Why we should be okay when Mike Trout doesn't win the MVP Award

Michael Nelson Trout is the best baseball player on the planet earth.  I truly believe that and I have yet to hear a convincing argument otherwise.  Not that there are many who are making that argument, but if there were, they would be wrong.  Mike Trout is phenomenal, and now that the Angels have added Andrelton Simmons, the best hitter and best defensive player in baseball will be gracing the field at the same time for the next five seasons.  I'd like everyone to take a moment to contemplate that.  Try to keep it clean.

...

...

Done?  Great!  Hopefully you're in a good enough mood now to refrain from excoriating me for saying that Mike Trout might not have been the most valuable player in the American League last season. Let me be clear - I still think he was.  But there's a very reasonable argument that he wasn't, and I'm going to run that argument by all of you fine folks.  And then I'm going to run away while you grab the torches and pitchforks.

First, I want to clarify what I mean when I say "most valuable player."  The actual ballot voted on by the Writer's Association has five criteria for determining the MVP.  These include: actual value to the team (strength of offense and defense), number of games played, character, former winners are eligible, and voters can vote for more than one member of the same team (since voters rank the top 10 in MVP voting).  The character clause is very nebulous, so I stay away from that one, and the last two are there to deter voters from withholding votes for spurious reasons.  It's the first two criteria - strength of offense and defense and number of games played - that I examine.

The reason I offer this clarification is because the writers like to take that word "valuable" and attach all sorts of extra stuff to it.  It's true that the ballot acknowledges there is no clear-cut definition of "valuable."  But I fail to see how someone can look at the listed criteria and purposefully rank a player who provided less value higher than another player.  That's how we get all sorts of twisted logic where the writers use narratives to justify voting for the guy they like.  It leads to arguments like this:

"Well, sure, he played vastly better defense and was better on the bases and put up very good hitting numbers in a difficult hitting ballpark, but did he lead his team to the playoffs?  A true MVP would have led his team to the playoffs."

Ugh.  Here's another one:

"Well, yeah, he's really good, but the team was so good around him that even if you replaced him, they still would have won 98 games.  This other guy singlehandedly willed his team into the postseason with sheer grit and moxie."

Blech.  I won't go into any more.  The bottom line is, I do not subscribe to the theory that a player's value can be determined by the performance of the players around him.**  Mike Trout did not make Huston Street give up a homerun to Jed Lowrie in September.  Josh Donaldson did not make Jose Bautista a beast, nor did he trade for Troy Tulowitzki.  Players have no control over those things.  Now, the context of a player's statistics matter (ballpark factors, game situations, etc), and I think it's fair to use concepts like "performance in a division race" as tie-breakers if you simply cannot distinguish between the value of one player and another.  But let's focus on what the player actually did.

**People like to use money as an analogy here.  They will say things like, "$19 is more valuable to a poor man than $20 is to a rich man." or "What good is a $20 bill to that guy if he doesn't have enough other cash to buy what he wants?  That $20 bill doesn't have any value to him."  No.  If wins were boxes of Cap'n Crunch, a $20 bill will buy just as many boxes for a rich man as a poor man, and it will buy just as many boxes whether or not you have enough UPCs to mail in for that secret Cap'n decoder ring.  The value of the bill does not change based on who has it.  It is still $20.  It will always be $20.  And it will always, always be more valuable than $19 or $10 or any other number of dollars that is less than 20.

I brought up Josh Donaldson in the last paragraph because he will be your 2015 AL MVP.  Ach!  Sorry, spoiler alert. But seriously, it's going to be Donaldson.  There is zero doubt in my mind.  He had a great season statistically and he has most of the narratives working for him: his team went to playoffs, he was the most valuable player on a playoff team, the Blue Jays hadn't been to the playoffs in a long time, he hasn't won it before, etc.  But the relevant question is: should he be the MVP?  And the relevant answer is: maybe he should.  Let's go to the numbers.

As usual when it comes to these sorts of comparisons, I like to start with WAR.  According to Baseball Reference (BBRef), Trout wins 9.4 to 8.8.  According to Fangraphs (FG), Trout wins 9.0 to 8.7.  So, game over, right?  Not yet, Bill Paxton.

WAR is a great tool, but it can be a bit of a blunt instrument.  It's great because it seems so simple, but like a duck calmly gliding on the surface of a pond, there's a lot going on below the surface.  There is a degree of uncertainty in these value calculations that we can't wave away as being meaningless.  The best evidence for that uncertainty is that we have two well-respected statistical websites that consistently have different values for players.  Trout's and Donaldson's 2015 WAR figures look similar on both sites, but, for example, Paul Goldschmidt's differs by 1.4 wins.  That's significant.  (Edit: Hat tip to Stirrups for linking this today - it's a good summary of the large differences in the two main types of WAR for 2015.  Interestingly, Goldschmidt doesn't even make the top five.)

Usually, when you see a difference like that of Goldschmidt, it is related to the fielding numbers each website uses to calculate WAR.  Defensive numbers rely quite a bit on human input and are thus subject to human error in measurement and interpretation.  To illustrate, how many of the balls a center fielder reached during the season were related to positioning versus speed?  How much does the fielding prowess (or lack thereof) of surrounding players affect another fielder's numbers?  Both BBRef and FG thought Trout was a bit above average on defense this year after being fairly well below last year (BBRef also thought Trout was well below in 2013; FG disagreed).  Has Trout really regained his defensive prowess?  Was 2014 a fluke?  Or was 2015 not really as good as the numbers might indicate?

Donaldson, on the other hand, has been above average his whole career.  By both the eye test and the numbers, Donaldson is an excellent third baseman, but his defensive numbers were down a little bit this season.  Do the numbers properly represent a natural decline in defensive skills with age?  Or was he actually just as good  defensively but the rating was off because he was playing his home games in a different venue and under a new manager who positioned him differently?  It's hard to tell.

Aside from defense, both versions of WAR use estimates to assign values to other components including positional value, baserunning contributions, and park factor adjustments.  There is actually quite a bit of uncertainty included in the calculations, although most of it is in the margins.  It means we can look at a guy who put up 10.8 WAR and with 99.9% confidence say he's better than a guy who put up 7.2 WAR (ahem, 2012).  But, due to those pesky uncertainties, it's really hard to make a definitive statement that, between two guys who are within half a win of each other, one had a better season than the other.  Do I THINK Trout had a better season?  Yes, but I'm not completely sure.

Putting WAR aside for a minute, Donaldson does have Trout beat in one advanced statistic: Win Probability Added (and it's cousin, Win Probability Added divided by Leverage Index - WPA/LI).  WPA adds context to a player's numbers.  How much did a player add (or subtract) from his team's chances to win each game?  Donaldson added 5.75 points, Trout added 5.32 (Donaldson and Trout were ranked number one and two in the AL in this stat, Anthony Rizzo led the NL at 7.57).

But wait, you might say, Donaldson probably had a lot more opportunities to add to his team's chances to win since he had more guys on base in front of him and his team was generally in position to win more games.  And you'd be right, he did.  Well, you continue, that's not fair to Trout!  Once again, I have to praise your insightfulness.  You are an awfully smart person.  Don't let anyone tell you otherwise.

Let's adjust for those opportunities, then.  WPA/LI removes this bias by dividing a player's WPA by the opportunity he had to affect the game.  Even taking into account Donaldson's greater opportunity to impact the outcome of his contests,  he still comes out ahead of Trout 6.28 to 6.19 (the next closest AL player was Nelson Cruz at 4.53).  This isn't a lot, and it comes with at least as much uncertainty as the WAR calculation.  However, the numbers indicate that Donaldson's hitting contributed more to the Blue Jays' wins than Trout's did for the Angels'.  You may be a skeptic about whether clutchiness is a repeatable skill, but clutch hits happen and they count for a lot in terms of wins and losses.  Trout had many important hits during the season.  Donaldson had one or two more.

Let me reiterate what I said above - it's not that I don't think Mike Trout deserves his second straight MVP award (and what should be his fourth straight).  It's that I think there's enough uncertainty in the numbers to make a case that Josh Donaldson had just as good a season.  And when you've got a toss up between a couple of excellent players, why not give those ol' pesky narratives a chance to have their say?

When Josh Donaldson's name gets announced on Thursday, feel free to curse and yell at the injustice of the world that has once again conspired against our favorite son.  But it's not that bad, I promise.