A Theory on Russell Martin’s Framing Numbers
Projection systems tend to look at reality a whole lot more soberly than us humans, who can fall madly in love with a player on the basis of aesthetic appeal alone. That’s why most offseason columns here at FanGraphs reviewing free-agent acquisitions tend to damper down instead of ramp up excitement.
So it is a meaningful testament to Russell Martin’s skills that, upon being signed by the Toronto Blue Jays to a five-year, $82M contract as a 31-year-old catcher — i.e. after Martin has already sustained several lifetimes of knee-shredding, cup-checking abuse in baseball’s most brutal position — the deal was graded positively in these pages by Mr. Sullivan.
As Jeff notes in the piece, not only is Martin’s bat an improvement over Toronto’s primary catcher last season, Dioner Navarro, but Martin has also provided the equivalent of the value of an above-average player via his pitch-framing skills alone. (And these skills are not presently included while calculating WAR here at FanGraphs.) This, while Navarro tends to provide negative value to his team, as one of the league’s least-proficient framers.
Even more impressively for Martin: he has annually sustained this framing excellence even while bouncing from team to team, with none of the long-term security of franchise cornerstones Buster Posey or Yadier Molina. After two-year stints with both the New York Yankees (2011-12) and Pittsburgh Pirates (2013-14), the Blue Jays are Martin’s fourth team since just 2010, his last year with the Los Angeles Dodgers, and he has been an excellent framer at every stop.
Using framing data from StatCorner, we see that Martin has had a consistently low percentage of pitches in the zone called for balls (zBall%), a consistently high percentage of pitches from outside the zone called for strikes (oStr%), and a meaningful amount of runs saved above average (RAA) each season. His league rank in RAA is annually superb, especially considering that Martin is not an offensive liability like plenty of the catchers above him:
| Year | Team | zBall% | oStr% | RAA | Rank |
|---|---|---|---|---|---|
| 2010 | Dodgers | 13.4 | 8.7 | 9.7 | 11 |
| 2011 | Yankees | 12.3 | 8.9 | 28.7 | 3 |
| 2012 | Yankees | 12.3 | 9.1 | 24.1 | 2 |
| 2013 | Pirates | 12.1 | 8.3 | 17 | 6 |
| 2014 | Pirates | 11.8 | 8.7 | 11.7 | 10 |
So how bizarre, then, that Martin ranks 56th among catchers in RAA thus far this season, actually providing negative value as a framer, sitting in the neighborhood of notoriously poor framers Jarrod Saltalamacchia, A.J. Ellis, and Kurt Suzuki in the rankings.
What’s interesting about Martin’s struggles this season is that his zBall% is at 12.0%, or right in line with his excellent rates of the past five seasons. And it should be noted: Martin’s zBall% is considerably better than the 2015 rates of catchers like Saltalamacchia (14.0%), Ellis (17.6%), Suzuki (14.2%), or Navarro (23.1%). When there is a pitch thrown inside the strike zone, Martin is still elite at not letting them get away.
It’s Martin’s oStr% that has taken a considerable tumble, down to 6.1% after holding steady around 8% since 2010. In 2015, Martin has been stealing strikes at virtually the same rate as Saltalamacchia (5.7%), Ellis (7.1%), and Suzuki (6.9%). (Navarro is at an incredible 1.0%, but I think the point holds true.) This year’s elite framers are stealing strikes at considerably higher rates, like Chris Iannetta (10.5%), Yasmani Grandal (10.1%), or the Pirates’ own Martin replacement, Francisco Cervelli (11.9%).
Does a difference of two percentage points really turn an elite framer into a framer that’s costing his team value? Well, yeah. Now, a difference of two hundredths of a point in batting average — like from .240 to .260 — hardly has any influence on a hitter’s overall value. That’s because a hitter is getting, at the very uppermost, 700 plate appearances a year, meaning that a change by those two hundredths is affecting, at the most, the outcome of 14 plate appearances on the whole season. Catchers, meanwhile, are receiving thousands of pitches a year, with the league’s most prolific one or two catchers getting around 10,000 pitches caught (that is: pitches that actually make it into the catcher’s glove). Martin is already at 2,264 pitches caught this season: by losing two points of his stolen strikes, that means he’s already lost 45 strikes that he usually secures for his pitchers. Hey, that’s a few good innings of nothing but balls that Martin has typically converted into nothing but strikes! If Martin gets around his full-season average of 8,000 pitches received, that’s a difference of 160 pitches — more than a whole game of could-be strikes that are going down as balls. Expressed in terms of total pitch count like this, it’s a lot clearer (at least for me) how framing really does add up to full wins and losses.
So what’s Martin’s deal? A few days before the season started, FanGraphs alumnus August Fagerstrom wrote on how the Blue Jays were going to experiment by having Martin catch knuckleballer R.A. Dickey. Well, that experiment lasted three whole Dickey starts, with Dickey’s personal hombre Josh Thole catching his last four starts.
Did these three starts do in Martin’s numbers? Using Baseball Savant’s PitchF/X tool, one finds that Dickey threw exactly 300 pitches over those three starts, with 203 of them making their way to Martin’s glove. Martin successfully stole 15 strikes, or 7.3% of the eligible pitches — above his season average.
So it wasn’t even catching the only active knuckleballer that threw Martin off-course. What about the other Blue Jays pitchers? The following table is not totally precise, as many of the relievers have pitched a small handful of innings to Navarro and/or Thole. These are all numbers for the full 2015 season except in the case of Mark Buehrle, whose two season-opening starts pitching to Navarro have been excluded:
| Pitcher | Innings | Eligible Pitches | oStr% | BB/9 |
|---|---|---|---|---|
| Drew Hutchison | 36.1 | 400 | 9.5 | 3.22 |
| Aaron Sanchez | 32.1 | 384 | 4.9 | 6.96 |
| Daniel Norris | 23.1 | 303 | 8.2 | 4.63 |
| Mark Buehrle | 27 | 227 | 13.2 | 2.08 |
| Marco Estrada | 20.1 | 217 | 7.3 | 3.1 |
| Roberto Osuna | 18.1 | 182 | 3.2 | 2.45 |
| Jeff Francis | 12 | 153 | 6.5 | 3.75 |
| Miguel Castro | 12.1 | 135 | 4.4 | 4.38 |
| Aaron Loup | 13.2 | 124 | 3.2 | 1.32 |
| Liam Hendriks | 13 | 114 | 7 | 1.38 |
| Brett Cecil | 10 | 107 | 2.8 | 3.6 |
The correlation isn’t perfect, but I think it’s pretty strong: the Blue Jays have a lot of pretty wild pitchers, and Martin & Co. are having a hard time stealing strikes for them. As a staff, the Blue Jays were projected for the highest BB/9 in the preseason, and they’re at third in the league so far, trailing only the Colorado Rockies and Philadelphia Phillies. As Zachary Levine wrote at Just a Bit Outside in mid-April, the Blue Jays have a phenomenally young pitching staff, which makes the wildness understandable.
So here’s my theory: Martin has not, suddenly, lost his considerable gifts as a pitch-framer. Rather, when he sets up the target on the edge of the strike zone, hoping to steal a pitch on the fuzzy edge of the strike zone, this young staff is unable to hit the target, making it difficult to establish a convincing frame. You could almost blame the whole decline just on Sanchez. Usually we only talk about catchers when we talk about framing, but perhaps there’s a bit more teamwork at play when it comes to stealing strikes.
Miles Wray contributes sports commentary to McSweeney's Internet Tendency, Ploughshares, The Classical and Hardwood Paroxysm. Follow him on Twitter @mileswray or email him here.
BP’s pitch framing numbers have Martin as a very good pitch framer this year, like in every other year.
http://www.baseballprospectus.com/sortable/index.php?cid=1819124
What’s the difference in methodology? Yadier Molina is -5.7 at BP and +2.6 at SC. That’s a pretty big swing.
On a separate note, do these numbers just look at pitches out of the zone and see how many times they’re called a strike? So if it’s a pitch in the dirt, or a foot wide of the plate, it’s treated the same as a pitch an inch off the plate? If so, I would think the pitcher has a pretty big impact on these numbers. With a very wild pitcher, no catcher will have good framing numbers.
BP introduced a new, insanely comprehensive methodology this year. I was duly impressed, at least at the stuff I could understand:
http://www.baseballprospectus.com/article.php?articleid=25514
Statcorner: “The catcher is simply used as a grouping point. There’s no attempt to control for the pitchers, the umpires, the counts, or anything other than which side the hitter stood on.”
BP: “This new model allows simultaneous consideration of pitcher, catcher, batter, umpire, PITCHf/x, and other data for each taken pitch over the course of a season, and by controlling for each of their respective contributions will predict how many called strikes above (or below) average each such participant was worth during a particular season.”
Not sure why you’d even use StatCorner, frankly.
And I don’t really see this mentioned much, but there seems to be a lot of errors in pitchf/x. Not sure I trust the system to be accurate enough for this sort of thing.
That was my first impression. BP’s new system seems to be a quantam leap in pitch framing statistics. But I’m waiting for some smarter stats guys to tell me why I’m wrong.
Interesting. So we should probably see these numbers start to agree more as the year goes on, I’d think.
BP’s system looks at out of zone called strikes and in zone called balls. But these are compared to the league wide % for every pitch location, much as defensive stats are based on buckets of varying difficulty. So if a wild pitcher throws a ball well out of the zone, that location will have a very low % of called strikes, and the catcher will lose very little value if the ball is called a ball.
Most of the pf value comes from pitches that are on the edges of the zone, where they are called strikes an appreciable % of the time. When it’s called a strike, a catcher gets significant value, and when it’s called a ball, he loses significant value.
I think the main way a wild pitcher could hurt pf is if most of his pitches were either clearly in or clearly out of the zone, not giving the catcher a lot of opportunities at the edges.
P.S. the amazing thing about BP’s numbers, in short, is that they take into account the strike/pitch influence of all four of the catcher, pitcher, hitter, and umpire.
Right, that’s the beauty of machine learning
Without being labelled as some tinfoil hat wearing dude, I profess the Blue Jays are routinely shafted by umpires.
I think Bautista’s comments about them and their mediocrity and now the suspension and non apology by Brook Jacoby is not helping Russell either.
When are we going to get more thorough umpire data? We know when a strike’s a strike and a ball’s a ball, how is it no site has information on which teams get the most balls called strikes and who gets the most strikes called balls? Are there no numbers for this?
Even if there were numbers like this. MLB and the umpire’s union would never allow it to be made public, or published anywhere.
Who says this information is not public? It is. But nobody compiles it. I compiled some of it here, today, ironically:
http://www.reddit.com/r/baseball/comments/362yn6/analysis_which_teams_got_the_bestworst_strike/
Interesting results — fans of every team think their team is getting shafted. I’m not saying this to insult the original commenter, I’m just pointing out something that is universally true.
Nice. Now lets see batters as well. And then look at both hitters and pitchers over a 3-4 year span. 3-4 years should be enough time to bring everyone back to mean but I bet the results would show that some teams get favorable calls while others do not.
I miss you guys.
To anyone who watches Jays games the answer is pretty obvious. Umps have been cheating both our hitters and pitches for as long as I can remember. I would love to see an analysis by team of cumulative incorrect calls vs. Pitch Fx for hitters and pitchers. I am pretty sure the jays would be on the far negative end of the scale every year. Ever since they added Pitch Fx to Jays televised games 3-4 years ago I have become more and more disillusioned with the umps. Russell Martin has not forgotten how to frame. Some umps just dont see the game the same way when you are wearing a Jays uniform. End sad butt hurt rant.
I am not sure if this is because the team is from Canada or because a large portion of the players are Dominican and are very vocal and bitchy when they get bad calls. Its probably a mix of the two but one thing for sure is that certain umps absolutely hate Bautista. I would bet he gets shafted more than any other player in baseball.
cozeners’ link from above.
http://www.reddit.com/r/baseball/comments/362yn6/analysis_which_teams_got_the_bestworst_strike/
And you can see my comment immediately after. In short, no different from what fans of all teams say.
I watch a lot of the Jays games, and I think you’re right. At least what I’ve seen so far.
Especially Aaron Sanchez has gotten the shaft from the umps. He doesn’t seem to get the benefit of the doubt since he’s so wild. I’ve seen a few really bad calls go against him this year.
Sanchez is a tough one to frame, and tough to blame the umps. His stuff is crazy good, but it’s veering all over the place and a lot of the time it’s like neither he nor Martin have any idea where the ball is going to wind up. Hard to frame when he has to move the glove two feet from target to catch the ball, must be hard for the ump to call, too.
I do think Jays hitters get shafted regularly, harder to complain for those pitchers who are all over the place.
Aaron Sanchez….
The good fortune of getting to watch Mariano Rivera’s entire career leads me to believe that pitchers often have as much to do with “framing” as catchers do. Catchers for Mariano would typically set up just off the plate, and Mariano would typically hit the glove exactly. He got called strikes out of the zone all the time, and it wasn’t because of anything the catcher did, it was because Mariano’s command was such that he made the catcher’s job easy.
Of course, Mariano is a special case, but I would be surprised if pitchers do not deserve a lot of the credit for getting strikes called outside the zone, as well as for getting balls called inside the zone.
I would not be surprised to find that wearing a yankee uniform is responsible for getting more calls then either the pitcher or the catcher.
Then you’ve never seen Jorge Posada’s framing numbers…
as well as a boost to the Phantom tag percentage. Cough Jeter cough
Interesting. Martin IS at an age where you would expect his framing skills to start to decline, albeit not as dramatically as they have.
Of course we all get less good at physically-/reaction-intensive things as we get older, but is there actually data on framing skills declining at a certain age? Of all things in baseball, it strikes me as one of the easiest to do as you get older.
A note, they have Thole catching Dickey again only because of the injury to Navarro, Martin was doing a good job and he’ll start catching him again soon.
Very soon indeed – about 7 hours from now in fact.
Maybe it’s something about the Rogers Center backdrop? If the blue jays fans feel like they’re always getting shafted…
Can you look at Home and away splits on the framing data?
But then both teams would be affected in a given game and thus there would be no shafting to speak of.
I would be interested in seeing if there is a correlation between pitcher age/MLB service time and Ostr%…
Plus general reputation of the pitcher’s control developed over that service time. It seems notable that Buehrle is getting plenty of good framing from Martin in a small sample, which makes you think his ability and reputation for hitting spots is part of the catcher’s framing ability.
I don’t usually like “wah, the umpires are screwing us!” conspiracy stuff, but I do think there’s something there with the Jays. Be it because they have been particularly poor at hitting their spots/not walking guys in general this year (especially with a guy like Dickey who is known to throw a random/inconsistent pitch and Sanchez who has been all over the map), maybe it’s because Bautista & co have developed a rep for whining or routinely questioning the umps over balls & strikes, maybe the Jacoby thing made it worse (though I saw the issue before), maybe it’s the # of rookies and young pitchers on the roster at various points who have yet to “earn” the benefit of the doubt from the umps. It’s probably a combination of everything. But I swear I have seen the zone be more consistently slanted in favor of the opposition compared to the Blue Jays this season than I ever have in the past. In fact, last night’s game in Houston was the first time I can honestly say this season that I’ve seen the opposition get the same brutal calls Toronto does and the Jays get favorable ones.
I’d be interested in seeing the pitchf/x data for the Jays games this year to see if Toronto gets a greater # of unfavorable ball calls on pitches in the zone and Toronto opponents get more gift pitches than is normal for them when they take the mound against the Jays. Because my instinct so far this season is to say that that there’s been a marked and unusual advantage in favor of Blue Jay opponents.
It doesn’t excuse Toronto’s brutal-with-a-capital-B pitching performance. But it certainly doesn’t help.
The data is out there, I remember seeing statistics showing the Pirates were getting screwed by the umpires for a couple years during their losing streak. I can’t remember what it was taken from though.
Miles, I don’t understand how you get these numbers where you state the following:
When I did the search on Baseball Savant, I got 122 pitches that should qualify. I count all called balls and strikes outside the zone, excluding 3-0 counts and excluding balls in the dirt. Of those 122, 14 were called strikes, which is about 11.5%. I was going to double check the stats, but the site is down right now.
Martin is catching for Dickey again tonight. I think we should look at the Dickey effect helping other pitchers on our staff but it can also work against the catchers as well. Martin has showed difficulty a few times catching other pitchers immediately after Dickey pitched, it took him several pitches to adjust back to catching other pitchers. So their is an adjustment that both the hitters and Martin have to get used to immediately after. Would also
Would also like to point out that Sanchez makes Martin look worse. One pitch will be on the complete wrong side of the plate or at the backstop then the next pitch right on the corner and they are not getting those calls considering the previous pitch.
The difference between .240 and .260 has “hardly has any influence on a hitter’s overall value”? That’s well over one standard deviation of difference. Being very conservative and assuming 12 extra singles over 600 PA, that’s about 5 runs of value. Realistically, it’s probably closer to a full WAR – hardly “no influence”
I’ve been wondering about this for awhile, and I think I have a way better theory. Pitch framing stats are teh new sexy for a year or so now. (They’ve been around a bit longer, but they got the most press ever during last season.) I would be surprised if this didn’t come up in whatever Annual Umpire Meetings umpires have. And what would they discuss? Well, that Molinas and Martins and Lucroys have been fooling them for *years* now.
So if I’m an umpire armed with this information, if a pitch is borderline and I can perceive the catcher trying to sell me AT ALL I’m calling it a ball. Why? Because this dude steals strikes all the time and he’s probably trying to do it again, and I don’t wanna lose my job to some RoboUmp2000. I saw it happen to Yadi Molina several times during the ESPN game tonight: repeatedly pitches were borderline in the zone, and Yadi framed them well, but the ump called a ball. Molina looked shocked. Matheny yelled.
I.e., the counterfactual may matter. Just like good catchers can “shut down” the running game — and thus lose outs because nobody challenges them anymore — they might “lose strikes” if umpires think they’re playing them.