The Jeff Mathis Factor
PITTSBURGH — To begin this post, I would like to remind the audience of a series of fortunate events detailed back in February, a timeline of transactions that appears to be quite consequential as related to the surprise Diamondbacks and their resurgent ace Zack Greinke.
On Dec. 2, new Diamondbacks general manager Mike Hazen announced the hiring of Mike Fitzgerald to lead the club’s analytics department, which was to take more prominent and influential place in the organization. It was Fitzgerald who was a primary table-pounder for free-agent-to-be Russell Martin and his presentation skills in the summer of 2012, when Fitzgerald was the No. 2 analytics official in Pittsburgh.
On Dec. 2, the Diamondbacks non-tendered Welington Castillo, the starting catcher who was most responsible for the Diamondbacks’ placing 26th in majors by framing runs saved last season. Castillo ranked 95th in framing runs saved last season, according to Baseball Prospectus.
And on Dec. 2, the Diamondbacks signed Jeff Mathis, the top-rated receiver available on the open market, to a two-year deal. In a part-time role with Miami, Mathis ranked 13th in framing runs last season. To date this season, he ranks eighth.
Dec. 2 was a busy day for the Diamondbacks, and some suspected it might also be an important day for Greinke, from whom the Diamondbacks needed a better season in 2017 to contend in the NL West and to justify his record contract.
Seven months later, Greinke is in the midst of a significant bounce-back season and by some measures — including strikeout rate, K-BB%, and swinging-strike rate — he’s never been more dominant.
There’s something of a trend developing. Last year, when Greinke was paired with one of the game’s poorest framers (combined with a suspect defense behind him), he endured one of his worst seasons of his career. In 2015, when Geinke was most often paired with an elite framer — Yasmani Grandal, who caught 26 of Greinke’s starts — he nearly won the NL Cy Young. In 2014, Greinke was paired with lesser receivers in A.J. Ellis and Drew Butera. He was still excellent but not as good as his career-best 2015 level.
Paired again with a well above-average receiver in 2017, Greinke is Greinke again.
Consider a series of pitches from last year that all crossed the plate in the PITCHf/x strike zone but were called balls nonetheless…
Here’s a slider that Castillo couldn’t stick:
There was this elevated fastball that took Castillo’s glove further out of the zone, giving the pitch an appearance of being higher than it was:
Here’s another pitch in the strike zone that Castillo failed to present cleanly. It resulted in a walk. Greinke really wanted it, as one can tell from his demonstrative hop on the mound:
Then consider Greinke’s better fortune in 2017, when Mathis has received all 12 of his starts…
This is a ball called as a strike:
And here’s another ball called as a strike:
Greinke is getting out-of-zone pitches called as strikes at a rate that’s 3 percentage points higher than last season. Greinke is getting more borderline pitches and is more often ahead in counts. But to how much of Greinke’s turnaround can we credit simply being paired with an elite framer?
I was curious how much Greinke thought we should credit Mathis, and I was curious what — beyond framing — Mathis might be doing to help the Diamondbacks staff. So I approached one of the game’s most cerebral and talented pitchers in the PNC Park visiting clubhouse last week with my questions.
Greinke praised Mathis for framing well, but he had doubts on the overall effect of his performance.
“I read an article [in the offseason] about how much better I was going to be because of Mathis,” Greinke said. “’If he steals 5% [more] strikes, he’s going to get ahead 5% more often…’ But in reality, that’s not really true. I had first-pitch strikes at 70% last year.”
Close. It was actually 67.8%, but it was a career high. This season, his first-pitch strike rate rests at 64.2%. (Curiously, opponents slashed .349/.344/.556 last season against 0-1 pitches from Greinke. This season, the line is .379/.367/.517.)
But overall, Greinke is ahead in the count on 32.2% of his offerings, the highest of the four years considered here. His out-of-zone swing percentage leads baseball, and he’s already induced 246 swings on pitches out of the strike zone, compared to 485 last season.
Greinke also has a career-low zone percentage (39.0%). Is that tied to having a better framer behind the plate, allowing him more confidence to nibble on the edges of the strike zone, knowing he stands a greater chance of getting the call?
Does a Mathis, did a Grandal, give Greinke more confidence? Greinke is aware of catchers’ framing numbers.
“If I’m commanding really good and I have someone who can really stick the ball, if I get two strikes on a guy, or I am feeling really good and I can really locate a pitch, then I might try occasionally to hit the perfect spot,” Greinke said. “But it’s only when I’m feeling really good, and I probably have a couple pitches to work with. Like, say it’s 0-1 or 1-2 and I try to locate. Instead of saying ‘OK, I am going to try and locate a fastball away and try to freeze him,’ I might say ‘I want to locate this ball at the bottom of the knees and on the outside corner where it might be a ball, but he’s so good [at receiving], he might get that call for me…’ It might work only once every 10 tries, but when a guy is good I will try it sometimes. But it’s still very rare that you change the game plan based upon who is [catching].”
Greinke did theorize about another possible benefit of framing that he didn’t think was quantified, and that is the influence of a catcher’s receiving reputation to compel batters to more often expand their zones, particularly with two strikes.
“When Grandal catches, yes, he’s stealing strikes, but he is also getting people swinging at more pitches,” Greinke said. “It might be tough [to measure] because guys might be ahead in the count a little more often [with a quality framer]. But it’s almost like when I’m pitching and the umpire doesn’t call a strike a strike, ‘Crap. I have to throw it more on the plate.’ If a hitter took a pitch and thinks it’s a ball but it’s called a strike, ‘Crap, now I have to chase that even further.’ It almost gets into the hitter’s or pitcher’s mind.”
Greinke believes that this is a phenomenon of framing but adds, “I don’t know if you can technically prove it.”
I took a look at his idea and did not find any overwhelming evidence in a search of Savant data.
According to detailed strike-zone data at Baseball Savant, of the pitches in the 50-50 hinterland, that gray area on the edge of the strike zone, batters swung at 57.2% of Greinke pitchers there last year, compared to a 56.6% rate in 2015 and a 56.4% rate this season.
We expect batters to be most anxious about framing skills when hitting with two strikes.
Overall this season, 37,769 pitches have been thrown out of the strike zone with two strikes. Of those, 14,657 have been swung at with two strikes (38.8%).
When Mathis is catching, it’s a 39.2% rate (264 swings on 673 pitches). Castillo’s rate, in Baltimore, is 37.5% (255/679).
But last season, when Castillo was catching, batters swung at out-of-zone pitches 41.5% of the time (1,065 swings vs. 2,561 two-strike pitches). In 2015, when Grandal caught Greinke, they combined for a 42.6% rate (962 swings vs. 2,258 pitches).
Are batters generally more aggressive?
Overall, MLB batters have a 45.01% swing rate this season. When Grandal is catching, batters swing at a 47.78% rate. When Mathis is catching, batters swing at a 46.11% rate. Castillo? 44.99%.
This brief study does not include the quality of a collective staff’s stuff or pitch type, or control for umpire bias. But it appears the much more important influence is the ability to keep strikes called as strikes and have balls on the edge of the zone called as strikes. And Greinke is benefiting from that this season.
Unlike when the Pirates signed Russell Martin largely due to his framing in 2012, framing metrics now are more available publicly and more widely understood and accepted. Martin wasn’t sure why the Pirates were so interested him after he hit .220 in New York. Mathis knew why the Diamondbacks were so interested in him in despite being a .196 hitter.
“That’s something I’ve done my whole life, try to make pitches look good,” Mathis said of framing. “I think it gives all pitchers confidence. We’re trying to stick pitches on the corner and make them look the best they can.”
Greinke said he has more confidence to pitch, say, inside to a left-handed hitter when he knows he has Grandal or Mathis behind the plate.
And while many of a catcher’s skills from blocking to throwing to receiving have been quantified, it’s possible that some of their more important attributes, like game preparation, like pitch-calling, like confidence-building are not able to be measured. Mathis has played a role in Greinke’s career-high slider usage, which is another considerable factor. Diamondbacks pitching coach Mike Butcher raves about Mathis’ game-calling, ability to read swings and preparation.
“I try to get on the same wavelength, stuff wise, knowing what they like to do, what they don’t like to do,” Mathis said. “The preparation side is what helps you build trust and confidence.”
Mathis has caught all 12 of Greinke’s starts this season. It appears as though he has gained his trust and done nothing to hurt his confidence level. And there are the quantifiable facts that he’s getting Greinke more calls at the corners, more often getting him ahead in the count. Greinke’s ERA- was a league-average 100 last year. This year, it stands at 74. While there are a number of factors involved in the turnaround, Mathis is a significant one. With the Diamondbacks looking like a contender in a watered-down NL, Mathis appears to be one of the more important offseason signings in the game.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
I know this is about Mathis and Geinke, but… what’s Castillo doing to the O’s pitching staff?
Jeff Mathis is the 5th worst hitter in the history of baseball (min 2000 AB’s).
Somewhere, the ghost of Bill Bergen is smiling at this.
( I love the sabr.org bio writeup on him)
Bergen’s lifetime batting average of .170 stands as the nadir of futility; no other major leaguer with 2,500 at bats finished his career with a batting average lower than .210. He collected only two home runs in 3,028 at bats, which is noteworthy even for the Deadball Era. Bergen’s lifetime on-base percentage is a deplorable .194, and his .201 lifetime slugging percentage is just as bad. One other peculiar statistic stands out: In 3,028 at bats he was never hit by a pitch, which may be indicative of his approach to hitting–and lack of success.
It’s ironic that Mathis was THE poster boy for the mocking of old school baseball thought.
He was often held up as an example of how stupid MLB was. This guy who can’t hit so the perception of his defensive value was inflated.
A decade later and Molina and Mathis become Sabr darlings. In many ways Sabr has come back around to justifying some of the seemingly
ridiculous beliefs of old school baseball types.
Its at least still ok to mock pitcher wins.
I was thinking the same thing, the Jeff Mathis/Mike Napoli/Mike Scioscia thing was a big punching bag around here for awhile. Scioscia was adamant that Jeff Mathis was some sort of defensive whiz, while the metrics we had at the time really hated him. It’s funny to see pro-Mathis articles on here
This is certainly true. The SABR guys’ opinions are often dripping with mockery but with better data, they are open to changing their mind. I doubt that’s also true for old-school guys, though (including Scocia). So I think mocking Scocia/LaRussa is still cool.
Mathis has a 20 wRC+. I’m sure he’s doing a great job framing, but overall is he that valuable given how awful his bat is? I’m not so sure.
The entire point of this article answers your question.
The mindset of most fans and even analyzers of the game has yet to change. We still want our catchers to be Mike Piazza at the plate when really, we should view them more like the pitchers.
And they clearly have a direct impact on the success of a pitcher’s performance, so why not think of them this way? Sure, if they can hit too, great! But that’s not necessarily priority anymore and for Arizona that is clearly a good thing.
No, the article doesn’t answer my question. My question was, did Mathis’ value on strikes saved make up for his bat? That wasn’t answered. So if he saves 8 runs on strikes, but gives up 20 on his bat (hypothetical numbers), then it doesn’t do you any good to use Mathis as a catcher because he’s still costing your team.
I saw that Mathis ranks 8th in framing runs. That’s great, but I wanted to know if Mathis, given how awful his offence is, is at a point where he’s still providing value. Yeah, he’s providing a lot of defensive value, but is he a net positive? I didn’t see that written anywhere.
On the site the author linked to, Mathis has 3.4 framing runs. That’s great but he has -9.8 batting runs. So unless I’m missing something, he’s not providing positive value to the Dbacks. They might be better off with worse framing, but more offence from their catcher.
Or if we throw in the 3 framing runs to his WAR, he’s still a -0.2 WAR catcher. So again, he’s at a point where despite the fact he’s providing great defence, his bat is holding him back.
I think the post did a good job of showing that framing has value, but I don’t think they explained how Jeff Mathis is providing value in spite of his 20 wRC+ bat, or how even adding framing runs to his WAR leads to him being an above replacement level player.
I value framing runs, but I think this is overrating it’s value when you’re suggesting a guy with a 20 wRC+ who is not the absolute best defensive catcher in the game is providing above replacement level value. Show me how he’s doing that, and I’ll agree with your statement.
I don’t think the article answered the question.
Even though Baseball Prospectus ranks Mathis 8th in framing runs saved, it still pegs him as a replacement level player (-0.2 WARP).
So unless he impacts his pitching staff in a way that is massively undervalued by current metrics, it appears he gives back all his defensive value (and then some) at the plate.
The article tries to uncover some of his “hidden” value by diving into opposing swing rates, but ultimately finds minimal evidence that he provides much more value than already indicated by his strong framing numbers (which still fail to boost him over replacement level).
Greinke’s WAR is equal to last year’s in half the number of games/innings. He’s on pace to increase his WAR by over 150%. And that’s just looking at one pitcher. If a lot of that is attributed to Mathis, I’d say he’s worth it.
It’s difficult to attribute a lot of this to Mathis, though. I’ll give him credit for the framing, but beyond that? Swinging strikes would be more a credit to Greinke, and those are way up. You could argue that framing better pitches means that a batter might think a pitch further out is a strike because the previous pitch was, and swing at a bad pitch (something Greinke alludes to), but that should be included as part of the value in framing runs. If it’s not, double his framing runs for all I care, but then he’s still replacement level.
The bottom line is framing certainly has value, but when you’re a 20 wRC+ bat, you’re not helping your team. The Dbacks would be better off with a 90 wRC+ bat who is a slightly lesser framer than a 20 wRC+ bat who is a good, but not elite framer. Greinke is benefitting from the good framing, but so would every pitcher in the game. That’s why the value is already factored in with the framing runs.
“If a lot of that is attributed to Mathis, I’d say he’s worth it.”
This is what the article is trying to assess – how much of Greinke’s success is because of Mathis? But there is no conclusive evidence that he provides significantly more value than is already included in his defensive and framing numbers.
If Mathis really is that important to the success of a pitching staff, then we must need a radical change in the way framing/receiving are valued. Because the current metrics show the sum of Mathis’s offense and defense is near replacement-level production.
personally imo the art of gamecalling is probably more important than pitch framing. and the art of pitcher-whispering is another separate important factor.
all in all i’ve learned to appreciate catchers for their receiving far more than i thought possible. it really is a huge deal.
and mathis is a great poster boy for it.
Scioscia was right all along.
Efforts to quantify game-calling have found a smaller range of variation between the best and worst than exists for framing, suggesting framing has a larger impact, I suppose until the point all catchers are so good at framing that there is no difference between them.
Pitch-calling is sometimes imbued with a magical quality that I don’t think it warrants. People talk about it like in each situation there is the perfect, platonic, pitch that is the key to unlocking the batter’s weakness in that moment. I think that is simply not accurate. What matters far more is pitch mix, and also unpredictability. Basically, it matters much more that Rich Hill throws 50-50 fastball-curveballs instead of 75-25, than exactly when he threw each one. Once you have a good mix, a random number generator would be a decent pitch caller, assuming it was weighted properly (i.e. more out-of-zone pitches ahead in the count, etc.).
It’s kinda impossible to quantify, but more and more i’m starting to believe in the ability to “read” batters and guess what pitch they’re looking for in a given situation by how they approach previous pitches. Some pitchers can do it by themselves, but a catcher that’s good at this can help those who don’t I think.
Here’s an admittedly anecdotal example of roberto osuna changing his approach vs Judge simply due to how he looked on a certain early pitch: https://theathletic.com/65015/2017/06/04/anatomy-of-two-at-bats-roberto-osuna-overrules-aaron-judge-in-blue-jays-victories/
and again, beyond pitch calling, there is also probably a skill involved in making pitchers as comfortable and confident on the mound as possible.
I completely buy into the fact that a good approach with knowing how a batter is reacting “in the moment” is significant. Knowing if a guy is late or early – that he expanded the strike zone. All of that feeds into specific pitch sequencing.
Of course, it is all speculation at this point. A hypothesis that would be really hard to test.
“Efforts to quantify game-calling have found a smaller range of variation between the best and worst than exists for framing, suggesting framing has a larger impact, … Pitch-calling is sometimes imbued with a magical quality that I don’t think it warrants.”
As has been observed above, a few years ago, we analytic types were dismissing pitch framing as magical thinking, too. Maybe we should learn from that experience and hold off on dismissing things we can’t definitively quantify yet, eh?
The first step to knowledge is to realize that you know nothing.
This is one of the BEST articles on Fangraphs. And that is saying something.
I like the idea that command pitchers may benefit more from framing than stuff pitchers do.
This is a really interesting investigation, with good discussion from Greinke. I wonder if an effect would be more visible over a larger sample of pitchers.
The more I read about pitch framing, the more I want mlb to automate the zone.
I should clarify that this article, as anything done by my fellow Cardinal Sawchik, was well written. I just don’t like how teams are making a strategy out of fooling an umpire.
I guess ultimately things will balance out when umpires pay attention to articles like these and note “hey, this guy is famous for framing, I’m just going to have to call everything borderline a ball.” Then it stops… or catchers learn to become more discreet and or discerning as to when they work their magic or whatnot.
So… if I read this correctly, the Jeff Mathis “factor” isn’t about analytics at all. It is the psychology and intangibles behind called strikes. Clearly framing has something to do with that, but getting into the pitchers head that he can expand the zone and the catcher may make a sweet play to get a called strike allows him to pitch with more confidence and aggression.
The flip side is that if a hitter gets spun up early in a count or game by a borderline pitch, it may make him more susceptible to Ozone pitches later.
The Runs Saved/Runs Created numbers still indicate that despite his prowess behind the plate, he is a net-loss in the lineup. However, because guys like Greinke are much better and getting deeper into games, their performance delta should be factored into Mathis’ value, which is more than WAR.
Next up, bunting against the shift….