06 December 2013

2013 Strikezone Analysis: The Catchers

The baseball analysis community has recently begun studying (in detail, not just "by eye") the effects that catchers can have on the called strikezone for a game. Jose Molina (Tampa Bay) has been noted for having a repeatable ability to frame pitches and expand the strikezone for his pitchers. When we looked at how Matt Wieters receives pitches via the "eye test" back in May, there appeared to be some issues. Most notably, Wieters has a very "noisy" glove during the windup and he seems to give up on pitches rather than attempt to frame them. Well, now we have the numbers for the four catchers who played for the Orioles in 2013: Matt Wieters, Taylor Teagarden, Chris Snyder, and Steve Clevenger.

Click through the jump for the details.

(Edit: This post has been edited to include a glossary of terms.)

First, let's take a look at the overall correct call rates for the Rzone and Tzone. Note that the backup catchers, obviously, have much smaller sample sizes than Wieters.

Total Orioles Opp.
Wieters, Matt 89.16% 89.31% 89.04%
Teagarden, Taylor 89.13% 87.89% 89.41%
Snyder, Chris 89.91% 89.24% 91.61%
Clevenger, Steve 89.24% 88.75% 89.70%

Total Orioles Opp.
Wieters, Matt 86.80% 86.89% 86.73%
Teagarden, Taylor 86.55% 87.11% 86.36%
Snyder, Chris 86.85% 89.49% 85.76%
Clevenger, Steve 87.21% 86.50% 87.88%

These numbers are not overly interesting or instructive. On the one hand, we do want a catcher to present pitches and get correct calls from the umpire most of the time. On the other hand, we want Orioles catchers to be adept at framing pitches and earning extra strikes for their pitchers.

Wieters, Matt 90.49% 9.51% 86.56% 13.44%
Teagarden, Taylor 88.72% 11.28% 85.61% 14.39%
Snyder, Chris 86.85% 13.15% 95.00% 5.00%
Clevenger, Steve 88.11% 11.89% 90.48% 9.52%
Opponents 90.83% 9.17% 85.53% 14.47%

Wieters, Matt 97.96% 2.04% 72.51% 27.49%
Teagarden, Taylor 97.82% 2.18% 71.97% 28.03%
Snyder, Chris 95.34% 4.66% 81.50% 18.50%
Clevenger, Steve 96.63% 3.37% 72.93% 27.07%
Opponents 98.14% 1.86% 72.06% 27.94%

Somewhat surprisingly, all of the Orioles catchers outperformed the opponents' season numbers. That's the good news. The bad news, if you can call it that, is that Wieters was not as adept at his backups at earning extra strikes for his pitchers. When Wieters has an off year with the bat, as he did in 2013 to the tune of an 88 OPS+, 1.7 oWAR, .302 wOBA, and 86 wRC+, he needs to provide extra value above and beyond average behind the plate.

Using the 0.14 expected run value from Jon Shepherd for a ball turned into a strike and -0.14 for a strike turned into a ball, we can calculate the following expected run values (ERV) added for the Rzone.

Wieters, Matt 687 418 37.66 0.28
Teagarden, Taylor 84 39 6.30 0.35
Snyder, Chris 38 6 4.48 0.73
Clevenger, Steve 27 8 2.66 0.68
Opponents 773 543 32.2 0.20

Borderline Pitches
Strikes Balls T-R Strikes T-R Balls
Wieters, Matt 50.89% 49.11% 41.09% 58.91%
Teagarden, Taylor 51.56% 48.44% 47.33% 52.67%
Snyder, Chris 60.19% 39.81% 50.94% 49.06%
Clevenger, Steve 66.15% 33.85% 42.86% 57.14%
Opponents 49.88% 50.12% 40.43% 59.57%

Strikes Balls ERV T-R Strikes T-R Balls T-R ERV
Wieters, Matt 1257 1213 6.16 572 820 -34.72
Teagarden, Taylor 116 109 0.98 71 79 -1.12
Snyder, Chris 65 43 3.08 27 26 0.14
Clevenger, Steve 43 22 2.94 21 28 -0.98
Opponents 1452 1459 -0.98 646 952 -42.84

Small sample size caveats aside, Snyder and Clevenger were both quite good at earning extra strikes for Orioles pitchers and therefore helping the team's bottom line for runs.


It's difficult to draw conclusions for roster decisions based on one year of data. However, this does provide some interesting baselines for future analysis. Examining all of the Orioles' opponents strikezones does give a reasonable expectation for "average" regarding called zones. Comparing Wieters, with his larger sample size, to this average does provide a reasonable first glimpse that contradicts (to a degree) previous concerns about his pitch receiving abilities. If he is going to be a catcher whose primary value is his defense, that's important to consider when his dWAR takes the significant dive that it did from 2012 to 2013. It will be important to isolate different pitchers (across multiple seasons) to see if various pitchers are affecting the called strikezone, as well. But, with all the caveats stated, here's where we stand:

Despite questions raised earlier in the season about Wieters' ability to frame pitches, the numbers indicate that he isn't as bad as hypothesized and even exceeds the average (if opponents' numbers in games they play against the Orioles with the same umpire can be considered average). That's good news. The bad news is, Wieters does not put up extraordinarily high numbers for pitch framing, unlike some other teams did against the Orioles in 2013 (ex. Tampa Bay and New York). Combine this with his arm (5th among 15 qualified catchers in allowing stolen bases, 4th in caught stealing percentage) and his bat (8th of 9 in qualified catchers for OPS) and it does look like it may be worth exploring what Wieters could fetch on the trade market. That isn't to say that the Orioles should dump Wieters as they did with Jim Johnson purely to save money, but if a non-rival team comes knocking with a package that can strengthen the Orioles in multiple positions, it would be worth listening. Wieters is still a very nice player to have in Baltimore. Off the field, there are absolutely no negatives and on the field he has his uses. It's just that his bat has never materialized into what was hoped and his defense is slowly tailing off.

As for the backups, the sample sizes are small, but it appears that the Orioles should do just fine with Clevenger backing up Wieters. It may be worth exploring bringing back Snyder (especially on a minor league contract), as he handled the staff well (2.62 ERA when he was catching, compared to 4.30 for Wieters and 4.63 for Clevenger), earned some value on borderline pitches, and even threw out a few baserunners.

Glossary of Uncommon Terms

Rzone - The Rzone is defined as the rulebook "real" strikezone. In the calculations, this is the from edge to edge of the plate, plus half a ball width on each side (to account for pitches where just a fraction of the ball crosses the plate.

Tzone - The Tzone is the typical strikezone as defined by Brooks Baseball. For left-handed, the Tzone is actually about even with the Rzone on the inside corner and extends almost 6" off the outside corner. For right-handed batters, the Tzone extends about 3" from each edge of the plate.

BcS / ScB - In this abbreviation scheme, 'B' is for balls, 'S' is for strikes, and 'c' is for called. So, "BcS" stands for balls (i.e. pitches outside of the zone) called strikes and "ScB" stands for strikes (i.e. pitches inside of the zone) called balls.

T-R Strikes / Balls - This is a method for examining borderline pitches. These are pitches that are within the Tzone, but outside of the Rzone and then called either a ball or a strike. The traditional borderline pitches definition includes looking for pitches within an inch or two of the edge of the zone. The T-R method allows for examining pitches that are typically called strikes, but aren't rulebook strikes. For example, it is actually easier to get a strike called against a LHB that is 3" off the outside corner than to get a strike called against a RHB that is just an inch or two off the inside corner.

ERV - Expected Run Value is calculated based on the expected runs that an offense would score in a given situation. Because balls are obviously good for the hitter, these result in a better expected run value for the offense. The converse is true for strikes. Jon Shepherd previously calculated the value of ball being called a strike as -0.14 expected runs for the offense and a strike being called a ball as +0.14 expected runs for the offense.


Stacey said...

It would be really helpful in the future if you spelled out/explained the abbreviations in this post (or even just linked them to a place that does explain it). I'm not familiar with things like BcB/ScS/etc. Makes the article difficult to understand. Thanks!

Dyon said...

I'll second Stacey. My suggestion: write a glossary page including definitions of the various advanced statistics abbreviations you use, with a tag at each definition, and link to that page/tag the first time you use that abbreviation in each article. If accessible definitions already exist elsewhere on the web, then the links could send us there.
The hyperlinks would not distract frequent readers, but would allow those of us who are baseball fans but not stat geeks (meant in the most affectionate sense) to easily follow along and at the same educate ourselves. (I was eventually able to figure out "Balls called Balls" and "Strikes called Strikes" based on context, but I'm clueless re: Rzone and Tzone.)
An alternative would be to include definitions of those terms that appear on MouseOver, but that might involve more repetitive coding than your writers want to do in each post.

Jon Shepherd said...

In columns like this it makes sense to define tjese statistics because they are not mainstream. We tend not to overly define things like DRS, UZR, rWAR, or bWAR because these statistics are common online.

I think Lou defined Rzone and Tzone in the past. R zone is the rulebook strikezone and T zone is the typically called zone. The abbreviations are not intuiTive and we should do better.

Lou Proctor said...

I had previously outlined these terms at the beginning of this series, but I will include a glossary in future posts for clarity.

I have updated this post and apologize for the confusion.