September 20, 2004

Guess who is buried in Grant’s Tomb.

Master of the Obvious

With plenty of interest and attention on the American League Most Valuable Player Award, the National League award seems unimportant.  It’s certainly true that “round these parts”, where we walk around sniffing, “Of course, Bonds is the MVP,” that NL MVP discussions aren’t very creative.

The question might be, just how dumb is the BBWAA?  Or, in a world without Barry Bonds, who would be king?

Last time out, I examined Ichiro’s chances with a player rating system designed to consider context-neutral performance on offense and defense.  While the rating system has some assumptions and faith in zone rating that many of you may not, it captures the essence of both, even if it’s precision isn’t perfect. 

Okay, Barry Bonds is easily the MVP.  That’s a given.  So, by how much, and who’s second?  If Barry hadn’t crossed the Rainbow Bridge, whispering “I’ll be back” to Heimdall, then what player would deserve our adoration?

[Note: this is c&p from last week]

I wrote about quantitating a total player contribution in 1998-99 in an effort to determine if Rey Ordonez was worth his glove.  He wasn’t.  But I did find that there are players that do (Pokey Reese in particular).

See this article.  There’s even a link to MGL’s early work (2001) in the comments from Vinay Kumar.

While my methodology isn’t as granular as MGL’s, I think we’ll end up with very similar results – mostly.

This table is comprised of the top three players at each position (plus Brian Giles):

RS is runs saved above average; XR is Extrapolated Runs above average

“+” indicates the RS are adjusted for a full season’s worth of chances and the player’s innings played and that the XR are adjusted to reflect what an average player would produce after making the same number of outs the player has made.  And the XR are park-adjusted – thanks to Studes for the Park Factors.

If you see two numbers that don’t appear to add up (like Melvin Mora), that’s because I don’t round numbers.  There are decimal places that make it look that way.

Roughly, the error in the metrics is a handful of runs.  Players within a few runs are just about equal.

What does it mean?

One of the things that jumps out at me immediately, because I have a defense fetish, is Eric Enders’ vindication on the extremely high quality of defensive play by the left side of the Los Angeles Dodgers’ infield.  Adrian Beltre has been raking the ball all season, but is also the most valuable defensive player as well.  Okay, he’s tied with his teammate Cesar Izturis. 

This is one of the reasons why I don’t buy the notion that good defensive players “rob” chances from the player beside them.  Particularly the third baseman and the shortstop.  Izturis and Beltre have been the top two individual defensive players this season, and in the chart, we have the most valuable defensive first baseman (Pujols), second baseman (Castillo), center fielder (Patterson), right fielder (Giles), and catcher (Schneider).

The only position leader defensively that is not in the top three players at his position is left fielder Geoff Jenkins.  Brewers fan(s) can tell you that Jenkins has always been an outstanding fielder and only held back by injury.

Notice Adam Dunn at a weak +6 runs above average.  One of the flaws with the use of “average”, as opposed to a replacement level, is the extreme player.  In my six or seven years of calculating this rating, this has only been a problem with a few players – Bonds and A-Rod are two of them. 

Excluding Bonds from the left field production, Dunn has posted a strong +32 runs above average.

Another interesting tidbit I uncovered is Lance Berkman.  Berkman “qualified” at both left field and right field defensively.  Obviously he’s going to fare better compared to right fielders.  What I found interesting was how much worse Berkman’s defensive rating was in left field as compared to right field.  Berkman actually has played pretty good defense the last few seasons, so I was surprised at how low his rating was.  In right field, Berkman was closer to average.

Jim Edmonds is the MVP runner-up with two weeks left in the season.  In addition to hitting the ball well, Edmonds has played good defense this season.  Edmonds, like many centerfielders has a better reputation than he does skill, but he is a good center fielder.  Sure, anything could happen in two weeks.  It wouldn’t take much for Adrian Beltre to overtake him.  But other than that, either Edmonds or Beltre is the second-MVP this season.

On another note, Craig Wilson is a terrible right fielder.  He’s fine at first base defensively, but doesn’t hit enough.  I have Wilson at a significant negative (-10) overall.  He’s just terrible in the outfield.  He’d come in around average at first base (+/-0).

As for pitchers, Randy Johnson has a VORP of about 66 runs.  I don’t know what average is above replacement, but I’m guessing around 16 runs (for a given IP).  Johnson, while valuable, would be around 10th in the NL. 

As MLB noted in last week’s discussion, Johan Santana should be considered, and given his VORP less the difference between replacement and average, he should be the AL MVP.

Bonds deserves the award, but Edmonds has made a respectable showing.  The NL scale and the AL scale are the same, yet there are a dozen NL players excelling like any AL player.

Hopefully, we’ll all be congratulating Bonds on another MVP Award.