Tuesday, October 28, 2008

IBA Ballot: MVP

Presented below is my ballot (and some justification) for one of the categories in the Internet Baseball Awards hosted at Baseball Prospectus. I’m just one person, and the whole point of having a vote like the IBA is to get a wide variety of (intelligent) perspectives, and so I will not feel in the list bit slighted if you don’t give a flip about this.

In the American League, it was a pretty underwhelming years for position players, at least as far as MVP candidacies go. This results in a large number of candidates but no real standouts. Here is a chart-form look at some of the top candidates. RAA and RAR are against an average hitter at the position, so the column “Def” is estimated of runs saved above average at the position, based on Justin’s stats:

NAME                           RAA                RAR                Def  

Rodriguez 40 59 7

Sizemore 34 58 9

Pedroia 34 58 9

Mauer 38 54 N/A

Hamilton 31 53 0

Roberts 31 51 6

Kinsler 34 50 -6

Markakis 26 50 -4

Youkilis 21 44 4

Granderson 22 43 3

Morneau 15 41 -9

One thing to note is that in choosing my order, I will not treat the fielding stats as 100% equally valuable to hitting; they are pretty clearly less reliable. Also, the approach of comparing to the average hitter can be argued to favor second baseman and shortchange center fielders, which is something to keep in mind. Also, ARod’s “clutch” numbers are dreadful--while I don’t place much weight on this, it’s something that can swing my opinion in the case of a virtual tie.

As a result of all that, I would go with Joe Mauer (who does well fielding, +7 according to Chone’s estimates, although they don’t account for the quality of the pitching staff) in a close race over Sizemore and Pedroia. However, the most generous possible final RAR for Sizemore (giving him the 58 runs for hitting, 9 for fielding, and 5 more to correct for undervaluing center fielders) leaves him at +73; for Mauer, +54 RAR +7 fielding + an indeterminate amount for being a catcher…leave both the pitching duo of Cliff Lee and Roy Halladay.

I usually try to avoid giving my MVP support to a pitcher if there is a very small margin. This is not out of any bias against pitchers being the MVP, but because I am more confident in the sabermetric evaluation of hitters. No fielding or bullpen support to worry about, less nagging questions about “hit luck” and peripherals, etc. But here there is a clear demarcation between the two pitchers and everyone else, and I have to respect that. So this is how I see it:

1) SP Cliff Lee, CLE

2) SP Roy Halladay, TOR

3) C Joe Mauer, MIN

4) CF Grady Sizemore, CLE

5) 2B Dustin Pedroia, BOS

6) 3B Alex Rodriguez, NYA

7) SP Jon Lester, BOS

8) 2B Brian Roberts, BAL

9) CF Josh Hamilton, TEX

10) RF Nick Markakis, BAL

If any of the nimrod crowd at BTF (note that I consider this a subset of the BTF commentariat, not the whole) ever see this, I’m sure they will complain about how there are few guys from the left side of the defensive spectrum and attribute this to VORP and its treatment of DHs. The fact of the matter is, the left side of the defensive spectrum players in the AL just aren’t that good. Here are the leaders in Hitting RAR (not accounting for position) in order by position, without their identities:

5, 8, 9, D, 8, D, 3

I listed seven because I gave three spots on the ballot to pitchers (merit-based; I don’t have a three pitcher quota or anything)--thus there are seven position player spots up for grabs. It should stand to reason that if the best a first baseman can do is seventh, without considering defensive value at all, they are not going to fare very well when you do consider it. The third baseman, the two center fielders, and the right fielder all make my ballot. The two DHs are a guy who played for a bad team (Aubrey Huff, BAL) and a guy who played only 126 games, but was extremely productive when in the lineup (Milton Bradley). In fairness to them, they each played a fair amount in the field, but are considered 100% DH because of the single position adjustment used here (Huff played 24 games at first and 33 at third, Bradley 20 in the outfield corners).

So one could certainly argue that one or both are worthy of a ballot spot--but who are you going to replace? Are you going to argue that Huff was more valuable than his two Oriole teammates who played the field all the time, and also hit well? Are you going to argue that Bradley is more valuable than Hamilton, who could have very easily been reversed in roles with Bradley had Texas felt that would make them a better team? You can, but I’d have a hard time buying it.

In the National League, there is a runaway winner. Albert Pujols led players with 300 or more PA in SLG, RC, RG, and all four of the “above baseline” categories I track. He was second in BA, OBA, and secondary average. He did all this while fielding well (albeit at first base) and helping his team stay in contention all year when many (including myself) thought they’d be bad. And while it was arguably his best season (I think I’d hold out for 2003), he didn’t do anything that was way out of line with his track record.

I realize that you, as an intelligent baseball analyst, realized all that and didn’t need a lecture. But as an intelligent baseball analyst, you probably don’t care much about my MVP preferences in any case.

For the rest of the ballot, Hanley Ramirez’ seemingly improved fielding makes him a clear #2--even if you don’t believe he’s +7 out there as Jin’s numbers do, there would have to be around a dozen run error in that estimate to make me place David Wright or Chipper Jones ahead of him (that’s not to say that there couldn’t be a dozen run error, but I’ll bet against it).

Wright and Jones were #1/#2 on my 2007 ballot, but they will be #4/#3 this year. Wright’s RAR edge over Jones is razor-thin despite having nearly 200 more PA, and the zone data has Jones ahead by ten runs in the field. I find that hard to believe, but I was learning towards Jones anyway. Again, you can’t go wrong with either of them.

Behind them, I slip the top two pitchers in, then go with Berkman on the basis of trusting batting stats, although Beltran and Utley, on the strength of +10 performances in the field, could very well be ahead of them. Jose Reyes rounds out my ballot; Justin has him at -6 in the field, and there are number of guys who you could also make a case for (Giles, Holliday, Ludwick, and McCann among them).

1) 1B Albert Pujols, STL

2) SS Hanley Ramirez, FLA

3) 3B Chipper Jones, ATL

4) 3B David Wright, NYN

5) SP Tim Lincecum, SF

6) SP Johan Santana, NYN

7) 1B Lance Berkman, HOU

8) CF Carlos Beltran, NYN

9) 2B Chase Utley, PHI

10) SS Jose Reyes, NYN

Four Mets in the top ten will rub some folks the wrong way, but I’m hardly the first to observe that New York is a team with several stars and a lot of mediocre filler around them. It is a testament to Wright, Santana, Beltran, and Reyes that they came as close as they did.

Finally, I apologize again for the terrible formatting. Blogger has made it damned near impossible to copy and paste from Word while maintaining a readable output. The "Meanderings" post looked awful and this one may be worse.

10 comments:

  1. Youkilis does not fare well according to your metric. He doesn't even crack your top-10. Part of this is due to the fact that you are giving him a first-baseman's positional adjustment. Youkilis is a third-baseman in my opinion, forced to play first-base becasue of Mike Lowell.

    I would not have a problem with a weighted positional adjustment. Perhaps this could be done with Games Played at a position, or better yet, defensive innings. Of course, we would have a problem with the DH if we chose to use defensive innings. Anyway, Youkilis' positional adjustment would be a weighted average of his defensive innings at 1B, 3B, and RF. This type of positional adjustment would also be fair to DH's who occasionly play the field, like Aubrey Huff. Why should Huff get the same positional adjustment as David Ortiz? As you point out, Huff played 33 games at 3B and 24 games at 1B this season.

    The only instance in which I reference "value-added" stats is for MVP and Cy-Young discussions. Morneau, Sizemore, Quentin, Mauer, and Youkilis were all within 1 value-added batting run of each other. At first, Youkilis was my choice for MVP. But you could make arguments for Mauer, Sizemore, and even Cliff Lee and Roy Halladay. I really have no idea how the MVP race will play-out. You were correct in stating that it was a very un-inspiring year for position players in the American League.

    ReplyDelete
  2. I agree completely that a weighted positional adjustment is better than a one-size-fits-all. However, for the purposes of assessing 2008 value, I don't care if Youkilis is capable of playing third base, other than the amount that he actually did.

    Weighting by games played, Youk's PADJ would be (125*119 + 36*101 + 2*112)/(125 + 36 + 2) = 115. That would put his RAR at 46, an increase of 2, which still leaves him tenth among position players. I just don't see him as a good MVP candidate.

    ReplyDelete
  3. Patriot,

    I was just offering my lousy, less-informed opinion on the AL MVP race. Also, this post gave me an opportunity to criticize positional adjustments, which I'm very ambivalent about.

    ReplyDelete
  4. I changed my MVP choice in the AL to Mauer.

    I thought that he should of won the MVP back in 2006 as well.

    ReplyDelete
  5. How did Pujols manage to get only 61 votes more than Howard in the NL MVP vote. Pujols' batting average was over 100 points better than Howard's, his OBP was 125 points better, and his SLG was 110 points better. Pujols is also a Gold Glove calbier 1st baseman, while Howard is one of the worst defensive 1st baseman I have ever seen. Yet, Pujols only manages to capture 61 more votes than Howard. Howard managed to get 12 1st place votes. Pujols only had 18 1st place votes.

    ReplyDelete
  6. Of course Howard had the RBI, and MVP voters of recent years have become fixated on them.

    What has always amused me about the "playoff team" qualifier with respect to Howard is that he won the MVP in 2006 on a non-playoff team. The real MVP that season was Albert Pujols, who was on a playoff team--in fact (although the voters did not know it at the time they cast their ballots), he was on the world champs. STL did have a relatively poor record that year, and won two less games than PHI, but hey, they were a playoff team. But Howard hit 58 homers and drove in 149, and the voters swooned.

    In fairness, at least that season Howard was a legit candidate--I have him at +80 RAR versus +83 for Pujols, although as you said, factoring in fielding and other factors would increase Albert's margin.

    ReplyDelete
  7. In case you're wondering why I am posting so late on a weeknight, I was laid-off last Thursday. Of course, when I had a day-job that didn't stop me from staying up late on weeknights.

    Yes, other factors do increase Albert's margin, in both 2006 and 2008. How about baserunning LWTS, courtesy of Dan Fox:

    Pujols 2006: +2 RAA
    Howard 2006: -3 RAA

    Pujols 2008: -0.2 RAA
    Howard 2008: -3.6 RAA

    Let's take a look at value-added batting runs:

    Pujols 2006: 74 VABR
    Howard 2006: 74 VABR

    Pujols 2008: 73 VABR
    Howard 2008: 39 VABR

    Finally, lets compare their Win Probability Added:

    Pujols 2006: 9.57 WPA
    Howard 2006: 8.10 WPA

    Pujols 2008: 6.39 WPA
    Howard 2008: 2.37 WPA

    So Pujols slightly edges Howard as the MVP in 2006, by virtue of his better baserunning and defense, and WPA.

    However, Pujols deserved to be the unannonymous MVP this season.

    ReplyDelete
  8. Sorry to hear that.

    On a lighter note, here is a hilarious editorial from a Philly writer (hat tip to BTF), hand-wringing about the travesty that Howard did not win the award.

    He actually does have a point about the potential conflict of interest that the writers voting pose, but I think that ultimately is based on a viewpoint that takes these awards too seriously. And he is right about Volquez, as I ranted about. But the rest of it is unreal.

    ReplyDelete
  9. Thanks for referring me to the editorial by the Philly writer. I'm still trying to make sense of this statement by the author:

    "Notice there are no decimal points involved there, only whole numbers that made a difference in real baseball games. That takes care of the logic. Now let's look at the process."

    ReplyDelete
  10. No decimal points? I guess we'll never have to hear about batting average again!

    ReplyDelete

I reserve the right to reject any comment for any reason.