Skip to content

Valuing Catchers: Or, Why Joe Mauer is Still Underrated

August 26, 2009

Fangraphs is arguably the most statistically advanced website on the internet. They have pretty much anything and everything you could possibly ask for—win probabilities, run expectancy, batted ball data, Pitch F/X,

Fangraphs' Win Values underrates Joe Mauer by quite a bit.

Fangraphs' Win Values underrates Joe Mauer by quite a bit.

linear weights, Ultimate Zone Rating…you name it, they probably have it. One of the most valuable statistics that they implement on their site is what they refer to as “Win Values,” which is the culmination of multiple advanced metrics (park-adjusted linear weights for hitting, UZR for defense, and UZR-based positional adjustments) in order to determine a player’s Wins Above Replacement (WAR). While this system is perhaps the best means with which we can objectively identify a player’s value, there are some players that are being consistently over or underrated—and those players are catchers.

Fangraphs has no metric for evaluating catcher defense and assumes that all catchers are exactly league average, when we know that this is simply not true. Now, there’s nothing wrong with omitting catcher defense, because there is no advanced defensive metric along the lines of UZR, and there are a lot of question marks as to just how efficient catching statistics are. There is no real evidence to support the notion that some catchers call a better game than others (although Dr. Carl Sagan might have something to say about that), because Catcher ERA can fluctuate from year to year and Keith Woolner’s study on a catcher’s “framing abilities” also showed wide variation annually. Also, a catcher’s caught stealing rate is not solely based on his arm strength and accuracy, but also on the delivery time of the pitcher. Wild Pitches and Passed Balls are more or less subjective calls by the scorekeeper and can only be taken as seriously as errors are. That being said, it makes perfect sense for Fangraphs to stay away from using a defense estimator for catchers. But if we incorporate a simple rating system, how much of a difference will we see in a catcher’s WAR?

I thought it’d be fun to take a look at the difference it would make for Joe Mauer, who is widely regarded as being the best catcher in Baseball.

The Methodology:

My approach is essentially the same as Justin’s—the difference being that I include “reputation runs” and the Probabilistic Model of Range and exclude the Fans’ Scouting Report. So nothing you see here is revolutionary or original by any means. All of these statistics can be found on The Hardball Times, and you can read Justin’s overview of and methodology for catcher defense here. “Reputation runs” is how often a baserunner attempts to steal off of the catcher, compared to league average—this is a Sean Smith method and not my own either. This is done with the notion that baserunners will attempt less steals off of good defensive catchers than poor ones. And while catchers rarely handle balls hit into the field of play, I’m including David Pinto’s Probabilistic Model of Range to estimate a catcher’s mobility. Here’s a quick rundown of each statistic included into the system:

Reputation: (LgSBA/G * (Player Innings/9)) – (SBA/G * (Player Innings/9)) * 0.086

CS Runs: (CS% * (SBA/G/9)) –( LgCS% * (Player SBA/G/9)) * 0.62

Glove: ((LgWP+PB/G * Player Innings/9) – (WP+PB/G * Innings)) * 0.28

Errors: (LgError Rate * Player Innings – Errors) * 0.48 for throwing errors, 0.75 for fielding errors.

Very simple and very straightforward. Now, let’s take a look at Joe Mauer’s numbers over the years:

Year

Range

Reputation

CS Runs

Glove

Errors

Runs Saved

2004

N/A

0.5

0.6

1.9

-0.1

2.8

2005

1.8

2.5

3.2

1.9

0.1

9.5

2006

-0.6

3.3

3.0

4.9

2.8

13.5

2007

-1.7

-2.5

6.4

0.5

2.6

5.2

2008

2.6

2.8

1.5

3.5

2.7

13.1

 

Mauer ranks (excluding PMR) 19th in 2004 (in only 32 games), 5th in 2005, 3rd in 2006, 7th in 2007 and 3rd in 2008. So we’re looking at an elite defensive catcher. In fact, Joe’s only 4 runs behind defensive mastermind Yadier Molina of the Saint Louis Cardinals. Once we add these figures in to his WAR calculations, this is what we get:

Year

With

Without

Difference

2004

1.6

1.3

0.3

2005

4.5

3.5

1.0

2006

7.6

6.1

1.5

2007

3.6

3.0

0.6

2008

7.2

5.9

1.3

Average

4.9

3.96

0.9

Total

24.5

19.8

4.7

 

Over a span of 5 seasons, Mauer is being underrated by nearly five wins, or about one win per year. That’s a substantial change in value. Generally speaking, catchers add between +/- 4 runs to their value, or about half a win. The difference between the best defensive catchers and the worst defensive catchers ranges from two to three wins per year, which is pretty substantial—so it’s pretty clear that defense behind the plate is pretty important. So here’s a look at the Gold Glove winners from 2004-2008 with this system:

 

Year

AL

NL

2004

Damian Miller (OAK)

Brian Schneider (MTL)

2005

Ivan Rodriguez (DET)

Yadier Molina (STL)

2006

Ivan Rodriguez (DET)

Yadier Molina (STL)

2007

Victor Martinez (CLE)

Yadier Molina (STL)

2008

Kurt Suzuki (OAK)

Jason Kendall (MIL)

 

Every single player here (with perhaps the exception of Victor Martinez) has a reputation for being either above average or outstanding behind the plate. Yadier Molina has an astounding 46 runs saved over the course of four years, and it’s conceivable that it’s closer to 50 (with his incredible ability to pick off runners). It’s good to see just how underrated Brian Schneider is (he saved 19.4 runs in 2004 and has averaged 8 runs saved between 2004-2008, almost a full win added per year). Just for kicks, here’s the list of the actual Gold Glovers and how they rated that year with the system:

Year

AL

Rank

NL

Rank

2004

Ivan Rodriguez (DET)

18

Mike Matheny (STL)

2

2005

Jason Varitek (BOS)

40

Mike Matheny (SF)

2

2006

Ivan Rodriguez (DET)

1

Brad Ausmus (HOU)

2

2007

Ivan Rodriguez (DET)

31

Russell Martin (LAD)

4

2008

Joe Mauer (MIN)

3

Yadier Molina (STL)

4

The AL talent evaluators seem to be a bit enamored with the high profile names, giving the award to players with good defensive reputations having poor seasons behind the dish. The NL talent evaluators seem to have a pretty good grasp as to which players are the best—although I do find it quite funny that Yadier won the Gold Glove in a year where he was fourth-best in the NL, while he was the best the three previous seasons.

This is not a perfect model by any means. There are many ways in which we could improve it—the first thing that comes to mind is that the system will underrate catchers that are exceptional at picking off baserunners, and this current metric does not take into consideration “gamecalling” abilities. Furthermore, we’re working off of league averages, and because of the variation in pitching staffs, there could be an advantage in using a WOWY (With Or Without You) method to adjust for the staff they’re handling. Taking into consideration the delivery time of the pitcher and the speed of the baserunner is also something that should be accounted for, but that data is not available.

If there’s anything that I want you to get from this, it’s this: catcher defense can be quantified, to a certain extent. While the data we have is limited, the results of the method seem to correlate quite well to players that have a reputation for being a good or bad defensive player. WAR is a great tool for identifying player value, but when it comes to catchers, it just simply cannot be trusted. I hope Fangraphs includes some sort of metric relatively soon; even if it’s something as simple as Justin’s system.

Here’s a spreadsheet containing the figures from 2004-2008 for your enjoyment.

Advertisements
2 Comments leave one →
  1. steve shane permalink
    August 30, 2009 1:36 AM

    an idea worth looking into for the reputation formula is seeing how often the top SB guys attempt steals vs a catcher vs how often the bengie molinas of the world attempt steals.

    and, if its possible, seems like there should be some variable as to who is pitching…ie certain pitchers (most closers in particular) are easier to steal off of than a lefty with a great move/slide step…. maybe a pick off rate and SB% rate variable for pitchers.

    • triplesalley permalink*
      August 30, 2009 5:29 PM

      You’re absolutely right, Steve- in fact, I’ve been working on a model of my own that takes into account handedness of the pitcher, the run value of the base they are attempting to take, and the speed of the baserunner (by using speed scores). I know how I want to do it, but putting together the data all at once is, of course, the difficult part. This new “digitized baseball” system they’re using should also give us a more accurate read on a baserunner’s speed and (hopefully) not only the speed of the delivery of the pitcher, but the pop time of the catcher. That should solve at least a few of the variables that you’ve mentioned.

      Like all defensive metrics, this aggregate system cannot be taken as gospel- there’s far too many variables that are not considered- but I’d like to think (and hope) that this is a step in the right direction.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: