This morning, we looked at the three potential MVP candidates from Texas and decided that Josh Hamilton was the best of the bunch, as he’s added about 1.9 wins over a league average player at this point. Does that make him a legitimate candidate for AL MVP?

Let’s look at the other contenders, using the same process described in this morning’s post.

Grady Sizemore: 2.7 WPA/LI, +0.5 position adjustment, +0.5 defense – +3.7 wins
Joe Mauer: 2.2 WPA/LI, +1 position adjustment, ? defense – +3.2 wins
Alex Rodriguez: 2.6 WPA/LI, no position adjustment, +0.1 defense – +2.7 wins

I think these three are the real AL MVP candidates as of July 23rd. Sizemore’s having a tremendous season while no one notices, and considering his offensive production matches Hamilton’s while playing in a less hitter friendly park and playing better defense, it’s tough to justify a real argument for Hamilton over Sizemore. Grady’s been the best CF in the league, and if you’re going to pick an MVP from that position, it has to be him.

The difference between Mauer and Sizemore really comes down to how much credit you want to give Mauer for his defense. We have almost no ability to quantify a catchers impact on a pitching staff, so while most people agree that he’s great back there, how much is that worth. Half a win? A win? Two wins? I have no idea. I’ll let you fill in that part yourself, and your answer to his defensive value probably determines whether you think the MVP at this point is Mauer or Sizemore.

A-Rod’s in the race, but he’s going to need a huge finish to make up the gap. He’s capable of it, but I wouldn’t bet on him right now.

Print This Post

Dave is a co-founder of and contributes to the Wall Street Journal.

13 Responses to “AL MVP Race”

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


    Quick question. Don’t you need to double weight WPA/LI in order to put it in the same context with other Wins Added stats. A team starts off with 50% WP. As I see it you only need to add 0.5 WPA in order to create a win.

    Please let me know what I am missing. Do position and defensive adjustments already adjust for this? Is my logic incorrect on the 0.5 WPA = 1 Win?

    Thanks for the help.

    Paul Calluzzo

    Vote -1 Vote +1

  2. Guillermo says:


    I understand it is still a “work in progress” when we are measuring defensive value, particularly on catchers.

    My point is: If you compare Mauer Vs. all the catcher’s in the AL and Sizemore Vs. all the center fielders in all the AL…. Who has the most value?

    I believe it’s Mauer.

    Of course, I know the argument is flawed, but it does serve a point.

    In a position Vs. position basis… Joe Mauer is creating the most value for the twins than any player in the league.

    Great work on the blog.


    Guillermo (Mexico)

    Vote -1 Vote +1

  3. Matt says:

    Carlos Quentin isn’t even an MVP candidate?

    Vote -1 Vote +1

  4. Steve says:

    Wouldn’t straight WPA plus some defense/position adjust b a better way to determine the MVP as nothing is more valuable than a win.

    The guy that has a game winning hit night after night is more valuable to his team then the player piling on runs in a blow out (see the Tiger’s 10 run 8th the other night). I know WPA will not give you the most skilled player, but it will give you the one that contributed the most that season.

    Vote -1 Vote +1

  5. Dave Cameron says:

    Don’t you need to double weight WPA/LI in order to put it in the same context with other Wins Added stats.

    This is a little counter-intuitive, but WPA/LI isn’t really WPA. Because the leverage index is divided out on a play-by-play basis, WPA/LI is really more like a seasonal linear weights by game state. So while it has WPA in the name of the stat, it’s really more like a run estimator that includes base-out state, recognizing that a single with the bases loaded is worth more than a single with no one on.

    WPA/LI, however, does not care what the score or the inning is, so a bases loaded single with a 15 run lead is the same as a bases loaded single in a tie game.

    To address your specific question about WPA, though, the answer is still no – there’s a difference between absolute wins and relative wins. Perhaps the best example is a bases loaded walk off home run in the bottom of the 9th down by 3 – the guy who hit it will get a WPA of +0.9 or so for the home run, but the team still only wins one game. I know its a bit tough to grasp (it took me a while, and several emails with tango, to get this), but 1.00 WPA is 1.00 win above average.

    I believe it’s Mauer.

    I won’t argue with someone who wants to vote for Mauer. I don’t think we know enough about catcher defense to have a leg to stand on either way.

    Carlos Quentin isn’t even an MVP candidate?

    Not really, no. Sizemore is really running laps around the rest of the outfielders in the A.L.

    Wouldn’t straight WPA plus some defense/position adjust b a better way to determine the MVP as nothing is more valuable than a win.

    Not really. WPA does a great job of telling the story of how a game unfolded (which is why Dave Studeman calls it the Story Stat), but it doesn’t do a great job of distributing credit for which runs matter and which ones don’t. If you win a game 2-1, why is the second run more important than the first one? You wouldn’t have won without either one, but if you hit a home run in the bottom of the 9th to win a game that was tied at 1-1, the second run will get far more WPA than the first run. From a point of view of a fan watching the game, that will match how you feel, but it won’t match the reality of what went into winning the game.

    Vote -1 Vote +1

  6. dan says:

    I’d personally use WPA, I don’t agree with David here. I like that it tells how the game unfolded. Some people (writers) tell us that A-Rod doesn’t deserve to be the MVP because he only hits well in blow-outs. Well, that’s what WPA tells us.

    “If you win a game 2-1, why is the second run more important than the first one?”

    That might be a valid point, although I’m not sure if it’s correct. I think the run later in the game matters more, but it’s too late for me to defend my point on that. My complaint is that WPA/LI treats a home run in a 15-0 game the same as it would in a 1-1 game in the 9th. WPA distinguishes between the two. While WPA/LI is a better measure of true talent, that’s not necessarily what the MVP is about. If the MVP were about true talent levels, then we’d decide it by running a Marcel on every player the last day of the season and see who comes out on top.

    Vote -1 Vote +1

  7. Dave Cameron says:

    I like how WPA tells how the game unfolded too, but we can’t ignore the fact that it does misappropriate credit for how the game was won after the fact. The entire role of WPA is to serve as a real time win estimate, much like the poker odds numbers you see on TV. However, no poker player ever decides that the king that hit on the river to give him a flush is the most valuable card in the hand, because he wouldn’t have been in the position to have the king mean anything without the earlier cards.

    The same thing here. If a run scores in the first inning that enables you to break a 1-1 tie in the 9th, the first inning run is just as valuable as the ninth inning run. WPA misses that, and that’s a big deal.

    Vote -1 Vote +1

  8. Paul Calluzzo says:

    Thanks Dave Makes sense now.

    The rangers have a record of 52-50. That is 2 games above average.
    Hitting WPA: +3.69
    SP WPA: -5.27
    RP WPA: +2.58

    Aggregate WPA: + 1.0 = 2 games above .500 (not average). However because after 102 games 51-51 would be average, 52 wins is really only 1 win above average.

    I suppose the low WPA #’s compared to all the other wins metrics can be explained because it is an above average not above replacement metric.

    Vote -1 Vote +1

  9. Nate says:

    As much fun as it is to use WPA, anyone else think it’s a little ridiculous to consider a guy who’s going to end the year with only 70 RBI from the #3 spot in the order? Just doesn’t seem right…

    Vote -1 Vote +1

  10. Matt says:

    I think it’s even more ridiculous that the four candidates proposed here (including Hamilton) are from a second place team, two third place teams, and a last place team. These stats are fine for how good a player really is but MVP’s come from teams that make the playoffs.

    Vote -1 Vote +1

  11. Sky says:

    Guillermo — Dave is taking position into account via the position adjustment.

    Nate — Mauer’s had Carlos Gomez (.283) hitting first all year and Brendan Harris (.316 OBP), Matt Tolbert (.307 OBP) and Alexi Casilla (.360 OBP) splitting time in the secod spot. Even Josh Hamilton couldn’t drive in 100 in that situation.

    Matt — Dave, as he should, is ignoring baseball writers’ lame interpretation of MVP. There’s no need to follow their lead.

    Vote -1 Vote +1

  12. Blackadder says:

    A minor point: WPA/LI actually does care what the score and inning are. For instance, Down by 2, bottom of the ninth, nobody out, WPA/LI will see a walk as worth almost as much as a home run (the only difference being the chance that the guy on first is subsequently removed running the bases). However, same situation, but down by 1, and a home run will be MUCH more valuable than walk. WPA/LI weights every PA equally, but it is definitely sensitive to the inning and score.

    For people who really care about “real wins”, I don’t think WPA is ideal either. Dave’s argument is a good one, and their is also the issue that, for instance, in a high scoring, back and forth game you can accumulate a ton of WPA and still have your team lose. I don’t know what stat would track “real wins”, but WPA doesn’t

    Vote -1 Vote +1

  13. Ted says:

    FYI…Justin Morneau is the MVP of the MN Twins and perhaps a contender in the AL MVP race.

    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>