By In Stuff

Talking about the V in MVP (again)

KANSAS CITY — Let’s start with this: Wins Above Replacement – that famous WAR statistic that has inspired so much war in the baseball community – has been manna for Mike Trout fans the last two years. It utterly confirmed what they (me among them) knew about Trout.

1. He was the greatest player in baseball.

2. He was one of the greatest young players in baseball history.

3. He absolutely, entirely and thoroughly deserved to win the MVP award over MIguel Cabrera, despite their differences in the three boring statistics that had been the lifeblood of baseball for too long: Batting average; home runs; RBIs.

WAR so perfectly illustrated what many of us believed about him and the game – that his all-around game simply overpowered Cabrera’s Triple Crown brilliance. They both hit for high averages, Cabrera some points higher. They both hit for power, Cabrera slugged more. But we felt sure that when you added in Trout’s huge advantages in defense and base running and the extra walks he drew and the double plays he did not hit into and the extra runs he scored, well, he was comfortably ahead as a player.

WAR confirmed this for us even if the MVP voting went the other way.

2012 WAR

Trout: 10.8 (Baseball Reference); 10.1 (Fangraphs)

Cabrera: 7.2 (refwar); 6.9 (fanwar)

2013 WAR

Trout 8.9 (refwar) 10.5 (fanwar)

Cabrera: 7.5 (refwar); 7.6 (fanwar)

So plain to see, right? WAR confirmed what we knew to be true – that a complete game like Trout’s was simply more valuable than a brilliant but incomplete game like Cabrera’s. Sure it was just one statistic — and I still remember future GM Farhan Zaidi telling me that the Oakland system actually rated Cabrera’s seasons ahead of Trout. But WAR just so vividly expressed those things about the game that we just knew had to be true, and most Trout fans used WAR liberally.

Fast forward to yesterday … and a little post I threw together about Alex Gordon as MVP candidate. My point in it was not to make Gordon’s MVP case (I’ll do a bit more of that here) but to point out WHY people in Kansas City want to view him as one. I thought that point was fairly clear, but I got a lot of response from, well, yeah, Mike Trout fans. The response was generally along the lines of:

— Come on, Alex Gordon’s a nice little player and all but he’s not HALF the player Mike Trout is.


— Really? You’re seriously comparing Alex Gordon with Mike Trout?


— Kansas City fans are delusional if they think Alex Gordon is an MVP candidate in a league with Mike Trout.

And so on. Now, let me start by saying: I think at this moment Mike Trout IS the MVP of the American League. I’d vote for him. I think he’s the best player in baseball by a pretty fair margin and have written that many times.

That said, the Trout fan responses sound exactly like, yep, the responses I would get from Miguel Cabrera fans whenever I made  the case that Trout deserved to be MVP. I mean, these responses are almost word-for-word like the Cabrera arguments in that for the most part they are not arguments at all. They are simple statements of opinion dressed up with certainty and incredulity to appear like facts. As I’ve written before, it’s like when people put the word “Period” at the end of their thoughts to punctuate just how right they must be.

“The Empire Strikes Back is the best movie in the Star Wars series.”

“The Empire Strikes Back is the best movie in the Star Wars series. Period.”

The second, I guess, is supposed to be more persuasive.

So, “Alex Gordon is no Mike Trout. Period.” seems to be the Trout argument these days, and the only real trouble with that is those stubborn folks at Baseball Reference and Fangraphs are still figuring that pesky WAR statistic. And that pesky WAR statistic suggests that Alex Gordon, in fact, IS playing almost as well as Mike Trout. It suggests that Oakland’s Josh Donaldson IS playing about as well as Mike Trout. And it demands a closer look.

What made Trout so absurdly wonderful his first two full seasons was, as mentioned, the variety of his contributions. That 2012 season, holy cow, the guy did EVERYTHING. He hit, he hit with power, he ran, he threw, he played breathtaking defense, he walked, he stole bases, he scored runs, he drove in runs, he was incredible in ways that that exploded the imagination. WAR reflected those things and his 10.8 refwar was Willie Mays like. In 2013, Trout was better in some areas, not quite as good in others, but again he was a bouillabaisse of wonderful, and again WAR reflected those things.

So what’s happening this year? Trout’s still amazing. Utterly amazing. But let’s just be blunt about it: He’s amazing in fewer ways. It’s impossible not to see if you look. In 2012 he stole 49 bases. This year he has 13. In 2013 his on-base percentage was .432. This year, it’s almost 60 points less. The last two years, he struck out 137 or so times. This year, he’s on pace to strike out 175. He’s not as effective a base runner – he’s going first to third on singles less, he’s scoring from second on singles less, he’s scoring from first on doubles less.

And defense … it’s different. In 2012, all the defensive numbers celebrated him … he saved 23 runs with his defense according to the John Dewan system where reserachers study video of every play. Every defensive stat showed more or less the same excellence. Last year, his defensive numbers were a lot more inconsistent. Dewan’s system actually showed Trout’s defense COSTING his team runs. By Fangraphs WAR his defensive contribution went down some, by Baseball Reference’s method it went down a lot.

And this year, all the defensive numbers I see say the same thing – Trout is, at best, an average outfielder and he’s trending as being at least slightly below average.

So what happened? Are the defensive stats wrong? Is the baserunning decline simply a rounding error? You decide but for me Trout seems to be morphing into a somewhat different player. He’s hitting more home runs. He’s driving in my runs. He is becoming more like, well, yeah, the great Miguel Cabrera.

Now, you look at Gordon and Donaldson. Are either of those guys the slugger that Trout is? No, absolutely not.

Trout:. .291/.376/.561 with 30 homers, 91 runs, 94 RBIs.

Gordon: .282/.356/..457 with 17 homers, 71 runs, 61 RBIs.

Donaldson: .255/.346/.470 with 26 homers, 81 runs, 88 RBIs.

Clear advantages across the board for Trout. So why does Donaldson have a higher Baseball Reference WAR than Trout? Well, WAR judges their baserunning to be about even, their tendency to avoid the double play to be about even, and Donaldson has a huge, huge advantage in defense. Donaldson is a marvelous defense, but you can see Trout fans arguing that even if Donaldson contributes more on defense it can’t possibly be THAT much more. And I would say that’s exactly what all the Cabrera fans said when WAR gave Trout such a huge edge on defense in 2012. Same system. Same methods of determining the completeness of a ballplayer. And right now, Baseball Reference has Donaldson ahead in WAR 7.1 to 6.5

Gordon’s advantages over Trout are more varied – both versions of WAR have him as the better base runner, better at avoiding the double play, markedly more valuable as a defender.  Trout still leads Gordon in refwar (0.9) and he leads by an almost insignificant margin in fanwar (0.3 wins). But it’s close. And Gordon has been playing better than Trout of late.

A little bit more about Gordon’a defense: The Dewan system has Gordon saving 21 runs this year with his defense, Trout costing his team six runs with his defense. This is judged against the average player, by the way. You can doubt this, but from what I can see – purely by seeing – this DOES match the eye test. Gordon makes utterly fantastic catches pretty much every day. He is a superb thrower, so superb that few challenge him any more. When he is playing alongside Jarrod Dyson or Lorenzo Cain, the left-center gap is vanishingly small. Meanwhile, I don’t watch Trout every day, but I watch him a lot and to my eye his defense does seem somewhat bland. He doesn’t seem to run down as many balls as I would expect from a player with his amazing athletic ability.

Now, let me repeat this in case anyone missed it: I still think Mike Trout is the best player in baseball, and I still think he’s a worthy MVP. But blanket statements about him being so much better than Gordon or Donaldson are sounding pretty flat to me. Trout is not having as good a year as he did his first two. He’s not, at the moment, as dynamic a player as he was those first two. The Trout-Cabrera arguments for me were never about the two players – both so sensational – but about this idea of myth and reality, about the question of what the eyes see and what the eyes miss. Now, I’m feeling the same way about the Trout-Gordon-Donaldson arguments. WAR giveth. WAR also taketh away.

68 Responses to Talking about the V in MVP (again)

  1. Richard Aronson says:

    Trout’s arm has gotten better this year. He has 13 steals to only once caught; his dropoff there IMO reflects batting second instead of leadoff more often. Also, batting second means fewer plate appearances. Finally, Trout has just finished a horrendous slump; his numbers now are about his lowest this season. If the slump continues, I may be persuaded, but the last few games seem to indicate a turnaround. We’ll see what’s what at the end of the year.

    • Artie says:

      Over an average full season, players hitting second have about 18 fewer plate appearances than those hitting leadoff. Not sure that’s the deciding factor here.

    • You don’t get points because you “could” steal more bases. You actually have to steal bases. Not sure if he’s had injuries, The mgr has chosen to run less, or whatever, but the reason doesn’t matter. He will probably end up with about 20-25 fewer plate appearances owing to batting one position lower in the lineup. So, that’s not it either. You can argue that Trout still leads the league in Total Bases, Home Runs are up and his OPS+ is right at his career average. But that slump was long, his strikeouts are WAY up and his BA and OBP are down significantly. And steals are now a non factor in his game. Still, he’s a great player, but don’t make excuses for his drop in output. A drop is a drop. Reasons don’t matter when determining WAR or when voting on MVPs. It reminds me of the argument that Cabrera WOULD HAVE more dWAR if they hadn’t moved him to first base….because first base actually loses points on the defensive adjustment, while third base adds points. The reason doesn’t matter. Cabrera is now a first baseman.

  2. fenster ludge says:

    I’m open to the new stats but still on the fence with WAR, simply because I don’t know how it’s figured. OPS, for example, is a much better single-number assessment of a hitter than batting average, and it’s easy to get on board with it because it’s a simple calculation and easy to compare across leagues and eras. I look at the WAR explanation on Baseball Reference and it looks like an especially boring part of a textbook. Also, I’m suspicious of the defensive metrics. Is it really possible Trout went from saving 23 runs as a rookie to being a below average defender, at age 23?

    • The problem with OPS, however, is that it adds together two numbers with different denominators. It’s a nice little shorthand, but a player with a .400 OBP and a .500 SLG will have the same OPS as a player with a .300 OBP and a .600 SLG, and I don’t think you could reasonably argue that player B is as valuable as player A when he gives up 100 points of on-base. There’s no doubt that WAR is imperfect, pretty much every stat is, but you have to respect its attempt to display the complete player in a way that something like OPS is not equipped to do.

    • It’s totally possible that Trout could go from being well-above average to well-below average, in the same way that Wade Boggs could hit .332 one year and .259 the next. Random variance and luck happen, and have huge effects. Trout is almost certainly still a very good fielder, but he’s having a very bad year in the field. If you think the MVP is for the most talented player, that’s still Trout. If you think it’s about how that talent manifests in measurable on-field performance, you need to consider Gordon.

      Fangraphs has a decent WAR primer, and you don’t really have to go past the first page of it if you don’t want to. ( But WAR is just marginally less complicated than OPS, and the latter only makes sense to us easily because we’re so familiar with the components. I’ve tried explaining how OPS (which requires explaining how and why slugging and on-base percentage are calculated, and how it could possibly make sense to add them together) to baseball newbies before, and it doesn’t go particularly well. Surprisingly, though, they’re receptive to WAR because the single number, expressed as a Win, strikes them as very clear and obvious.

      • Grant says:

        Yeah, it’s really disconcerting how many people throw out WAR because it looks boring and complicated to calculate.

        It’s funny how we don’t do this with other stats. Does anyone know how to calculate a QB’s passer rating in football? That formula is a mess (and, to be fair, it’s a flawed stat), but we don’t question it because it looks challenging to compute.

        • Brent D'Elia says:

          I, personally, have always questioned that stat. In fact I think it is probably the very worst stat ever created.

          Is is simple? No. It’s absurdly complex.
          Is it transparent? Omg no.
          Is is easy to explain to someone? heh. Um, no.
          Is it scaled to an easy to understand number? Woba, WRC+ FIP. These all work despite being opaque, complex and tough to explain because they are SCALED to numbers that average baseball fans understand. Watch how easy this is: “THink of Woba as OBA. 400 is great. With WRC+ 100 is average every other point is a percent above or below that. A 125 WRC+ is 25% above average. FIP is the same as era. 3.00 is good, 4 is bad below 3 is awesome”

          As for War I like it. I use it. Im partial to bref because in general it SEEMS that defense is less heavily weighted.

          I’m extremely skeptical of the weight defense is getting. I don’t like that the defense is adjusted to an average player at the specirif position.

          You drop any decent fielder in LF and he looks like a freaking world beater. You shouldn’t get extra credit because your team had to drop you in left. Outfielders should just be compared to outfielders. There is no compelling reason Brett Gardner should only be compared to LF’ers.

          Also the defense stats for 1B and C are next to useless clearly not refelcting reality and not taking one of the most important factors in first base defense into account. There is no bonus for preventing errors by digging throws out of the dirt. Thats basically the main thing that separates a good from bad first base defender.

          And while guys have good and bad years on defense the huge swings from year to year should make one very skeptical. Especially because the samples are small and until field fx and hit fx comes along this entire system is utterly subjective or far to simplistic.

          A line drive crushed into the “catch-able” zone that isn’t caught because no one would catch it loses the player just as much points as if Cecil Fielder was playing cf and couldn’t get to a pop up 20 feet in front of him.

          WAR is still the best currently available tool but there are some clear and obvious flaws in the defensive metrics and as such they should be weighted much less.

          We KNOW for a fact how OBA and SLG affect the chance of scoring runs.

          We DON’T know how much this defensive component does.

          As such we should not have enough confidence to weight them equally.

  3. Stephen says:

    Wonderful article, Joe. Period.

  4. deviator77 says:

    Trout is still the best player in baseball. He’s not the most valuable player right this moment, but isn’t almost unfathomable that THIS is considered a down year for him? That guy is unreal. Also, don’t trust CF defensive metrics, like, ever.

    • Best player in baseball on the team with the best record in baseball, but not Most Valuable? Apparently you have another definition of Most Valuable? Share please.

      • BobDD says:

        It is incongruous that you want to question someone’s definition when in the same sentence you have tried the very same thing. It sounds to me like you are trying to change the meaning of “most valuable player” to “the best player on the best team”. MVP is for the league, not just the team with the best record. So you’ve answered my question (and not the way I wanted it answered) if there are still fans who think that value only occurs on one team ever.

        • Excuse me, I said it wrong. He said best player in baseball. So, best player in baseball on the best team in baseball is MVP by any definition, no?

          • BobDD says:

            Yes, of course you are right. I jumped on your caveat of best team, thinking you were one of those who myopically made that a requirement. I accused you of jumping prematurely to unjustified conclusion when I was the one to have done that to you.

            (blush) You were Trout fishing, and I’m the Rube who tried to swallow the bait. To make up for it, I’ll agree with you for most of the rest of the month.

      • deviator77 says:

        “Best” means that he has the most consistent talent. He has the greatest likelihood of replicating success and consistently producing high dividends at the major league level.

        “Value” is entirely results based. It doesn’t matter if the best player had a sprained knee all year and didn’t hit well. It doesn’t matter if he hit into bad luck all year. All that matters is what he did on the field. This year, Mike Trout is probably not the most valuable player. Josh Donaldson is. It doesn’t matter if Trout is a better player.

  5. BigSteve says:

    Is it possible that Trout peaked at age 21?

    • David says:

      It’s almost CERTAIN that Mike Trout peaked at 21. Would anyone seriously expect a guy to put up a 10+ WAR year and then get BETTER? I mean, it’s fun to think that would happen, but it’s just incredibly unlikely. If he settles in to become a nice 6-WAR player for the next decade, he’s STILL an MVP-contender every year and has a great, “inner-circle” Hall of Fame career. The fact that his first year was so outrageously good shouldn’t make his other years seem like a disappointment. So yeah; I’d say he peaked. But who cares? It’ll still be a heckuva ride.

      • Minor quibble: it’s almost certain that his age 20/21 PERFORMANCE was his peak. Trout could still get better, but it’s unlikely that he was actually a 10 WAR player at 21 or that he’ll actually be a 10 WAR player in the future. He probably needed a lot of luck to pull that off, and that luck is unlikely to carry him to more 10 WAR years, even if his skills continue to improve.

    • John says:

      He wouldn’t be the first.

    • Grant says:

      It’s quite possible, and it happens more than you might think. Michael Jordan’s highest scoring year happened at age 23. Tiger was 24 during his 2000 season. Trout may have been a little earlier, but it’s not completely unheard of to peak that young.

  6. Grant says:

    Good article. The thing with defensive numbers is, while they may be flawed, they’re vastly superior to the Eye Test. The human brain simply does not have the ability and memory to recall every single play made by a fielder – and then to assess those plays against every other major leaguer at his position (many of whom you probably haven’t seen that much anyway). It’s an impossibly difficult task, and our eyes and brain simply aren’t up to it. That’s why advanced defensive statistics, which objectively track every single ball in play, are a much better method.

    Also, defense can ‘slump’ just like offense can. Even if Trout’s true talent level on defense is higher than he’s shown this year, the numbers indicate that he hasn’t quite performed up to that talent level. When this happens on offense, no one questions the numbers, we just accept that a player is performing below his true talent level/peak ability.

    Trout may well still be the MVP of course, but let’s not throw out the defensive statistics just because they don’t jibe with our perception or some old highlights of him robbing home runs.

    • Tom says:

      The fallacy here is that advanced defensive statistics are not at all “objective.”

      Any statistic that is derived by a human being viewing a tape and rating the defender’s play based on what the result “should” have been, or “usually” is, is by definition subjective.

      Rather, the advanced fielding metrics attempt to standardize the level subjectivity, this making the result less subjective, which is clearly an advance. But it is still subjective, nonetheless.

      That is why putting ones believe in these defensive metrics, which cannot agree with each other, and the various WAR(P)s, which also disagree with each other, reveals a faith which, though perhaps not blind, is definitely clouded. Actual facts simply do not justify the hard and fast conclusions which are drawn.

      Only a statistic that measures factual occurrences and has no tolerance for the involvement of any human decision can ever be called an objective metric.

      We all wish there was one objective number to sum up a player’s fielding ability, but their simply isn’t. It is far more genuine to accept that unknowns remain than it is to succumb to ones desire for certainty by pretending that there is an “objective” answer.

      • Grant says:

        Sure, they’re not 100% objective…but they are graded without any pre-conceived bias towards a player, and for that and many other reasons, they’re more accurate than the ‘eye test,’ that’s all I meant.

      • David Eberly says:

        I agree with this completely. I do NOT think that defensive performance, absent injury, changes this much year to year for a young player. I don’t think players really get into a fielding “slump” — at least not like a hitter might. So, I am super duper suspicious of defensive WAR because of things like the fact that Trout has turned from a world beater defensively to below average.

        Think of it this way — I can see a golfer getting into a slump — that is a complicated swing (like a baseball swing) that those guys have, and if timing is just a little off, things can change.

        But fielding is more like, blocking in football. Or a running back, perhaps. It is about angles and speed. “Catching” the ball, once you get there for an OFer, is about the simplest thing a baseball player does. Trout still has ridiculous speed. His defense has likely not changed, but his D-WAR has, but an amazing degree? I am at least skeptical.

  7. J says:

    If this is about WAR, why is Felix not part of the discussion?

  8. Mark Daniel says:

    Am I looking at different double play numbers than everybody else? Raw DPs in 2014 – Trout 6, Gordon 10, Donaldson 12.
    This alone suggests that Trout is better at avoiding the DP.

    And with DP opportunities (i.e. PAs with men on 1st and less than 2 outs):
    Gordon: 10 DPs in 93 PAs
    Donaldson: 12 DPs in 121 PAs
    Trout: 6 DPs in 121 PAs.

    Trout is is twice effective at avoiding DPs than the other two. Where are your numbers coming from, and how is WAR so off base?
    Does WAR not take into account plate appearances where a double play is actually a possibility? If not, then WAR must think leadoff hitters are exceptionally good at avoiding DPs in the 1st inning.

  9. Patrick Dunn says:

    Great article Joe. I doubt it will change anyone mind. Once the big mo gets rolling, it’s a juggernaut, and it seems to be rolling toward Trout this year, period. However, you are undoubtedly the MVW once again. Don’t need any stats for that.

  10. Pat says:

    Baseball is tough to quantify with stats because so many stats are dependent on the players around you. RBIs don’t assume the same number of opportunities, fielding percentage doesn’t allow for the fact that some infielders throw to a first baseman better equip to handle bad throws, etc. Below are the most important stats to view a players ability to affect a games outcome on his own:

    1. SLG%
    2. Extra bases taken (including steals)
    3. Ratio of SLG% to Extra bases taken
    – this is to level the playing field between guys who hit a lot of doubles vs guys who hit singles and steal a base creating the same opportunity for their team.
    4. % of runners moved
    – A player should be rewarded in some part for their ability to move a runner. Currently, all stats punish a player for making an out, but there aren’t many available stats that reward a player for “better” outs than outs with no added outcome.
    5. OBP
    6. OBP w/ runners in scoring position
    – BA w/ runners in scoring position is less affective because some players aren’t allowed the opp. to swing in these circumstances. Also, getting on base with a runner in scoring position can be just as valuable.

    1. % of runners prevented from taking an extra base
    2. % of outs recorded on balls hit to your position
    – this of course assumes that over the course of a season players at certain positions will have equal amounts hit to their “area.” compared to players of the same position.
    3. % of plays an additional out(s) was made
    4.Ratio of errors made to total team errors
    -This will allow visibility into the level of skill in the surrounding defense to allow for more independence of other statistics.

    This sums up all the things a player can do to help his team win with as much independence as possible from his teams ability to affect opportunities or lack thereof

  11. dlf9 says:

    Rather than arguing that Trout is the best, PERIOD, let me add one piece of data in his support: his leveraged performance has been great this year. There may be no predictability about future clutch performance, but past clutch performance puts real wins on the scoreboard. The more advanced clutch stats (WPA, WPA/LI, RE24, REW) all reflect that Trout has been a monster when it has counted the most pretty much running away with the AL lead in all those categories. In high leverage PAs, his OPS is 1111, in medium leverage ones, he is at 965, and in low leverage PAs, he is at a relatively pedestrian 854. That has real world wins and losses impact and should go to any discussion of MVP.

  12. Fin Alyn says:

    Trouts decline I think can be summed up fairly easily; he’s bulked up two years in a row. He’s lost a lot of the athletic speed he had, causing a loss of stolen bases and outfield range.

    • NevadaMark says:

      I was about to write something like that, Fin. I thought the guy had put on some weight. He DOES look like he’s bulked up some.

  13. richie bklyn says:

    The biggest problem i’VE always had w/ war is it over rates stats… the hardest thing to do in any sport is to hit. There have bee a bunch f studies that show hitting a baseball consistently for a 300ish avg, or even even a 250ba w/ high obp and/or slg pct. war treats defense, base running way to much. as if its equal to hitting. but its not. If an athlete in good shape trains hand enough and studies the game eough, he can learn to play defense, run the bases, throw to the right base, steal etc etc. as good as any1.
    but if u cant hit, u cant hit.
    To say player X hit 300BA 400OBP 600SLG
    and player Y HIT 250BA 330 OBP 430SLG can have the same over all WAR is almos impossible to me uness player X is a full time dh, slow footed, station to station runner like big papi and player Y is GG who w/ 30+ sb w/ a 90+ sb%.
    Otherwise War for position just doesnt calculate right for me. . Defense Hiiting Base running, and Baeball iq arent all equal 25pct each.
    To me for position players hiting is atleast 50pct or even up to 70pct of the players value. everything else = the other 30-50pct and even that might be to much. even if your not the most atletic guy, or even fastest guy. you can teach yourrself to be at least an avg to above avg defender and base runner, by taking the right routes to the ball, playing hard, knowing where to position yourself, throwing to the right base, cut off man and what not and not making mental mistakes. everyone can do that if they tried…. but not every1 can hit.. u can take the best athletes from any other pro sport and make them a good fielder, baserunner but even mihael jordan couldnt hit.
    WAR doesn valu hitting enough and over values everything else to much. And guys like gordon who is a great lf are over valued defensively b/c dwar.compares playes to other players at that pos. LF is where teams put their good hitters w/ the worst defense so of course gordon will look like a plus 21 defender, Put gordon in RF or CF, and is war value would drop alot.
    I’m not discounting D, BASERUNNING, BASEBALL IQ… just saying war desnt calculate it right.
    I rather look at a players OWAR nd DWAR separately and then make my own judgement then trust WAR as ”the stat”
    LIKe every other stat WAR has it flaws and witers need to stop writing as f its the be all end all stat. you can compare hitters w/ owar, u can compare pos players d vs othe players at the same position. but when u use war to compare every1 across all positions
    it doesnt work

    • dlf9 says:

      Fundamental error in the above post: WAR does not, despite the BR’s post above, treat defense, hitting, and base-running as equal.

      Fundamental error number two in the above post: the BR ignores the position adjustment. Gordon is 21 runs (about 2 wins) above an average LF, but in the WAR calculation, there is a position adjustment, here -5 (half a win), because he plays an undemanding position.

    • Kris says:

      I agree to an extent, in that I don’t think positional adjustments quite do the trick. But that’s just based on my feel of the situation. However, fWAR (at least) presently breaks down to giving about 60% value to hitting vs. 40% for everything else around the league, which is in your bounds.

  14. richie bklyn says:

    wow lots of type o’s up there.. sorry about but my point is
    WAR works when comparing players at the same position… if you want to compare Griffey Jr. to using war to Mays it works… but comparing a CF to a LF you run into problems. B/c right off the bat the reason why most LF are in LF is because defensivly es not good enough to play CF. The same goes for infielders, Kieth Hernandez is probably the best defensive 1b any of us have ever seen. but if i put ozzie smith at 1st i have no doubt he’d be alot better. from the dwar part of war it doesnt work comparing players from different positions. look at the best defensive 3b’s … as good as schmidt nettles rolan, chavez, zimmerman may be at 3rd, the reason theyplay 3b is because they outgrew or cant handle ss. But put ozzie at 3b and i’d bet he’s like brooks robinson w/ 16gg.

    • dlf9 says:

      I’d encourage you to study Rpos, a subset of WAR that specifically adjusts a player’s defensive contribution. It awards runs to SS, C, and 2B while taking away runs for DH, 1B, and LF to account for the demands of the positions. If you think the number of runs it awards across positions is insufficient, please say why. But you erroneously seem to believe it doesn’t exist at all.

  15. richie bklyn says:

    I Didnt mean to suggest that way equates all factors of the game equally, just that it doesnt factor hitting enough and gives to muchcredit for d, baserunning etc … and when the pos adjustments dont translate correctly.
    Hence i use owar for pos players hitting, and dwar for fielding.
    try as they may, u cant make 1 stat be the go to stat as if its the be all end all.
    i doesnt work, hast worked and will never work b/c even w/ all the improvements in D stats… they still arent trust worthy.
    Thats why fangraph and baseball ref wars are different….both keep adjusting the way they calculate the numbers…but neither can pinpoint it correct….. owar is a accurate dwar doesnt work, it has gotten alot better but it cant work….a ss for the most part lay any pos other then catcher and hold his own…cf can atleast be a +defender in lf and rf…. but u cant say either of the above for corne inf or corner of’s. catchers need a complete different skill set defensivly then other pos players.
    like i said above war is great comparing same pos players to other players playing the same pos but when you compae middle inf to 1b or cf to lf and especially of to any inf or catchers to any other pos war is very flawed

  16. Gesge says:

    One thing that Joe does not touch on here is that Gordon plays left field, and unless he plays left field with a jetpack, it’s hard to believe he can have that much defensive value.

    Joe has written two columns that kind of contradict each other–the previous column explains why we should be skeptical of WAR and why suggesting Alex Gordon is the best player in the AL is silly. Now he writes another column that comes off as feeling bad for criticizing WAR. And he seems to commit the same sin that he talked about in his last column, and the sin that he criticizes Trout partisans for now, namely, believing that your opinion is The Truth.

    Just as a reminder, WAR is so inexact a stat that even those who believe in it as The One True Way to evaluate all baseball players cannot agree on how to calculate it. A couple of years ago WAR thought Darwin Barney was one of the best players in the league–you know, Darwin Barney, that guy that the Cubs cut. Maybe, when WAR tells you that Alex Gordon is the best player in the American League, the correct response is not to wonder why we didn’t all know that Alex Gordon was the best player in the American League. Maybe the correct response is to remember that WAR spits out nonsense sometimes.

  17. Edward Brown says:

    My first post ever. I’ll throw Adam Jones into the MVP debate, and I don’t care that his WAR is lower, significantly, than Trout’s or Donaldson’s.

    The reason Cabrera should have won the MVP two years ago wasn’t because he won the Triple Crown, although that helped. He won it for me because he was willing to switch positions – first to third – even when he knew that he would be less comfortable. It’s real easy to say, “But isn’t that what athletes are supposed to do?” Yes, it is, do whatever is necessary to win, but how many big named superstars have voluntarily switched positions and made themselves open to looking foolish by playing out of position? Not many. Jeter wouldn’t do it. Yount did, but he was an all-star outfielder.

    Making this move, Cabrera showed leadership, something that doesn’t show up in WAR. If the biggest name on a team makes an unselfish move, what message is sent to the rest of the team?

    Gordon would be my MVP, and if not Gordon, then Jones, for the Cabrera effect. I don’t know enough about Gordon this year to mention specific examples, but Adam Jones runs out every ground ball, and I mean every ground ball. He slides into second cleanly but with purpose. How many superstars (and I’m not sure Jones is a superstar) do that? Cano doesn’t. It’s the leadership that Jones, and I’m sure Gordon as well, shows that can’t be counted in WAR. I’m not against WAR, but there’s also the eye test. Somebody is doing something right in Baltimore because the O’s have the third best record in baseball after losing an all-star catcher, having an inconsistent appearing third baseman, and dealing with a pitching staff that has gotten better but isn’t seemingly in the upper echelon of staffs I would argue that is because of Adam Jones, among others. Buck gets some credit, but if games are won on the field, then Jones has to be considered. He’s the best player on the third best team in baseball.

    Just something to consider. If we’re going to make everything in the world scientific, then let’s get rid of the voting and simply look at the numbers. If we’re going to do that, why play the games. Put players’ stats in a computer and simply tell me who should win based on science. Science and facts get us most of the way there in an argument, but there’s always room for the intangibles.

    • desorgher says:

      Does Mike Trout run out every ground ball or slide cleanly into second? Does Alex Gordon? does Josh Donaldson?

      This is the problem with using completely subjective data to arrive at a conclusion. I’m not a huge WAR fans as much beyond a starting point, but when you rely on the “eye test” too much you think that Jeter is a good fielder and Robinson Cano is “lazy.”

      • Edward Brown says:

        I understand what you mean. What I’m trying to say is that “leadership” and other intangibles count for something, sometimes. I don’t think a voter went wrong two years ago voting for Trout or Cabrera. I don’t think a voter would be wrong this year voting for Donaldson, Jones, Gordon, or Trout. There may be others to consider also.

        Idealistically, what might be fun is to pick the MVP finalists by September 1, and then have like a playoff with their results over the rest of the season. This idea isn’t fleshed out, but it does remind us that we still have a month to go in the season. Let’s see what players, and teams, do over that period.

        The award is for most valuable, a very subjective word. It’s not for best statistical. Not as subjective.

    • Does he also play the game the right way? If so, you’ve convinced me.

    • BobDD says:

      MVP because he’s willing to switch positions unlike mortal players? So that’s why Pete Rose kept changing full-time positions 2B to LF to RF to LF to 3B to 1B. Babe Ruth only changed full-time positions twice: P to RF to LF – obviously less valuable than Charlie Hustle. {sarc off}

      • Hah, exactly. Most players have to change positions at some point in their careers. Usually it’s because the manager is trying to find a position they won’t butcher as bad as the one they’re currently playing….i.e. Cabrera.

    • brad says:

      Adam Jones’s numbers this year; (fangraphs)120 wRC+, same as Brett Gardner and Sterling Marte, good for 60th place in the bigs, combined Off of 17.0, same as Posey and Santana, good for 31st, Def of 6.6, which has him 2nd among CFs to Hamilton, so he’s got that going for him. Overall bWAR (I know), 4.7, 18th place in the bigs (among position players). Over at B-Ref; OPS+ 121, placing 54th in MLB, around the likes of Morneau and Ellsbury, oWAR of 3.9, placing 25th, between Reyes and Aybar (?!), dWAR of 0.7, which I’m not sure the ranking among CFs but is a whopping 82nd overall, combined total bWAR of 4.3, 29th overall (position players only).
      Nice little player, but you’re on crack calling him an MVP candidate just because the O’s have the division lead. How’s the weather in Baltimore?

      • brad says:

        Ooops. I wrote bWAR the first time when I meant fWAR, in case that’s not obvious.

        • brad says:

          And to possibly move from the snarky to the constructive…

          My point is not to speak ill of Adam Jones. The best numbers we have suggest he’s been somewhere in the range of the 20-25th most valuable position player in the bigs this year, to the tune of 4 and half wins in 4/5ths of a season. That’s fantastic.
          But you waving away the numbers doesn’t mean the rest of us have to, and from my perspective what you’re arguing is that certain ‘intangibles’ amount to an added 2+ WAR, in saying he should be seen as more valuable than the guys the numbers point to. Even just agreeing, for the sake of argument, that that much value is possible to add by running out routine grounders and inspiring your teammates, how can you possibly prove that Jones has that much of a lead over Trout or Gordon or Donaldson or even, say, Endy Chavez in i(ntangibles)WAR?
          Either you’re arguing Jones is unique in this ability, which you don’t seem to be, or you’re crediting him with added value without acknowledging every other player in the bigs has an iWAR, too, to be added or subtracted.

  18. Ron Warnick says:

    I was inclined to doubt Donaldson’s WAR defensive rating … until I actually looked at his stats.

    His range is way above the league average, which is impressive enough. But he’s already got 40 double plays, and he still has about 30 games left. That’s insane.

    For comparison, Mike Schmidt never had more than 40 double plays in a season. And Brooks Robinson — whom we can all agree is the gold standard for third-base fielding — surpassed 40 double plays just twice in his career.

  19. John Gale says:

    Good article, but the opening passage is entirely disingenuous. RBI, batting average and home runs were not the *only* reason Cabrera won those MVPs over Trout. The fact that the Tigers made the playoffs both years and the Angels did not (yes, I’m aware that the 2012 Angels won a game more than the Tigers but missed the playoffs because they finished third in their division, and I don’t care for the purpose of this discussion) had a lot to do with it too, maybe more so.

    Now, if you want to argue that team success or the lack thereof should be irrelevant, and that the MVP should always go to the “best player” regardless of any other context, fine. I don’t particularly agree with that line of thinking, but that’s a discussion for another time. But to simply pretend (perhaps unintentionally) that it wasn’t a big factor? Come on.

  20. What’s happened is that 300/400/600 with 49/5 SB and huge +dWAR has become something less than that. Joe’s right – the dropoff is significant and impossible to miss. What’s happened is the Saberheads (count me as one) have become too invested in Trout-as-advanced-metric God. Those same metrics – and the traditional ones as well – show Trout as a diminished player compared to 2012 and 2013. Trout’s “struggles” aren’t a failure of advanced metrics, they’re a celebration of them.

  21. drew says:

    Good ole Joe Posnanski still thinks the triple crown doesn’t necessarily get you an MVP these days. Apparently leading the world in the ‘boring’ stats of avg, hr, rbi is not enough. I’m assuming he means the on-base, OPS, and slugging are more exciting? This is ludicrous.

    Posnanski is borderline unreadable. I don’t know how you guys can this stuff.

  22. John Leavy says:

    I’ll risk being accused of hijacking the comments section of this particular topic.

    Once again, many of us find WAR in conflict with our own personal Eyeball Tests. It happens. Not as often as some people think, of course- usually, WAR confirms what our eyes tell us. I mean, if you asked a bunch of innumerate old-school reporters and fans “Who were the best defensive players you ever saw,” they’d almost certainly throw out names like Ozzie Smith, Brooks Robinson, Willie Mays, Bill Mazeroski and Johnny Bench And sure enough, Defensive WAR shows all those guys at or close to the top.

    USUALLY, when WAR conflicts with the Eyeball Test, it’s at the margins. WAR may tell us a guy we thought was great was merely very good (or vice versa) or that a guy we thought was lousy was just a little below average (or vice versa).

    It gets interesting when WAR and the Eyeball Test are miles apart. And there has NEVER been a player whose WAR and whose Eyeball Test numbers are farther apart than… Dave Winfield.

    In his prime, Winfield won numerous Gold Gloves (yes, I KNOW the Gold Gloves aren’t to be taken too seriously). Almost all fans (including me), reporters, managers and players thought Winfield was a superb outfielder. He was fast, he hustled, he had a great arm, he almost never made embarrassing errors on routine plays. I saw him play a LOT, and genuinely thought he was a GREAT right fielder for the Yankees.

    And yet, Winfield has the worst Defensive WAR of any Hall of Famer at any position.

    In this case, either WAR of woefully inaccurate our millions of people (myself included) were looking at a TERRIBLE outfielder and telling ourselves he was a star.

    Can anyone explain such a VAST discrepancy? How could the Eyeball Test and the stats be THIS far apart?

    • dlf9 says:

      I’d posit two things, one about the metric itself and one from my own ‘eye ball test’ having watched Big Dave since he played for the Pads in the 70s.

      The latter first … I always thought Winfield played very deep. He rarely allowed doubles over his head and he not infrequently made HR robbing catches. But it seemed to me that he allowed a fairly large number of balls to drop in front of him for singles. My eyeballs aren’t smart enough to weigh the tradeoff between X more singles and Y fewer 2Bs, but dWAR suggests that it was not beneficial.

      The former … Winfield’s career was completed in the pre-PBP days. That is, the opportunities is only an estimate based on the team’s K’s, right vs. left handed pitching, estimates of ground balls vs. flyballs, etc. It is certainly possible that the estimates are off. That is why pre-PBP defensive value is regressed significantly towards the median compared with the more current versions.

      Oh yeah, of Winfield’s large negative performance in Rfield, over 25% of it was accumulated after his missed age 37 season when he had back surgery. By that point of his career, he was certainly not the fleet footed athlete he was when he was drafted in three major sports coming out of Univ of Minn.

    • Patrick Bohn says:

      Generally, speaking, I think it’s just hard for a fan to know that some fly ball into the gap is run down by Brett Gardner but falls in for a double if the outfielder is Gary Sheffield. When you’re watching a game on TV, for example, the camera follows the ball, not the outfielder. It’s nearly impossible to tell (unless there’s a replay) if a player failed to catch it because he got a bad jump, took a poor angle, or simply didn’t have the foot speed to run a ball down.

      Regarding Winfield specifically, you talk about his speed, but I really wonder about that. I mean, he stole more than 200 bases, yes, but he had just 90 after the 1980 seasons. I mention that because prior to 1980, he was +13 in fielding runs, whereas he was -104 runs after that. Perhaps Winfield was a good fielder, but lost his speed.

      Also, Winfield was a nearly full-time outfielder at 38 and 39, which is rare for most players, and were two of the worst defensive seasons of his career. So the defensive counting stats look a lot worse than they would for a player who retired or moved to another position.

  23. Alejo says:

    Wow. How can you measure players using a stat that has at least three versions? (B-R, Fangraphs and Oakland´s system)

    Do you realise how obscure this argument is?

    Simplify. Numbers people need to create a standard that can be universally used and understood.

    • dlf9 says:

      Do you know how many different ways economists use to calculate Gross Domestic Product? Does the fact that you can measure distance in feet or meters make the concept of distance invalid? If you can use Celsius, Kalvin, or Farenheit does it make the day less hot?

      • Karyn says:

        I think comparing the various versions of WAR to the various measures of temperature isn’t all that valid. Temperature measures just use different scales–0C always equals 32F. The various WAR measures weight different factors differently–you can’t say that a bWAR of 4.5 will always equal an fWAR of 5.3, for example.

        You GDP comparison is much more on point.

  24. Alejo says:

    Yeah, because GDP is super clear. No discussions ever about the truth of GDP.

    You just are making my point all over again.

    Simplify mate, develop a stat that has one version and create a standard. I am not against advanced stats, I am against muddy results.

  25. Alejo,

    They’re different because they’re not actually the same stat. Call them “bWAR” and “fWAR.” They’re both measured on the same scale (wins), but they’re not the same thing.

    Nobody says they won’t pay attention to batting average or on-base percentage until “numbers people” go back and come up with only one batting statistic measured in three-digit percentages.

  26. 2 things bothered me about the Trout/Cabrera debate other than the obvious who’s better. Particularly in Cabrera’s triple crown season, so many fans I knew touted his triple crown as the greatest season since Yaz or Robinson. How quickly people forgot Pujols, A-Rod, Bonds, and Larry Walker’s seasons. Using triple crown stats all of those guys have better seasons than Cabrera, but failed to win the triple crown. Throw in a couple of Ichiro seasons and hopefully people realize MVP means more than just power numbers. The other thing was that guys like Robinson Cano and Josh Donaldson who are clearly better all around players than Cabrera were forgotten. McCutchen and Posey too, but at least they got the award on the National side.

Leave a Reply

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