA (xwOBA) would make it even more predictive, and would still only be part of the equation anyway.
Also, value added on defense is a huge part of the WAR equation for position players, but that is wholly irrelevant for fantasy baseball purposes. So if you do insist on looking at WAR in your fantasy draft prep, at least be sure to hone in on the Offense category on FanGraphs and the Offensive WAR (oWAR) category on Baseball-Reference, as those are at least what you’re aiming for.
Really, though, you should probably keep your distance from WAR for fantasy purposes.
Frequently Asked Questions
Why are fWAR and bWAR sometimes so divergent for the same player?
Simply put, they use different underlying, home-grown data.
Baseball Reference has OPS+ while FanGraphs has wRC+. B-R uses Defensive Runs Saved while FanGraphs leans on Ultimate Zone Rating. And while fWAR goes with FIP for pitching, bWAR’s backbone for pitching is Runs Allowed per 9 IP (RA9).
In each case, the goal of the stat is to sum up a player’s value added, be it at the plate, in the field or on the mound. But the subtle yet fundamental differences in those initial formulas can have something of a butterfly effect through the WAR calculations that ultimately results in considerable differences between fWAR and bWAR.
Usually, if you dig into the specific player’s profile, you can see why the split exists. We previously mentioned Ronel Blanco had a 2.1 fWAR and a 4.5 bWAR for 2024. Well, that’s largely because he had a 4.15 FIP and a 3.01 RA9, which is an unusually wide split. Meanwhile, MacKenzie Gore went in the opposite direction in 2024 with a 3.53 FIP compared to a 5.03 RA9, and he ended up with a 3.2 fWAR and a 0.8 bWAR.
(For what it’s worth, most feel that fWAR is better than bWAR for assessing pitchers.)
TBH, it’d be great if we could just normalize averaging a player’s fWAR and bWAR. That’s probably the truest single-number measure of a player’s value added.
Just don’t call it gWAR, though, or you’re going to confuse a lot of heavy metal fans.
What is the highest WAR in MLB history?
What about the worst WAR in MLB history?
I can only give you the FanGraphs versions of this answer, but the worst single-season fWAR by a qualified hitter was Jim Levey with a ghastly mark of negative-4.0 in 1933, while the low point for a pitcher with at least 100 innings pitched belongs to Leo Birdine at negative-2.1 in 1929.
For career worst, Bill Bergen takes the cake among position players with a negative-16.2 mark, in which he had a negative WAR in all 11 seasons played. Among pitchers who logged at least 700 innings, the worst fWAR belongs to Pat Mahomes at negative-3.0. At least his son can throw a football, though.
Take Your Fantasy Baseball Game to the Next Level with Sleeper!
Now that you know a little more about WAR, you can try out Sleeper, AKA the fantasy company that leads the industry in wins above replacement.
The app is top notch, and the fantasy offerings are plentiful, from breaking news, to Sleeper Picks to season-long leagues.
If you’re still relatively new to the world of baseball or just looking to expand your knowledge of the game, we’ve got all sorts of articles to help you out, from baseball rules for beginners, to an explanation of baseball statistics, the most popular teams, a primer on the positions in baseball and, of course, tips for playing fantasy baseball.
