The Most Important Thing I Learned from WAR

Wins Above Replacement isn’t a household statistic, and it never will be. It’s too complicated, too theoretical, too unnecessary for most baseball fans. With that said, today more people know about WAR than ever, with the metric routinely coming up on websites and on television. It is, rather understandably, controversial, and plenty of people out there dismiss it without a second thought. While it’s been around for years, I think what really launched WAR to the greater public was the first go-round of the Mike Trout vs. Miguel Cabrera MVP debate. That’s when WAR started getting a lot of play at places like ESPN, and more people were introduced to the framework and to some of its declarations.

Most of the ways WAR gets used are for comparative purposes. People love to see rankings, and people love to know which players are better than others. It also gets used in contract analysis, especially around these parts, and it’s hard to remember how we used to write those posts before we had the numbers we have today. With WAR, it’s really easy to get wrapped up in the details, because the metric allows for such detailed interpretation. But the one most important thing I’ve personally learned from WAR is probably the most general of its points. That is, WAR has provided me with an idea of what baseball players are worth to a team.

Take a player. Take any player who’s a regular, and go to one of his team’s message boards and write a little thing about his WAR. Some people are going to disagree with it. Some people are going to ignore you. But others, still, will engage you and tell you why the player is better or worse. Plenty of people disagree with individual WAR ratings, and that’s OK, because WAR is imperfect, for a variety of reasons. At least, the inputs are imperfect. The framework’s just fine. But anyway, while people don’t always see eye-to-eye on the individuals, I haven’t seen many people bicker with the idea that an average player is worth about two or three wins. It’s accepted that a good player is worth about three to five wins. It’s accepted that a star player is worth more than five wins. And the truly elite can get up around eight or 10. WAR, clearly, also is controversial. In general, people seem to accept its scales. Most of the fights are over decimal points.

For the longest time, that’s something I never had any concept of. I understood some players were better than others; I understood some players were fractional approximations of superior players. Getting actual values, though, changed almost everything. The absolute best player in baseball is worth about 10 wins over a 162-game season — relative to some ordinary Triple-A type. A good player could be worth a third of that. I’m floored by the number of ways in which this has informed my thinking.

The reality is that any given player means relatively little. Maybe that ought to be obvious: A team has several regular position players, and several regular starters, and an entire bullpen full of guys. Baseball, certainly, is a considerable team effort. But if you consider the goal, say, making the playoffs, a given player makes a difference of only some percentage points. Even big splashes can look kind of like small splashes. Small splashes can be virtually inconsequential splashes.

If a talented player has to go on the disabled list, obviously it’s more bad than good, but it can almost never be crippling. If the player’s only gone for 15 or 20 days, the damage might be almost entirely negligible. A 15-day DL stint represents one-twelfth of the regular season. Good players are worth a handful of wins over a full regular season. I doubt I need to lay this out.

I know I used to stress out about midseason moves, if I was following a contender, which has happened. Fans really don’t appreciate it when their teams allow the trade deadline to fly by without significant improvements being made. By definition, an improvement is an improvement and improvements boost odds, but by the middle of the year you’re looking at half a season left, and by the deadline you’ve got two months. It’s a narrow window for a player to make a difference, and such a limited impact can be worth only so many resources going in the other direction.

And there’s something to be said about organizational success windows. People think the Kansas City Royals have to contend this season, before they lose James Shields. People think the Seattle Mariners have to contend before Robinson Cano begins his decline. People think the Baltimore Orioles have to contend before they face difficult questions about some of their own stars. Windows do exist, to some degree. To use the Royals as an example: They’re closer to contention with Shields than they are without him; but even without him, they’re left with a similar team, just without one starting pitcher. If the Royals don’t make the playoffs in 2014, they don’t have to tear everything down because Shields became a free agent. They could try to build a contender again for 2015, and it would be far from impossible. A decent team with a good player is a mostly decent team even without him.

Writing this out, I’m left wondering whether it’s even been worth writing this out. It all sounds so simple and obvious to me, and that isn’t what FanGraphs is supposed to be about, except for when the next team signs Yuniesky Betancourt. We’ve accepted the point of WAR for years. But take a moment to appreciate how that’s had an effect on you and on your own fandom. I’m a completely different baseball fan than I used to be. The emotion can still be there for the games, but in between, I’m a lot more mild-mannered. I worry less about individual players, because individual players are cogs in a giant machine. Groups of moves can make a big difference, but particular moves often just make the needle wiggle a little bit. Seasons aren’t won and lost on singular transactions.

That is, unless we’re completely and hopelessly wrong about clubhouse chemistry and those sorts of intangibles. I don’t know — that’s a separate frontier, and I’m not among the explorers. Maybe players make an enormous difference after all. I’ll await the evidence.

I can see the way WAR has had an effect on my following hockey. I’m not so concerned about offensive defensemen being mediocre in their own end; and I’m not so concerned about defensive forwards being mediocre in the offensive zone. It has to be all about overall value, with all of the attendant pluses and minuses. That’s an example of a way in which WAR has crossed a boundary.

In baseball, though, as much as I like to use WAR for comparisons and evaluations, the most important thing it’s taught me is simply what average and good players are worth. And mediocre players, and great players, and all the players in between. It feels like a fairly fundamental lesson, but you’ll never get anywhere without the fundamentals.




Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


62 Responses to “The Most Important Thing I Learned from WAR”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Forest Gump says:

    Keep your feet dry. Like Lutenant Dan told me.

    Vote -1 Vote +1

  2. Randy says:

    WAR, Pigs.

    Vote -1 Vote +1

  3. Pre says:

    Isn’t the beauty in the sequencing? Sure, you might have a 3ish WAR true talent player but that doesn’t keep him from racking up the WPA. Or coughing it up. Those walkoff HRs with awesome bat flips might not be worth more WAR but they are worth so much joy/sadness.

    Vote -1 Vote +1

  4. Dan says:

    The Royals need to win something with James Shields because they traded long-term pieces to get him for the short-term. If you’re still mediocre in the short term, then the trade will have been for naught.

    Vote -1 Vote +1

    • skipperxc says:

      That’s just a sunk cost at this point. It may reflect poorly on management if their actions didn’t lead to the results they were looking for, but those actions are done. All they can do now is look at what they have and how to translate those best to wins, be it in the present or future.

      Vote -1 Vote +1

    • Jon L. says:

      Sometimes there may be effects we overlook. Maybe the Royals peak after James Shields is gone, in which case it looks like the trade was a mistake. But maybe without the intervening respectability, the team doesn’t develop the same way. Even leaving aside the intangibles entirely, maybe a player signs with a decent Royals team that has James Shields who would not have signed with them otherwise.

      A broader example of this is when a team that’s not yet competitive overpays for a star who may decline soon. I wonder sometimes if these might be important, positive moves, even in some cases when analyzing $/WAR makes them look mediocre or worse, (considered in a vacuum).

      Vote -1 Vote +1

  5. MK says:

    At the same time I think the simplicity discussed here demonstrates why WAR is incomplete in my mind. I think the interactions among individuals are the holes in WAR. Individually, Mike Trout has been superb, yet you’ll have a hard time convincing me that he actually contributed 10 wins last year. His outfield skills are less valuable (to the team’s overall success) with a high GB pitcher and/or low FB pitcher on the mound. His base running skills are less valuable (to the team’s overall success) when the hitter behind doesn’t work the count (fewer opportunities to steal), hits fly balls or home runs (fewer opportunities to go first to third). I think due to interactions he may have only contributed 5-7 wins, a superb effort, but he’s simply a cog in a poorly-oiled machine. While we can isolate hitting skills fairly well, I think we are missing the interactions for base running and fielding that may affect WAR. Alternatively, a GM may be able to design a team around a certain player and enhance their WAR beyond what we would expect to see (Pirates and Andrew McCutchen?). Losing a key player could harm the team more than WAR indicates due to interactions. James Shields does not get credit for the innings the BP does not pitch on his night. Shields allows the BP to pitch better the next night.

    Vote -1 Vote +1

    • David says:

      As I understand it, those interactions are all taken into consideration in WAR. Players are only credited for the defensive or baserunning opportunities they are given, which would cause factors that limit those opportunities (such as a GB pitcher or the hitters following him) to reduce his value and his WAR.

      Vote -1 Vote +1

    • Baltar says:

      What you say is basically correct. Nobody would claim that the Angels won exactly 10 more games than they would have without Trout. The calculation of exactly how many wins difference a player did make is impossible for all the reasons you state and more. After all, without that one player everything would have been different.
      Still, a measurement of each player’s worth in likely wins more than a border line major leaguer would have contributed is extremely valuable and interesting.
      Since I became a rabid major league baseball fan at age 9, I yearned for one statistic that would measure a player’s overall worth, considering as many things as reasonably could be included.
      When I found out about WAR about 50 years later, I was ecstatic and still am.
      It is not a perfect statistic. No statistic is. It will be improved with future research and technology. But, God damn, I love that we have it.

      +18 Vote -1 Vote +1

    • Jake says:

      Your discussion of Trout shows a misunderstanding of what WAR is measuring. WAR does not take someone’s true talent level and then project out how many wins that talent would produce. In fact, it does everything you say it does not. If Trout doesn’t get as many opportunities in the field because of an excess of groundball pitchers, that is reflected in his WAR through his not having as many opportunities to accrue value. If Trout has less opportunities to steal or to go first to third, that affects his WAR as well (as he again has less opportunities to accrue value).

      So, if you are correct that Trout has less opportunities in the field and on the basepaths than he should (questionable claims that require further investigation), then it would actually support the claim that his true talent level is GREATER THAN 10 WAR.

      +17 Vote -1 Vote +1

  6. Alvaro says:

    The most important thing I learned about WAR is that is purely theoretical, a 10 WAR player wont necesarily make his team 10 wins better.

    Vote -1 Vote +1

  7. hahiggins says:

    I’m pretty confident about WAR’s scale, but I’m even more confident in it as an ordinal exercise, especially with hitters. Pitchers I go back and forth between RA9 and FIP so it’s a little tougher, and the innings question isn’t easy because you get value for pitching an inning even though a more optimal mix for the team might not have given you that inning.

    Vote -1 Vote +1

  8. DanJennings says:

    Maybe a dumb question, but how many wins is a replacement player worth? If I have a team of all replacement players, how many games are they expected to win?

    Vote -1 Vote +1

    • chuckb says:

      I think it’s 41 or 42 games.

      Vote -1 Vote +1

    • harry says:

      a replacement level player would be worth 0 wins, bc he is replacement level and WAR is wins ABOVE REPLACEMENT. For a team of replacement level players, I believe the baseball changes each season but is about 48 wins

      Vote -1 Vote +1

    • Plucky says:

      A replacement level player byb definition is worth 0 WAR.

      The scale construction assumes that an all-replacement level team is expected to win 48 games (there used to be different scales, with fangraphs & bbref having different definitions of replacement level, but those were harmonized about 18 months ago). This is an assumption and not a proven fact (because it implicitly relies on the underlying assumption of talent distribution and fungibility embeded in “replacement level” as a concept) , but based on practical experience of extremely bad teams is generally accepted to be about right.

      Vote -1 Vote +1

    • Beasy Bee says:

      Please check me on this, but doesn’t the theoretical replacement player have (and I’m not talking WAR here) more than a 0 win contribution? I guess what I mean is, if the complete replacement squad gets 40-48 wins, then isn’t each player worth between 1 and 2 wins per season? Maybe more if its a position player versus reliever, maybe not, I’m not smart enough to figure that out.

      But I don’t bring that up trying to nitpick a number. I guess, to address Jeff’s statment about WAR being mass consumed, I feel like this general idea works against a more widespread understanding of WAR.

      Personally, I love it (as well as the feelings about the individual/team elements of the game Jeff intimates in the piece). I also kind of understand when another fan rejects the idea of Mike Trout being worth only 10 wins when he’s out there 162 games doing flashy things all over the field.

      And while saying that he’s now worth 12 wins would probably still undersell that fan’s idea of Mike Trout’s greatness, does it do anything to help that fan accept the contributions (we might say lack thereof) of a Lyle Overbay or Ryan Doumit to say they contributed a win or two last year in total?

      I don’t know, maybe not. Like most criticisms, I offer very little in terms of solutions.

      Vote -1 Vote +1

      • Preston says:

        The theoretical replacement level player is definitely the hardest thing to except about WAR. But when you try to come up with a different baseline you realize why it exists. If you did it as a measure of above or below average, an easily discernible and real number, then you aren’t properly allocating value. Under such a system an average everyday player would accrue no value. He could play 162 games and be worth exactly zero WAR. At the same time a bench player who has a heavy platoon split could collect only 150 ABs and if he’s better than league average in that span, than he would be measured more valuable than the guy who’s good enough to be an everyday player. This would obviously not be an accurate measure of a players worth. If you based it on an absolute zero model, saying that the replacement level team would win zero games then you are setting the bar immeasurably low. The sample size of a baseball season is too long and the variable in each contest to great to guarantee any team would lose all 162 games. A College or Rookie level MiLB team would probably win a handful of games over the course of a season. Assigning a positive contribution to a player who is good enough to stand on an MLB field and look like he has played baseball before shouldn’t really be the goal of a good measure of value (and I say this understanding that most of us would not be able to stand in the batters box, on the mound or in the field and accomplish it). Just looking at the Astros recent strategy to try and dig up AAAA gems we have seen that a team comprised mostly of guys considered non-MLB talent can still win some games at the MLB level. The purpose of replacement level is to attempt to judge at what point is a players performance at a level where he should be replaced with a AAA player who is readily available. And when you look at the alternatives I think it’s clear that this is the most accurate and more importantly the most meaningful way to measure value.

        Vote -1 Vote +1

        • Beasy Bee says:

          This was a great response, thanks! And again, I personally am down with WAR. As soon as I started spit-balling, I saw I wasn’t getting very far (anywhere really). I guess I just think about how I used to be into Vorp, now War….something’s likely to come next, no?

          Vote -1 Vote +1

      • Floyd says:

        No, because a replacement level team is worth 47-48 wins in 162 games (think a AAA team playing a full MLB schedule). What WAR measures is wins ABOVE that replacement level. Wins 49 thru 81 for a .500 team are a team’s wins ABOVE replacement level.

        Vote -1 Vote +1

  9. TKDC says:

    These discussions always leave me thinking about the differences in the economics and player values between MLB and NFL. If the NFL were more of a market-based league, wouldn’t a guy like Peyton Manning, who must be like a 5-6 WAR player (out of 16, so basically the equivalent of 50-60 baseball WAR per season) be worth just an incredible amount of money per year? I mean, the NFL does not at all work that way, but if it did, would a contract paying him over $100 million per season be so far-fetched?

    Vote -1 Vote +1

    • Plucky says:

      Football is a really hard sport to have WAR for, not just in the sense of being able to calculate win value, but also because players aren’t fungible in the way baseball players are. In baseball, when we talk about context depended stuff like GB/FB for pitchers with better or worse infields, we’re talking about half a win in a 162 game season. In football, difference in scheme fit can be the difference between a pro-bowler and a cut at some positions.

      On economics, in the context of a salary cap in the ballpark of $130m, elite QBs are getting paid in the neighborhood of $25-30m, 19-23% of the total for a team with a roster size more than twice as big as an MLB roster. That is not an obvious underpay.

      Furthermore, the salary considerations in baseball reflect the reality that a baseball team’s revenue is much more sensitive to its W/L record than a football team’s is. Despite all the TV money flooding into baseball, park revenue (tickets & concessions) is still a major slice (for most teams still the majority I think) of the pie. In football, TV money is the overwhelminly large source, and with stadium revenue most games are highly attended. The revenue difference in football between 10-6 (usually but not always in the playoffs) and 8-8 is nowhere near the revenue difference in baseball between 93-69 and 81-81.

      Vote -1 Vote +1

      • Cool Lester Smooth says:

        All of this.

        Football is almost wholly context dependent.

        Vote -1 Vote +1

        • Cool Lester Smooth says:

          Yeah, those stats aren’t good, though.

          Vote -1 Vote +1

        • TKDC says:

          Do you think there is a GM in football that doesn’t believe Peyton Manning would turn what would be an average team with a mediocre (replacement level) quarterback into a contender?

          The point is that the economics and structure of the NFL are so different, but if it were structure like baseball a guy like Manning would be worth just an unbelievable sum of money.

          Vote -1 Vote +1

        • Plucky says:

          The 2011 Broncos are exactly what you describe- an 8-8 team with a replacement level QB (Tebow). Manning probably would have gotten paid more if they knew he was going to be healthy when he signed his contract, but in not in the 40+ range. He’s great (and possibly the greatest ever) but he’s not so good that he could turn an all-replacement team into a playoff team. Stick him on the Jags or Bucs and they probably struggle to go 8-8. Nor is he so good as to be in a class all by himself. Brady, Brees, and Rodgers would also have extremely high WARs.

          It’s important to remember that replacement level in the NFL is way worse than it is in baseball, or at least that it’s much harder for an inferior team to beat a good one in football. An all-rep-level MLB team still expects to win more than 1/4 of its games. An all-rep-level NFL team is lucky to win 1/8 and very often will win none at all. Furthermore, a team with an all-rep level O-line will not get a full season from its QB (see 2012 Arizona). Football teams cannot win by fielding a great QB and nothing else. If you overpay for your stars and fill out your roster with scrubs, you’re the Dallas Cowboys and miss the playoffs for going on a decade. WAR for Manning is as high as it is becase the baseline for wins is 0 or 1. A baseball player who singlehandedly turns a rep-level team into a .500 one is 33 WAR, not 50-60. Both of course would be unprecedented, but your 50-60 baseball comp is off by almost a factor of 2.

          In terms of dollars, when you say ‘structured like baseball’ you essentially have to mean teams owning their local media revenue (which is a controllable thing), but even if NFL contracts were structured like baseball there are factors that would moderate salaries- in baseball, the guaranteed nature of contracts means teams bear the financial risk of long-term injury, whereas in football the player does. The Phillies can’t cut Ryan Howard and shed his remaining salary because his achilles will never be the same. In football that kind of thing happens all the time. If NFL teams had to build in that injury risk into their contracts, the length and dollar levels of contracts offered to even the top players would go way down. Teams just could not afford the risk of committing 20% of their payroll to a single person for 5-7 years with no way out given the injury rate in the NFL. It’s risky enough for a baseball team (again, see Ryan Howard) but would be insanely reckless for a football team

          Vote -1 Vote +1

      • Ruki Motomiya says:

        The hardest thing about advanced NFL stats is it is much harder to seperate an individual contribution in the NFL than the MLB. Not impossible, but it makes a good difference.

        Vote -1 Vote +1

  10. JMo37 says:

    WAR compares Mike Trout to an average replacement player. Meaning Mike Trout is 10x better than the league average.

    -18 Vote -1 Vote +1

  11. xeifrank says:

    Nice article. Is Buster Posey part of this giant machine you mentioned?

    Vote -1 Vote +1

  12. X says:

    I think it’s too pessimistic to say that WAR will not catch on with the general public. Consider the quarterback passer rating, which (a) sucks and (b) is complicated to compute. Despite these drawbacks, it was popular for a long time, because it captured in a single number something people care about: how good is my guy?

    Vote -1 Vote +1

    • Cool Lester Smooth says:

      Was popular? It’s still, along with sacks, TDs, yards and interceptions, one of the only football stats that most people use.

      Vote -1 Vote +1

      • jpg says:

        Funny thing is QB rating is actually not a bad stat provided you understand. It gets mocked because people don’t understand it or how to compute it themselves, much like WAR. If you have a QB rating north of 90 for the season, odds are you had great year. If it’s north of 90 for a single game, odds are you had a great game. The reason for this is simple: you need to throw TD passes, avoid interceptions efficientit big plays. If you aren’t doing all three, it’ll show in the rating.

        “QB ratings are a garbage stat. Elvis Grbac has a higher career QB rating than Joe Namath so it must be garbage”

        “Gerardo Parra has a higher WAR than Ryan Howard? Even with all those homers? That stat is garbage”

        Lesson here is that there is no such thing as a true end-all-be-all stat and context matters. QB rating is an efficiency stat. Grbac played in more pass friendly era and didn’t throw a lot of picks. Namath was one of the few gunslingers of his time and threw a lot of picks. He also played during a time that the fullback was the feature back and linebackers couldn’t run like track stars. Grbac was more efficient than Namath but it doesn’t make him better and he certainly wasn’t.

        Vote -1 Vote +1

        • Cool Lester Smooth says:

          Exactly. It measures all of the important stuff. It just needs to be adjusted to context.

          Vote -1 Vote +1

        • Visitor says:

          QB rating also does not include sacks, overrates interceptions, does not include rushing, does not include fumbles, and double counts completion percentage. It’s actually mocked as a bad stat for the reasons you’re saying as well, but people who DO understand the context of football hate it just as much for different reasons.

          If it “measured all of the important stuff,” it would be fine, but it doesn’t. It skips about half of it and weights what it does measure badly.

          Vote -1 Vote +1

        • Visitor says:

          And just so you know I’m not just someone who complains about all stats, adjusted net yards per attempt and ANY/A+ are excellent QB stats. There is still context to consider because that’s just how football works, but those actually do include everything important and weight them properly, unlike passer rating.

          Vote -1 Vote +1

  13. TK says:

    Fangraphs for hockey!!!

    Vote -1 Vote +1

    • Cool Lester Smooth says:

      Tom Tango has worked on an adjusted version of plus/minus.

      Vote -1 Vote +1

      • YABooble says:

        And the basketball plus/minus is already pretty well established. Love to see what Mr. Tango can do in light of this.

        Vote -1 Vote +1

    • stretchfest says:

      behindthenet.ca and extraskater.com have some pretty interesting analyses.

      They have basic (advanced) stats like Corsi and Fenwick, but also stats that control for garbage time, quality of competition, overall/PP/PK time on ice.

      It’s all cool stuff, but I recently read one of Grantland’s hockey writers say that hockey stats are 15-20 years behind baseball. So none of this stuff is close to gospel, but you have to start somewhere.

      Vote -1 Vote +1

  14. Ivan says:

    But every once in a while a midseason acquisition for a fringe prospect will play out and help in a big way (Scutaro for example). That’s why baseball is so much fun – it’s not all about the (saber) numbers, because if it was – it would be boring as hell! Not that I want to take anything away from the saber numbers, I am a frequent reader for a reason. :)

    Vote -1 Vote +1

  15. GilaMonster says:

    This is a very Nihilist view on WAR.

    Vote -1 Vote +1

  16. GilaMonster says:

    But in a more serious way, WAR does change your outlook on things. Before WAR, I feel like everyone want the big free agent. He was a star. He was going to carry your team. But using WAR, you see that Cano or Tanaka and realize they are very good, but you also realize having a non-replacement level outfield(Mariners) or infield(Yankees) is just as valuable.

    In makes you appreciate a well balanced deep team filled with role players.

    Vote -1 Vote +1

  17. channelclemente says:

    I’ve always wondered, if players are compared that have a quoted WAR value, what’s the variance in each of those values.

    Vote -1 Vote +1

  18. Billy says:

    Be more optimistic on the ability for people to understand WAR, Jeff!

    It is complicated, and some people may have trouble understanding exactly how it works; I didn’t understand how SLG% worked when I was 8 years old, but I grasped that it was a measure of power hitting. There might be some people who are a bit cynical or closed-minded with regards to WAR, but in time it may become pretty broadly understood.

    I am often a cynic about human nature and how dumb it seems people can be. But then I look at myself and realize I’m not particularly intelligent or strong-willed (try as I might), yet I understand this stuff. So if I can get it, that mean’s there’s definitely some chance that a great many people can. You might be right, Jeff, that WAR will never be a household statistic. But I think there’s also a legitimate chance that it does become one.

    And I doubt Jeff reads the comments down this far.

    Vote -1 Vote +1

  19. pft says:

    WAR is fine as long as you realize the following

    1. the error bands are large. Stats like 1 yr or less UZR are not regressed 50% as its creator demands. Park adjustments are crude and don’t take into account a batters/pitchers handedness and batted ball profiles. A RH FB pull hitter gets enormous benefit from playing at Fenway, a LH FB pull hitter gets killed at Fenway. Both get park adjusted the same.

    And there are different WAR’s, fWAR, bWar, etc which can give much different results for the same player

    2. its a counting stat

    3. it does not account for context or leverage. A clutch hitter for a team in a pennant race is treated the same as a player on a last place team

    4. the worst defensive SS in baseball gets a 2.5 win handicap over a DH. An inordinate amount of WAR differences between players at different position has nothing to do with performance but simply where their manager or team chooses to play them. Like RBI’s are opportunity drive, a position adjustment for WAR gives credit to players at positions who get more opportunities

    Vote -1 Vote +1

  20. Peter B. says:

    Is there an upper bound on WAR?

    Vote -1 Vote +1

    • Justin Bailey says:

      Not in an infinitely long season!

      Vote -1 Vote +1

    • Preston says:

      Yes and no, as a hitter you can’t compile more value than hitting a home run in every at bat, and as a pitcher you can’t do more than strike out every batter you face. So for a hitter 800 ABs and 800 homeruns would be the upper bound, obviously in this situation he’d get no base running credit as all he would ever do is jog, but he could also theoretically field every ball hit, which would give him astronomical value. But it is also theoretically possible for a player to get an infinite amount of at bats and fielding opportunities because games don’t have to end. If the Angels played 27 innings every night because they were constantly tied, Mike Trout could get 3 times the WAR (although he might also die of exhaustion). Practically it’s hard to imagine a player accumulating much more than Babe Ruth’s 15 WAR in 1923.

      Vote -1 Vote +1

  21. Tim says:

    I expect to read a hit piece on Orix any time now.

    Vote -1 Vote +1

  22. Dr Morris says:

    WAR has changed the way I view everything, from the performance of my colleagues at work to the performance of the barista at my local Starbucks. And, as the author suggests, the net effect is increased equanimity about the long haul.

    In WAR = Zen.

    Vote -1 Vote +1

  23. DNA+ says:

    My issue with WAR is that it is overly simplistic, rather than overly complicated. Most of the interesting things get lost by contracting everything down to one simple and error-ridden number. It is almost certainly the case that the current, or next, market inefficiency is in GM’s who understand WAR well enough to recognize when to believe it and when not to. An average or below average hitter with a high WAR because of outlier defensive measurements should be sold high. An above average hitter with low WAR because of outlier defensive measurements should be bought low. When comparing two players with equal offensive WAR components, you should favour the hitter who accrues WAR through less team dependent ways (power) if you’ve got a below average offense, while you should favour the hitter who accrues WAR through more team dependent ways if you’ve got an above average offense. …that sort of thing.

    Vote -1 Vote +1

    • Preston says:

      I don’t think this will happen. First any GM who places value in advanced statistics, looks at and understands the components of WAR. Second, the things that “we” aren’t as sure about, like fielding, GMs have far superior data on. So they are not likely to be “fooled” by inflated single season UZRs.

      Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *