When Pitch Framing Travels

You might have noticed that a lot of catchers have been on the move this winter, and you wouldn’t be mistaken if you thought that every single one went through San Diego at some point. (The Padres traded Yasmani Grandal to the Dodgers and Rene Rivera to the Rays, briefly had Ryan Hanigan before sending him to Boston, and ended up with the duo of Derek Norris and Tim Federowicz.) You probably also noticed the Blue Jays gave Russell Martin a huge contract, that the Cubs are thrilled to be going from Welington Castillo to Miguel Montero and David Ross, and, well, yes, this is going to be about pitch framing.

It’s pretty easy to understand the impact pitch framing has made on the game over the past few years, and even if teams aren’t necessarily paying for it on the market, most of the catchers teams acquired this winter are considered upgrades over their predecessors when it comes to framing. When the Pirates lost Martin to Toronto and understood they’d never be able to approximate the offense he’d provided, they instead turned to former Yankees Francisco Cervelli and Chris Stewart, both considered plus framers.

It’s fascinating to think what impact plus framers might have on their new pitching staffs. Clayton Kershaw, for example, just had one of the most impressive pitching seasons of all time, and he did it while mainly pitching to A.J. Ellis and Drew Butera — neither considered particularly stellar framers. (Ellis, in fact, ranked among the worst in the game, an issue he was well aware of when I chatted with him about it last spring.)

The extra strikes Grandal can offer over Ellis can only help. Certainly, the Blue Jays were thinking about giving young arms Marcus Stroman, Daniel Norris and Aaron Sanchez as much assistance as they could when they spent to upgrade from Dioner Navarro to Martin. Jake Arrieta and friends will benefit from Montero and Ross, and so on. To pick one randomly-selected example, Matt Shoemaker faced 219 hitters with the well-regarded Hank Conger behind the plate, and 324 with the below-average Chris Iannetta. With Conger, Shoemaker’s K% was 26.0% against a BB% of 3.6%. With Iannetta, those numbers dropped to 20.6% and 4.9%, respectively. It’s not quite as simple as that — that doesn’t adjust for opponent, park, umpire, etc. — but it’s a noticeable difference. This is a real thing.

I thought about trying to identify which pitchers could improve the most from their new backstop partners, but it’s extremely difficult — if not impossible — to say which pitcher on a staff will benefit the most without any actual data of the pair working together. That is, Grandal is expected to be an upgrade over Ellis, but how do we know if he’ll help Kershaw more than Zack Greinke or Hyun-Jin Ryu? We can guess, but it’s difficult to do more than that. A catcher’s framing skills aren’t necessarily going to help every pitcher the same way.

So instead of looking forward, let’s look back. A year from now, I’ll be interested to see how Grandal helps Kershaw. A year ago, we might have been wondering the same thing, just with last year’s crop of catcher changes (plus one or two as-yet-unknown changes that would come around thanks to injury). And as luck would have it, several of the best and worst pitch framing catchers played their first full season for a new team in 2014, giving us a nice opportunity to compare pitchers who had been with the same team in 2013 against what they did with the new catcher in 2014.

Three of the four worst-regarded framers — Kurt Suzuki, Navarro and Jarrod Saltalamacchia — were new to their teams last year. Three of the best were either new to town or getting their first real playing time in Brian McCann, Rivera and Caleb Joseph. For pitchers who threw at least 50 innings for the same team in both seasons, did we see tangible differences? I hope so. I don’t know where this is going to go. I sure hope it’s somewhere interesting.

To do this, I’m going to borrow one of Jeff Sullivan’s favorite quick-and-dirty toys, which is ‘expected strikes,’ a simple way to look at zone rate and out of zone swing rate to figure out how many more (or fewer) strikes a pitcher received.

As with all things that include relievers and single-year data, sample size is an issue, though it’s somewhat unavoidable here because the entire point is we’re talking about a single season worth of data. Also of note, this accounts for all catchers on a team, not just a new one; but for the most part, these teams had one primary catcher. Does that make this too unreliable? Perhaps, but it will still be interesting to see what comes of it. We press on.

Poor framers (Navarro, Saltamacchia, Suzuki)
Blue Jays
Season	Name	Expected Strikes/1000	Difference
2013	Aaron Loup	-29.93
2014	Aaron Loup	-13.59	16.34
2013	Brett Cecil	-10.20
2014	Brett Cecil	-32.64	-22.44
2013	J.A. Happ	-5.17
2014	J.A. Happ	-16.78	-11.61
2013	Mark Buehrle	11.67
2014	Mark Buehrle	-0.17	-11.84
2013	R.A. Dickey	-8.58
2014	R.A. Dickey	-2.34	6.24
2013	Todd Redmond	-11.24
2014	Todd Redmond	-20.90	-9.66

Going from J.P. Arencibia, 2013’s primary catcher, to Navarro was a clear step down, since framing is one of the few things Arencibia does well. Four of the six Jays pitchers to throw 50 innings both seasons took a big step back, particularly Cecil, although a big velocity spike allowed him to succeed nonetheless. So far, this is essentially as expected.

I’m not even going to pretend to know what happens when you cross a knuckleball with pitch framing — total world destruction, I assume. (Josh Thole caught a considerable amount of Dickey’s innings, anyway.) As for Loup, that huge jump was surprising to me, but I imagine a walk rate that more than doubled (and a release point that never stayed the same last year) introduced some additional variables to this.

Marlins
Season	Name	Expected Strikes/1000	Difference
2013	A.J. Ramos	-32.12
2014	A.J. Ramos	-39.41	-7.30
2013	Henderson Alvarez	-11.83
2014	Henderson Alvarez	-11.65	0.18
2013	Jose Fernandez	-26.99
2014	Jose Fernandez	-15.09	11.91
2013	Mike Dunn	-23.84
2014	Mike Dunn	-23.37	0.47
2013	Nathan Eovaldi	-15.99
2014	Nathan Eovaldi	-13.79	2.20
2013	Steve Cishek	-14.72
2014	Steve Cishek	4.47	19.19
2013	Tom Koehler	-26.24
2014	Tom Koehler	-13.02	13.21

The 2013 Marlins featured five catchers, but mostly Rob Brantly (a poor framer) and Jeff Mathis (a mediocre one, despite his reputation). Saltalamacchia was the single worst framer in baseball in 2014, according to StatCorner. And yet there’s considerable improvement here from three of the seven pitchers shown.

What gives? This one gave me pause. This one made me go over the numbers four times, particularly in regards to Koehler. I’m not sure I have a great answer here, but I can offer some hypotheses: First, Saltalamacchia’s performance has been somewhat all over the place, going from pretty decent in Boston to awful in Miami, to the point that Jeff felt the need to specifically investigate it late in the year. I also wonder if something about Koehler just makes him particularly difficult to catch. Here’s a July 2013 “midseason framing update” post that noted Koehler was the pitcher getting hurt the most of anyone in baseball by indifferent framing, and that had nothing to do with Saltalamacchia. Unlike Cishek, the numbers in 2014 for Koehler aren’t so much good as they are less bad.

Twins
Season	Name	Expected Strikes/1000	Difference
2013	Anthony Swarzak	-4.64
2014	Anthony Swarzak	-10.93	-6.29
2013	Brian Duensing	-31.95
2014	Brian Duensing	-31.61	0.34
2013	Casey Fien	-12.82
2014	Casey Fien	-6.75	6.07
2013	Glen Perkins	-24.60
2014	Glen Perkins	-35.71	-11.11
2013	Jared Burton	-32.56
2014	Jared Burton	-35.07	-2.50
2013	Kyle Gibson	-4.89
2014	Kyle Gibson	-3.77	1.12

The main Twins catchers in 2013 were the atrocious Ryan Doumit and Joe Mauer’s last gasp behind the plate. Doumit is probably the worst framer of this century, and Mauer’s final season was about average. Baseball Prospectus had the Twins ranked as the worst framers in 2013 and 2014 — a fact I wish I’d had handy when I was talking about lousy Minnesota defense last month — and so there’s not much improvement here. Though Doumit wasn’t a full-time catcher in 2013, it says a lot when you can move on from him and not get all that better.

What about the teams that imported better catchers?

Good framers (Rivera, Joseph, McCann)
Padres
Season	Name	Expected Strikes/1000	Difference
2013	Andrew Cashner	-2.28
2014	Andrew Cashner	24.73	27.02
2013	Dale Thayer	-27.13
2014	Dale Thayer	-1.60	25.53
2013	Eric Stults	19.43
2014	Eric Stults	30.94	11.51
2013	Robbie Erlin	11.05
2014	Robbie Erlin	17.24	6.19
2013	Tim Stauffer	0.58
2014	Tim Stauffer	3.57	2.99
2013	Tyson Ross	-1.79
2014	Tyson Ross	6.78	8.57

Unsurprisingly, the differences here are huge, because Rivera was beloved by framing metrics and Grandal also scored well. It’s not that 2013 primary starter Nick Hundley was bad — StatCorner had him as only slightly below-average — but Rivera, as has been documented, was elite. Stults, for example, received 87 extra strikes. That’s essentially an entire game’s worth of outs, just from having an excellent framer. It didn’t necessarily help Stults have a good season, because even great framing can’t save a HR/FB that doubled, but it’s valuable.

Orioles
Season	Name	Expected Strikes/1000	Difference
2013	Brian Matusz	-17.09
2014	Brian Matusz	-0.37	16.73
2013	Chris Tillman	-10.70
2014	Chris Tillman	-5.69	5.01
2013	Darren O’Day	8.74
2014	Darren O’Day	-2.34	-11.08
2013	Miguel Gonzalez	-22.01
2014	Miguel Gonzalez	0.77	22.79
2013	T.J. McFarland	-8.57
2014	T.J. McFarland	14.51	23.08
2013	Tommy Hunter	-12.28
2014	Tommy Hunter	2.74	15.03
2013	Wei-Yin Chen	-2.10
2014	Wei-Yin Chen	2.96	5.06

In Baltimore’s case, getting a new catcher wasn’t the plan, but then again, Matt Wieters having Tommy John surgery after playing just a few weeks wasn’t the plan, either. Hundley came over from San Diego to add support, and there’s some Steve Clevenger in there as well, but Joseph took over and provided value behind the plate despite a woeful 72 wRC+. This is the situation that had the least “regular starter,” but since Wieters had ranked poorly in framing for the past few seasons, to go from that to the average Hundley or above-average Joseph, well, you can see the effect.

The O’s showed improvement across the board, save for exactly the pitcher you would expect to be the outlier — sidearmer O’Day and his unusual delivery. As with Dickey, we’ll mentally note that and move on. We’re constantly talking about what the Orioles do to outperform our projections, the things that don’t show up in WAR, and here you are. This is one of them.

Yankees
Season	Name	Expected Strikes/1000	Difference
2013	Adam Warren	-0.59
2014	Adam Warren	2.03	2.63
2013	David Phelps	2.61
2014	David Phelps	4.87	2.26
2013	David Robertson	21.63
2014	David Robertson	29.34	7.71
2013	Hiroki Kuroda	9.92
2014	Hiroki Kuroda	9.31	-0.61
2013	Shawn Kelley	-2.82
2014	Shawn Kelley	16.33	19.15

There’s less improvement for the Yankees even though McCann was considered a very good framer, despite his offensive woes, but that makes sense: The Yankees have valued framing ever since Jorge Posada retired. Martin’s skills are well-known, and even though the 2013 duo of Stewart and Austin Romine was hideous offensively, they were solid framers. So even though McCann is also very good, there’s simply less room for improvement here, and that’s why the numbers creep up only ever so slightly, other than Kelley, who changed his pitching style considerably.

So what does all this mean? For the most part, in these six examples, we see what we expected. Pitchers got a lot more help in San Diego, Baltimore and New York. They didn’t, really, in Minnesota or Toronto. The Marlins’ results were a little unexpected, though that’s a young staff going from one flavor of bad to another.

When we look back on this at the end of the year, it might be hard to see if Kershaw gained at all, because it’s almost inconceivable that he could improve. But for someone like Nathan Eovaldi, moving from one of the worst framers in the game to one of the best could make a serious difference. The trio of young Jays may be thrilled, going from Navarro to Martin. The D-Backs starters might not know what to do with themselves, going from Montero to… Tuffy Gosewisch? Or whoever ends up there.

However this shakes out, we’ve learned enough about framing at this point to know it’s real. We might not be sure about the actual run values, and some catchers work better with particular pitchers, but the effect is real. For a lot of teams, that specifically went in to 2015 thinking.