The Real Gold (and Lead) Gloves

OPA! is more or less finished.  OPA! is my Retrosheet compatible fielding system that is based on out probability added above average (OPAAA… get it?  OPA!)  Over the past few weeks, I’ve been refining the system little by little, and I’ve finally gotten it where I want it.  Now, I can have some fun with it. 
For my first trick, I wrote some code that assigned a run value to all of the events I looked at.  Let’s take a simple grounder somewhere in the neighborhood of the third baseman.  If the third baseman makes the play (and the first baseman catches his throw), an out has been recorded, a runner is not on base, and any runners on at the time of that ground ball have less of a chance of moving up.  If the 3B lets the ball go through, there’s no out recorded, the batter reaches base, and the runners may just go crazy.  There’s a lot riding on that one play.  I looked at all ground balls that the third baseman either got to or were reasonably within his area of town (if the left fielder actually fielded it, the third baseman got part of the blame) and looked at the run expectancy (how many runs actually scored afterward) after a completed play on a grounder vs. a play that went for a hit.  The difference was the run value of a third baseman converting a ground ball into an out.  I did the same for all the other events for which I coded.
I then applied the run values to the players in the 2007 data file to answer the question, “Who were the best fielders in baseball at their respective positions in 2007?  And while I’m here, who were the worst?”  The Gold Glove is given to the player voted to be the best fielder at his position in a given year.  So, here I present the Lead Gloves!  I’m not the first Sabermetrician to present these types of awards, but I am the third-coolest.
I limited the contenders to those who played more than 729 innings (81 games) at the position in question in 2007.  (For pitchers, I simply asked for 50 IP.)  I ranked the players by how many runs they contributed over the course of the year, which does reward those who played more often.  I suppose I could do runs per inning and set a cutoff for innings played, but I’m lazy.  I continued to follow the tradition of having two winners per position, one from each league, and the tradition of not distinguishing by outfield spots.  One drawback: my system does not rate catchers.  Most of what catchers do is not immediately apparent on Retrosheet, particularly blocking balls in the dirt.
Pos   AL OPA! Gold Glove (OPA! runs)  AL OPA! Lead Glove (OPA! runs)
P      Shaun Marcum (2.76)                       Bartolo Colon (-2.71)
1B    Kevin Youklis (12.02)                       Carlos Pena (-9.67)
2B   Mark Ellis (20.85)                             Josh Barfield (-13.11)
SS    Tony Pena (16.88)                            Brendan Harris (-21.82)
3B   Nick Punto (9.13)                              Alex Rodriguez (-2.47)
OF  Curtis Granderson (22.74)               Jermaine Dye (-26.26)
OF  Coco Crisp (22.42)                             Raul Ibanez (-15.04)
OF  Ichiro Suzuki (16.20)                        Torii Hunter (-13.61)
Pos  NL OPA! Gold Glove (OPA! runs)  NL OPA! Lead Glove (OPA! runs)
P     Jeff Francis (3.03)                              Wandy Rodriguez (-2.78)
1B   Todd Helton (27.88)                          Lance Berkman (-14.27)
2B   Kaz Matsui (30.40)                           Dan Uggla (-25.60)
SS   Omar Vizquel (23.18)                        Hanley Ramirez (-20.56)
3B   Pedro Feliz (21.53)                            Ryan Braun (-33.18)
OF  Austin Kearns (30.76)                       Juan Pierre (-28.86)
OF  Jeff Francoeur (28.56)                      Ken Griffey Jr. (-22.21)
OF  Alfonso Soriano (26.35)                    Chris Duncan (-21.57)
Of the actual winners, only Youkilis, Ichiro, and Francoeur won OPA! Gold Gloves.  Torii Hunter won a Gold Glove in real life (his seventh in a row!), but was actually the third worst fielding outfielder in the American League last year.
A few other notable happenings:

  • To give you an idea of how bad Ryan Braun was at third base last year, the second worst fielder out there was Edwin Encarnacion, who checked in about 9 runs below average.  To give you an idea of how good Pedro Feliz was last year, Punto was in second place among all Major Leaguers.
  • The National League infield includes three Rockies (Francis, Helton, Matsui) and a fourth (Tulowitzki) just missed.  Speaking of, Omar Vizquel is nothing short of amazing.  Last year, he was 40 years old, and he’s still the best defensive shortstop in baseball.
  • Someone call the irony police: Nick Punto was tops at something and A-Rod was last!
  • Juan Pierre had the most horrid arm of anyone in baseball last year.  He cost whoever it is that he plays for (yeah, I know, the Dodgers), almost 24 runs by virtue of the fact that teams ran at will against him.  But, he’s fast!
  • The NL Lead Glove infield all made the 2008 NL All-Star team… including Dan Uggla.  Guess what happened.
  • Francoeur got most of his points off his arm.  Austin Kearns caught a lot of line drives.

Now, for a small laugh at the expense of the Gold Glove voters.  Let’s pretend that the actual winners played on a team and see what would happen with them defensively.
Pos  AL Winners (OPA! runs)   NL Winners (OPA! runs)
P      Johan Santana (1.46)         Greg Maddux (0.28)
1B    Kevin Youkilis (12.02)       Derrek Lee (-0.61)
2B   Placido Polanco (-5.62)      Orlando Hudson (1.13)
SS   Orlando Cabrera (-5.17)    Jimmy Rollins (8.35)
3B   Adrian Beltre (-1.65)          David Wright (5.04)
OF  Torii Hunter (-13.61)         Carlos Beltran (8.18)
OF  Grady Sizemore (-0.56)    Andruw Jones (17.64)
OF  Ichiro Suzuki (16.20)         Jeff Francoeur (28.56)
(Aaron Rowand, 11.56 OPA! runs, tied with Frenchy.  I gave the voters the benefit of the doubt that they really wanted Frenchy.)
It’s forgivable if the top guy doesn’t win if he loses to a guy who’s #2 or #3 at his position.  It’s another to think that the majority (5 out of 8) of the AL Gold Glove winners actually rated below average in the OPA! system.  Even worse, the AL team, if it were real would have a grand total of 3.07 runs above an average team on defense.  In the NL, the writers at least picked mostly above average defenders (with Lee coming in just a bit below), but Hudson, Rollins, and Wright, while above average, are hardly “the best.”
The fact that the Gold Glove voters have no idea what they’re talking about can be proven by a quick examination of Derek Jeter’s trophy case, and the argument has been made before.  But, now with OPA!, I can look to see how good or bad the folks voting on these things have been in all the years that Retrosheet has available.  Perhaps I will.


2 Responses to The Real Gold (and Lead) Gloves

  1. jinaz says:

    You’ll have to forgive me, because I’ve been away for well over a month and haven’t read all of your articles in this series. But I was wondering if you are applying a position adjustment on your lumped outfielders in recognition of how much better the average center fielder is vs. the average right- or left fielder? Tom Tango’s work indicates that “the average center fielder” would be roughly a +10 runs/162g if he played in one of the corners. That’s strictly range, of course, because it’s based on UZR–I have no idea how arms would vary (though I’d guess right fielders are better than left fielders).

  2. Pizza Cutter says:

    In this particular article, I’m not. So, Kearns is runs over the average RF and Soriano is runs over the average LF. This is more just a poke at the Gold Glove voting.
    However, I found that when I studied guys who played both CF and one of the corners during the course of a year, there was little correlation between the ranges there (both in my system and last year using RZR). Common sense says that range is going to be important, and the average CF is put there because he has a lot of range. It just doesn’t show up in small sample sizes such as a season.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: