15 July 2016

An Alternative Starting Pitching Solution

Did you know that Matt Wieters' career wRC+ is 98? It surprised me, because when I think of a solid offensive catcher Wieters comes to mind. He reached double digit doubles/home runs for four straight seasons. 2011 and 2012 were especially great, but going by wRC+ they were actually the only seasons that he was an above-average batter.

A wRC+ around 100 is pretty, pretty good for a catcher, but in baseball most backstops fall into one of two categories: bat-first or glove-first. (Excepting Buster Posey, Russell Martin, Salvador Perez and maybe Francisco Cervelli.) If your bat is average or below average as a catcher, it's likely that you're on the roster for your arm and/or your pitch framing abilities. In rare cases, such as Josh Thole, you're on the roster for your ability to catch a certain pitch. Washington's Wilson Ramos is a bat-first catcher. Boston's Christian Vazquez is a glove-first catcher.

Matt Wieters, at this stage of his career, is neither. And that doesn't mean he's useless trash; many teams have worse situations than a ~100 wRC+ catcher with deteriorating defensive skills on their hands. But the other day I was listening to Ben Lindbergh and Sam Miller talk about The Extra 2% Rays on their excellent daily podcast, Effectively Wild, and it got me thinking about the Oriole's catching situation. Lindbergh and Miller were discussing the market inefficiencies and radical strategies the impoverished Rays used to realize that extra 2%, and one strategy they brought up was punting offense for elite defense at the catcher position. Just look at some of the names that Tampa Bay has employed to catch over the last few years: Rene Rivera, Jose Molina, Jose Lobaton, Ryan Hanigan, Hank Conger, Kelly Shoppach. Those guys swing twigs at the plate but each of them rank near the top in advanced catching metrics.

Ever since the hype train arrived in 2009, Wieters has started behind the plate whenever he's been healthy. It hasn't even been a question really. From 2011-2013, he led all catchers in number of pitches received. He had tons of potential, a plus arm, and graded as an OK pitch framer. He's still playable now, but with the ridiculous firepower of the 2016 O's lineup, would Baltimore be better off taking a page from the Rays' book and giving more at-bats to a catcher who might be an offensive black hole but will make up for it with caught base stealers and extra called strikes?

I presented that as a hypothetical, but the O's already have just the guy for this experiment, and his name is Caleb Joseph. After suffering a horrific injury in May, Joseph was recalled on June 30 and slid back into his back-up role behind Wieters. It would no doubt be an offensive downgrade, but Joseph doesn't even need to be average at the plate to be a net improvement over Wieters.

His .409 OPS over 75 at-bats this season has been a huge disappointment after a promising 2015 campaign in which he smacked 16 doubles and 11 homers in just 100 games. Most of his offensive struggles this year are due to a .214 BABIP that is due for a rebound, and a sudden change in his batted ball profile.


Goodbye fly balls, hello ground balls. Caleb Joseph is still a fly ball hitter, and his elevated 11.1 IFFB% indicates that he's trying hard to get under the ball but not squaring it up right. ZIPS projects him to post a .640 OPS going forward, compared to Wieters' projected .734 OPS. If, for the purpose of this experiment, we assume he gets regular playing time, Joseph it's not unreasonable to think Joseph could finish closer to his .693 OPS from last season.

Now, here's the stuff that needs to convince you, Buck Showalter, and Dan Duquette that Joseph can more than close that offensive gap with his defense.

The Pitch Framing

It's a relatively new field of research, but Matthew Carruth provides a sortable table of pitch framing metrics at his excellent site StatCorner. At the basis of all framing metrics lie two stats that are super intuitive: zBall% and oStr%.

  • zBall% is the percentage of pitches caught within the strike zone that are called balls. The lower the zBall% the better - you're getting the strikes you should get. The average zBall% in 2012 was 14.5%.
  • oStr% is the percentage of pitches caught outside the strike zone that are called strikes. The higher the oStr% the better - you're getting the strikes you shouldn't get. The average oStr% in 2012 was 7.2%.

Carruth uses these two stats to create an metric called +Calls, which is the number of extra called strikes a catcher is responsible for cumulatively. There's also PerGame, which is like +Calls but on a per game basis. If it sounds simple, that's because it is. The complex part that leads to some error is the fact that the strike zone is tough to define rigorously for the purposes of statistical analysis as it changes game to game, even hitter to hitter.

There has not been too much research into the stabilization point of catcher framing metrics, but Jeff Sullivan of Fangraphs, who has written a bunch about the topic, estimates it takes around 1,000 - 2,000 receptions to get a good idea of the talent.

Here is how Joseph stacks up against Wieters over their careers:


Over the past three years, Joseph has gotten 168 extra strikes called for his pitchers. Wieters has "lost" 128 strike calls for his pitchers. This graph makes it look like Joseph's skills are deteriorating, but if we look at the per game stats, a clearer picture emerges.


Among catchers with at least 1,000 receptions this season, Joseph's 1.41 PerGame mark ranks sixth in baseball. His 11.4 zBall% is particularly impressive, trailing only Yasmani Grandal, Kevin Plawecki, and Buster Posey. His 9.0 oStr% is also in the top ten. 

Here is Joseph getting a called strike for Kevin Gausman on a borderline breaking ball around the outside edge:


Wieters, meanwhile, has been bad to awful. Most troubling is his oStr%, which ranked dead last among catchers with at least 4,000 receptions in 2015 and has been below average since 2012. 


I can't claim to have done enough research to be able to quantify the impact that +2.51 extra called strikes per game makes, but given the seemingly unsolvable struggles of the starting rotation, the Orioles should be open to trying out any solution. Especially one that doesn't require making a trade in an extreme sellers market

The Arm

Wieters never boasted elite arm strength, but his CS% was among the league leaders from 2011 to 2013 and I think the perception of his arm strength today has been skewed by that stretch. His arm started to decline in 2013, and he hasn't been the same since undergoing Tommy John in 2014.


Then there's Joseph, who led the AL with a 40.4 CS% in 2014, and his 33.6 CS% from 2014-2015 ranks fifth in baseball. Joseph's track record isn't very long, but he's actually the same age as Wieters. He took six years to get the majors and if you fold in his minor league track record, there's no reason to think the arm isn't legit.

The Orioles are not the Rays, but sometimes small-market strategies make sense for high-payroll teams. With Davis, Machado, Trumbo, Schoop, and Jones slugging the lights out in the middle of the lineup, Baltimore is in a great position to sacrifice a marginal amount of offense for a potentially huge gain in catcher defense that could trickle down to partially address the team's greatest weakness.


16 comments:

Boss61 said...

Thank you for an insightful blog that quantifies what my eyes have been seeing - Joseph's superior pitch-framing and arm make him a better overall catcher.

Time to let Wieters walk and spend that money elsewhere. Thanks for the memories, Matt.

tsiz said...

I jumped on this bandwagon when Wieters took the QO. It broke my heart when they even extended the QO. It's been clear to me for a long time that Joseph's defensive skills more then make for the lack of offence. Especially this year on a team with monster bats and poor starting pitching. If Calib is healthy he should be the number 1 guy.

Ken M said...


Very interesting stuff. Brings 2 questions to mind...

Has there been a study done that approximates additional outs/runs per additional strike called?

And how do those stats look per inning caught, because it seems as if Wieters would have more short partial games than Joseph?

Jon Shepherd said...

Roughly, an additional strike saves 0.16 runs. That number gets pused around due to several variables, but that is that starting point. You can google and find dozens of studies on it.

Roger said...

There may be pitchers that would be better served by Caleb. Watching Gallardo last night was tortuous but with a few extra strikes, he would have been passable. Eventually, Tillman and Gausman may be able to operate "catcher independent" but Gallardo and, ugh, Jimenez never will. The junkballers need the extra piece of the plate. They should make Caleb Gallardo's personal catcher and go from there. Wieters will get another QO and will likely move on this year because his hitting is returning. The statistics may be a little unfair as Wieters has been injury prone during his "poor defense" time. If he returns to being healthy longer term, his defense should improve. He is still not catching 2 days in a row, right? Tommy John takes two years to come back from even if the players are playing during the second year. Our catcher of the future, though, is a bat-first catcher (Sisco) so get ready. Pena seems to be a good defensive catcher.

Elie Waitzer said...

I like the idea of using Joseph with Gallardo/Jiminez - anything that will get him more time behind the plate will benefit the O's, or at least is worth a shot.

Also, according to the framing numbers, Pena is below average. In fact, the O's haven't had an above average framer on the level of Joseph since Chris Snyder. Lavarnway, Hundley, Clevenger, Teagarden, Fox, Esposito, and especially Paulino were all below average. Some of them had good arms, but according to the stats Joseph stands out. I meant to include this video in the post - Joseph explaining his framing approach: https://www.youtube.com/watch?v=m55T_ywCE2Y

GRob78 said...

Great post on the two backstops. I was surprised with Wieters took the QO, but don't think the Orioles will offer him more than a passing deal when the end of the season comes up. They seem to be happy allowing Joseph to take the primary role for the next couple of seasons and transition Sisco or Pena into catcher roles. The market for Wieters should be good to watch, I don't think he'll get near the offers he could have several seasons ago. The FA market looks to be a little thin at catcher, so we'll see how he does. I doubt Wieters is with the Orioles in 2017.

Roger said...

Caleb didn't help Bundy much today. I think he wasn't familiar enough with the way Bundy's fastball moves. He set up so it would move right back across the middle of the plate. In this case, I think Wieters might have been better because he's been catching Bundy when he was pitching well. It's not enough just to give a solid target, you have to know how each pitcher's ball moves to make the best out of his outing. Bundy got hit hard a few times because his fastball would start off the plate and move right back to the geographic center of the plate making it easy pickings.

Jon Shepherd said...

Bundy was not hitting his spots. That is not about being unfamiliar with movement. Bundy was typically elevating and falling off.

Roger said...

Sounds vaguely like a chicken and egg argument and I also heard the announcers talking about too long a layoff, which could have been an issue too. It didn't look to me like he had any problem with control not like Wright or Jimenez or Gallardo. At least two of the homers (maybe all three) were on 2-strike pitches and one of the walks was of the four pitch variety right after a homer. And he had 4 K's in 3+ innings not the lack of K's he had earlier. Pitch calling and setting up with two strikes was more of a problem. Even Dempsey said he didn't throw the curve enough. The results were not great but the start was very encouraging. So much potential for later in the season. And Donnie Hart held his own too. I really hope the O's don't bring in O'Flaherty - he really doesn't have it any more. They need to take a longer look at Aquino. Maybe give him a spot start. One of the best options the O's have of improving is to release Jimenez to create a space for younger guys with options. Especially when O'Day comes back.

Jon Shepherd said...

I am at a loss that it is a chicken an egg argument. I guess Joseph could have buried his glove under the dort to get a low strike.

Roger said...

Jon, specifically on Longoria's HR, Caleb set up on the outside corner and Bundy threw it right at the mitt and the ball tailed over the plate right into the middle. If you anticipated the tailing effect, you should set up farther outside so the pitch would look off the plate and tail back over the corner. At times, Bundy's fastball looked almost like a screwball.

Jon Shepherd said...

That is not how you present for a pitcher. You present the end destination...the target. The pitcher adapts. I mean, you would never see a catcher setting up for the 12-6 curveball by holding the glove straight up in the air. You want as little movement out of the glove as possible. If you sweep like crazy you are going to lose called strikes left and right.

This is basic.

Toby Miller said...

In the article it mentions the Extra 2% book. Another great book on defensive sabremetrics and pitch framing is Big Data Baseball and the 2013 Pittsburgh Pirates.

https://www.amazon.com/Big-Data-Baseball-Miracles-20-Year/dp/1250063507

Jon Shepherd said...

We have reviewed both of those books for the site and, little Depot fact, our former writer Stuart Wallace is given a shout out as a new hire of the Pirates at the back of the book.

Anonymous said...

Great post. Is it possible to include Weiters in a trade deal that nets us a pitcher? If it is, we need to jump on it. Assuming you are correct that Weiters essentially amounts to a very average catcher, then I would think he could work as pitcher bait. After all, we don't need to get an ace to improve the rotation. In fact, we don't even need an average level pitcher who would slot in at 3-4 on most rotations. All we need to make a significant improvement would be a serviceable 5 man. At this point, I basically just assume every 5th game will be a loss. A Weiters for 5 pitcher trade would be good and a Weiters+Jimenez+prospect trade for a legit 3-4 pitcher would be outstanding.

The bats have been historic, the bullpen all but impenetrable, but it's still hard to win a division losing 4 out of 5 games your 5 man starts