World Series time! Enjoy Premium-level access to most features through the end of the Series!
September 28, 2012
A Sabermetric Case for Miguel Cabrera's MVP Candidacy
In high school, I had a very wise history teacher whose motto was that the best way to win a debate was to be able to make the argument for the other side better than your debate opponent. So, I decided to challenge myself. Can I make a case that Miguel Cabrera deserves the American League MVP over Mike Trout using sabermetrics?
(Note: Keith Law threw down this challenge on Baseball Today on Wednesday, although—I promise—I had already sketched some things out in my head before then. The careful reader will note that I’m using some rather slap-dash math in my estimations below. I know. I wrote this in a Kerouac-ian Benzedrine addled state, except without the Benzedrine, to make a point rather than to hit the third decimal place.)
For those who have recently emerged from a month-long stay in an isolation chamber (and even then, most isolation chambers have wi-fi now), Mike Trout has a commanding 2.5-win lead over Miguel Cabrera in Wins Above Replacement Player. However, Miguel Cabrera has emerged lately as a legitimate contender for baseball's vaunted Triple Crown (ooooooooh….), and many are saying that if he wins, he should receive an extra crown in the form of his first MVP Award.
It's not that Cabrera is having a bad year. Please note: under any legitimate definition of value, he's having a fantastic year, maybe even an MVP year in some other seasons. It's just that according to WARP, Mike Trout is having a year that is just that awesome. If only we were still allowed to make Chuck Norris references.
WARP, for the uninitiated, is a single number that tries to sum up a player's contributions to a team offensively, defensively, and on the basepaths. It compares him to the value that we might expect from an average Triple-A call-up or bench player who played the same position. (Come to think of it, Mike Trout was technically a Triple-A call-up this year.) Mike Trout has been worth about 2.5-3 more wins (roughly 25-30 runs) to his team than has Miguel Cabrera, depending on what win-value statistic you use. In fact, according to Baseball Reference's WAR values, Cabrera doesn't even win Most Valuable Tiger (Justin Verlander has him beat).
But let's focus on Trout vs. Cabrera for a moment and see if there is some hidden value in Cabrera or overstated value in Trout. And no, it can't be in the form of intangibles, clutchitude, or "what he meant to the team." I'll use only things that I can back up with numerical gymnastics.
Running Scoreboard: Trout +30 runs.
Warning! Gory Mathematical Details Ahead!
One thing that we try to do in sabermetrics is separate the effects of performance that a player controls from the effects of luck. Believe it or not, Miguel Cabrera has had a bit of bad luck this year. When Cabrera hits a line drive (and he has hit 99 as of the end of last night’s games), he is hitting “only” .636. League-wide, about 72 percent of line drives go for hits. I used a model similar to what I used here to determine what percentage of credit/blame the batter, pitcher, and dumb luck have in determining whether a line drive will be turned into an out. The answer is that it is mostly (81 percent) a matter of luck. However, the fact that the ball is a line drive (rather than a grounder) is much more in the batter’s control. So, Cabrera gets credit for the line drives, but we should strip out the bad luck that sometimes there’s a guy standing right where he’s hit them.
Cabrera is hitting about nine percentage points below league average and has hit just shy of 100 liners, so let’s turn nine of his lineouts into hits. He has 43 singles and 20 doubles on liners, so let’s give him six extra singles and three extra doubles that he “should” have had. Converting an out into a single is a swing of about three quarters of a run, and an out to a double is roughly one run. Cabrera should have an extra 7.5 runs in his account. Of course, one could argue that what did happen is the only admissible evidence when it comes to determining the value of past performance. But we’re trying to make Miggy look good here, so we’re arguing the opposite.
Running Scoreboard: Trout +22.5 runs
Trout, on the other hand, is hitting .757 on his 75 line drives. We should probably take four percent of his line-drive hits (two of them) and convert them into outs. Trout has 39 singles on liners and 16 extra-base hits. We’ll just steal two singles from him and debit his account 1.5 runs.
Running Scoreboard: Trout, +21 runs
But should the base runner get all that credit? I’ve found that advancement on air balls (read: sac flies) has much more to do with how deep the ball is hit than how fast the runner is. If we sabermetricians are going to carp about how RBI are a lousy stat because they tell more about who hit in front of the player, we can’t credit Trout for the fact that Albert Pujols hit some deep fly balls to score him. Also, I’ve found that the batter deserves some of the credit (about a quarter of the offensive credit) when a runner goes from first to third on a single.
If you look at the numbers, Cabrera is relieved of only a fraction of a run of blame (most of his negative value on the basepaths has been in not advancing on grounders). Trout, if we take away his air advancements and dock him a quarter of his hit advancement credits, loses two runs.
Running Scoreboard: Trout, +19 runs
What that means is that when I look and see that Miguel Cabrera has been somewhat below average as a third baseman this year (when we all expected that he would be Ryan Braun, circa 2007-esque), I have more confidence in that number than I do when I see that Mike Trout has played some Gold Glove defense. (Then I actually watched a game and realized that this is just silly.)
Numbers that aren’t reliable have to be heavily regressed down to the league average, and because outfield numbers are so unstable and we have only one year’s worth of data for Mike Trout, we have to take a very statistically agnostic view of his fielding prowess. And it’s going to cost him.
FRAA credits Trout with only seven runs above average on defense this season, but subtracting seven runs from his lead won’t make a big enough dent. So let’s sample some other defensive systems that rate Trout more highly. Baseball Info Solutions says that Trout has saved the Angels 25 runs. Although other systems, including UZR (13 runs) and TotalZone (15 runs), are more bearish, if we sap Trout of his defensive contributions and bring him down to the level of the average outfielder, the balance starts to tip. If we believe BIS’s 25 runs saved, and believe that we have to regress things 80 percent of the way, Trout loses 20 runs, and suddenly…
Running Scoreboard: Cabrera, +1 run
Victory! I was able to construct a case for Miguel Cabrera over Mike Trout. Or, owing to the shoddy nature of my calculations, I was at least able to bring the numbers close enough that you could start to say “close enough”. Over in the Senior Circuit, Andrew McCutchen, Buster Posey, Yadier Molina, and Ryan Braun are all 6-7-win players. I wouldn’t begrudge any one of these players a first-place vote in the NL.
The problem is that in order to make the case that Mike Trout wasn’t quite as good as Cabrera, I had to completely bury my head in the sand on Mike Trout’s defense. (I also had to cherry pick where I made my adjustments to their respective values.)
And that seems to be the crux of the debate. The only universe in which Miguel Cabrera wins the AL MVP is one in which baserunning and defense have no value. In order to swallow my own argument, I have to believe that Mike Trout either is a merely pedestrian center fielder, or I have to assume that there is precious little difference between outfielders in their ability to play defense. Oddly enough, this sabermetrician now encourages people to watch an Angels game or two.
There’s a deficit of language in how to express that a player is good at playing defense. I can rattle off stats both flawed (*cough* batting average *cough*) and advanced that describe offense. Personally, I think most of it goes back to playground days where defense was the boring part, but that’s another article. As a sabermetrician, what I worry about is that when I come to a deficit of language, I will act like, well, an American, and just yell louder. But what I have to realize is that my own language is not very coherent, and even if it were, I’d still need to teach people how to speak it.
Miguel Cabrera might win the AL MVP when the ballots are counted, even though Mike Trout was the better player. And the sun will rise the next day.
Now, if you’ll excuse me, I need to get back to my isolation chamber.