Quick and Lazy Team-by-Team 2014 Framing Projections

Just because there aren’t any pitches being framed these days doesn’t mean we can’t still talk about pitch-framing. When the Yankees picked up Brian McCann, it got me thinking about how well they’re going to frame as a team. When the Pirates picked up Chris Stewart, I tweeted out they might be the best framing team in baseball. It was a statement off the top of my head, and there are in truth a handful of contenders, but I decided, why wait? Why not try to figure this out now?

Somewhere over the past couple months, we all stopped looking at 2013 statistics and started thinking about 2014 statistics. We don’t have any 2014 statistics, but we do have projected 2014 statistics, so that’s how we’re informing a lot of our thoughts. Right here at FanGraphs, we provide projected batting data, projected pitching data, and projected fielding data. Something I haven’t seen, though, is projected pitch-framing data. So this is a first pass, which I assure you is both quick and lazy. But a start nevertheless.

A couple things I’ll be up front about: we don’t know how much of framing is the catcher, and how much is the pitcher. Wilder pitchers might generate different data from pitchers with command, so the same catcher can look different on two different teams. And we don’t know how heavily to regress framing data, aside from “a little more heavily” when you have a smaller sample size. We don’t know quite what one year of framing data means. I don’t, anyway; team employees might, but they don’t drop me informative emails out of the blue. What’s going to follow is kind of sloppy, but I think it still gives a good general idea in the end.

What we have is 2013 pitch-framing data. What we also have, right here, is projected 2014 playing time behind the plate. I did the simplest thing I could think of: for each catcher I applied 2013 data to 2014, given a sufficient 2013 sample size. Lacking that, I took numbers from 2012. For players with little or no big-league track record, I just put in a zero for extra strikes per game. With those numbers in place, it became a matter of doing team-by-team calculations, weighting by playing time. The result is the following sortable table. Shown are projected extra strikes per game, along with the change in that stat from the same team last season.

Team PerGame Change
Brewers 1.83 -0.17
Pirates 1.22 0.43
Padres 1.16 0.22
Rays 1.10 0.10
Yankees 0.87 -0.71
Cardinals 0.72 -0.25
Indians 0.72 0.30
Giants 0.59 -0.03
Nationals 0.52 0.63
Angels 0.52 0.29
Rangers 0.46 0.87
Mariners 0.09 0.95
Red Sox 0.01 -0.27
Blue Jays 0.00 -0.69
Athletics -0.07 -0.20
Reds -0.13 -0.44
Diamondbacks -0.20 -0.17
Marlins -0.32 0.64
Orioles -0.32 -0.18
Astros -0.33 -0.16
Mets -0.41 0.82
Phillies -0.45 -0.44
Royals -0.47 0.08
Tigers -0.52 -0.85
White Sox -0.56 -0.60
Dodgers -0.67 -0.12
Braves -1.08 -1.56
Cubs -1.33 -0.44
Twins -1.40 -0.31
Rockies -1.54 -0.33

The Pirates project to be really good. The same goes for the Padres, Rays, Yankees, and others. But they don’t hold a candle to the Brewers, who return Jonathan Lucroy and Martin Maldonado. This shouldn’t come as a surprise to people familiar with these sorts of numbers, since Lucroy and Maldonado have been cult framing All-Stars. If they do what they did a year ago, the Brewers should get almost two extra strikes per game, relative to average. Other teams are down closer to one. The Pirates do indeed take a step forward, it’s worth noting, having gotten Stewart to replace Michael McKenry.

At the other end of the table are the Rockies, Twins, and Cubs. And also, for the first time in forever, the Braves. It’s news to no one that Wilin Rosario is a bit raw and shaky defensively. For the Twins, Josmil Pinto had some rookie trouble getting borderline strikes for a lousy pitching staff. The Cubs are going to give a lot of time to Welington Castillo, and thankfully he can hit some and throw some. The Braves don’t have McCann anymore.

And it’s the Braves who have the biggest negative projected change from 2013. When it comes to framing, it’s a double-whammy to lose McCann and give more playing time to Gerald Laird. I don’t actually agree with the Braves’ FanGraphs depth chart, myself — it’s giving too much playing time behind the plate to Ryan Doumit, who should catch hardly at all. Change that and the numbers get better, and Evan Gattis actually looks good by the framing data. But Laird is bad, too, in the same neighborhood as Doumit, so the position’s still worse in the receiving department. The Tigers also slip, replacing Brayan Pena with Bryan Holaday. Holaday, however, is a bit of an unknown.

The biggest positive projected change from 2013 belongs to the Mariners, who’ll be starting Mike Zunino all season long. Right behind them are the Rangers, who added the defensively-improved-but-offensively-miserable J.P. Arencibia. In third are the Mets, looking ahead to a whole year from Travis d’Arnaud. The second tier of projected improvements includes the Marlins, Nationals, and Pirates. For more than half of baseball’s teams, there’s a projected change of less than 0.40 extra strikes per game in either direction, because of the lazy methodology used. I want to emphasize how lazy this is.

Again, it’s not right to just use 2013 data as 2014 data. And again, we don’t understand how much of this is actually pitcher-dependent. Maybe, say, Pinto’s quite a bit better than his numbers, and the numbers will reflect that in 2014 when he starts catching an actual major-league pitching staff. And players can change from year to year, especially if they’ve never really had framing-focused instruction before. Arencibia took a big step forward between 2012 and 2013, and other players could make improvements now that this data is out there and more and more accepted. This is an exciting field, but there are a lot of moving parts, and I can’t pretend to have it even half-figured out.

But as a first pass of projected framing numbers, there you go. All 30 teams, in order. A better methodology would probably generate a very similar table in a similar order, with bigger numbers regressed to slightly smaller numbers. There are teams who are weaker in this department, there are teams who are stronger in this department, and there are the Brewers. The Brewers look like they’re going to be the best framing team in baseball, and then we’ll just have to wonder why they still don’t pitch very well.

Print This Post

Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

33 Responses to “Quick and Lazy Team-by-Team 2014 Framing Projections”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Boris Chinchilla says:

    Good work Jeff, have a beer now

    Vote -1 Vote +1

  2. Heh says:

    Do the A’s framing projections have Jaso at catcher? Because he’s being taken off of catcher and moved to the DH spot, it looks like 99% of innings will be caught by Norris or Vogt.

    Vote -1 Vote +1

  3. Catoblepas says:

    Interesting to me that, even with Yankees picking up McCann, the loss of Stewart is enough to make them the team with the third-largest drop in extra strikes/game.

    Vote -1 Vote +1

  4. pft says:

    19 of the 30 teams look to be worse in 2014

    Vote -1 Vote +1

  5. Z..... says:

    I’m guessing the Marlins +change projection is based on the idea that Saltalamacchia is more like his previous years than 2013?

    Vote -1 Vote +1

    • FanGraphs Supporting Member

      Saltalamacchia was basically average last season, and Brantly was bad, and Hill was just dreadful. Limited time, but big negative impact. Some could be a relatively inexperienced pitching staff.

      Vote -1 Vote +1

      • Z..... says:

        I should have realized it was the fact that Brantly likely wont see much time there at all. It was really amazing how often strikes were taken away from our pitchers last year. Seeing that article here on Jose Fernandez and those issues in August wasnt surprising to me at all

        Vote -1 Vote +1

  6. Brad Johnson says:
    FanGraphs Supporting Member

    I find the Doumit acquisition to be a bit confusing. He was clearly acquired to be a catcher, but as you say, he may not spend much time behind the plate. There’s no place for him in the outfield unless Heyward or JUpton get hurt, same with first base and Freeman. So he’s just going to be an expensive backup on a team that is resource constrained? Or is he actually going to catch with some frequency?

    Vote -1 Vote +1

  7. Franco says:

    Framing has to Jeff’s favorite geek stat at this point.

    I wonder as it gains popularity if there is some kind of push back by umps.

    Vote -1 Vote +1

    • Kirk Davenport says:

      I have never seen an ump admit that he as swayed in any amount by a catcher “framing” Nor can I even see film of an ump even making the slightest glance at the glove position of the catcher. So I put framing in a league with the tooth fairy and Jolly Old St. Nick. I will leave out the myth of the “G” spot since I have seen it to be true.
      Even if framing was a serious reality, does 1 strike in a game or a percentage point of a strike in a game enough of an influence for such investigation? There are probably 150 pitches thrown by each side and maybe 100 strikes each (rough estimates – I know some of you can give exact numbers) and certainly many more things influence the number of strikes more than the decimal numbers presented – pitching decisions, umpire bias, batter stance, game situation, men on base, defense positioning to think of a few

      Vote -1 Vote +1

      • Yan Fucking Gomes says:

        Ha, Kirk.

        Vote -1 Vote +1

      • William Hargis says:

        Let’s make some poor assumptions. Let’s assume every fifth extra strike helps produce an out. With 162 extra strikes a year and every fifth one producing an out, that might lead to approximately 32 outs. Considering that a game is approximately 27 outs, I would say thats a signficant amount for an entire season. Again this was very lazy because we didn’t observe any of the extra strikes and the direct impact on the game.

        Vote -1 Vote +1

      • ankle explosion hr celebration says:

        It’s up to about 3 strikes/game at the extreme, but for a decent pitch framer, let’s say only 1.5. Those strikes come at the expense of balls; a ball is negative, a strike is positive (for the pitcher).

        1 Ball -> 1 Strike is worth ~.15 runs.

        Do the math, 1.5 strikes per game at .15 runs per strike and 162 games, the total runs is ~35 runs over a full season.

        That’s roughly 3.5 wins per season. And, as I said, that’s conservative; a very good pitch framer will get more strikes per game, and compared to a very bad pitch framer, the difference will be (obviously) higher.

        That’s why it’s worth investigating.

        Vote -1 Vote +1

      • Balthazar says:

        No.1) Of course no ump is going to admit he was influenced by a catcher. Calling the strike is literally the umps ‘product’ and they are very territorial about it. Even though (well, _especially_) we can tell from park cameras that human beings, in this case umpires, are rather B-A-D at making fine visual judgments at velocity in space.

        No.2) Far more importantly, the influence of pitch framing is perceptual rather than cognitive. That is, an umpire likely literally can’t tell when he is being influenced by framing because everything happens so fast perceptually relative to cognative understanding of what was percieved by occular physiology. There are any number of studies of visual cognition which indicate that there actions to quick to be understood in fact impact behavior because they are not too quick to impart physiological trace perceptions. The ump may just feel he got a better look at a borderline pitch without being able to tell you why or how. If one has any understanding of how ‘good framing’ is believed to work, this is easy to grasp, since good framing has less distracting motion that will blur the cognitive impression, and even more so good framing is stable where bad framing moves the final grasp of the pitch away from the strike zone so that the final cognitive impression would be of a pitch less of a strike.

        No.3) What umpires think they are doing is not really all that important unless they think they are deliberately calling a pitch ‘in favor’ of a particular outcome for reasons having nothing to do with where the pitch goes. I think that is seldom the case in baseball. In football and basketball on the other hand . . . .

        Vote -1 Vote +1

      • Jasper says:

        I can’t speak from experience as a level any higher then HS ball myself, but as an umpire at that level and having worked with guys who have some college and minor league experience I can tell you they say the same things I say.
        Catchers absolutely make a huge difference in framing pitches vs pulling pitches.
        Most quality umpires are looking for strikes and a catcher that frames properly will get every strike we (as umpires) can give him. A catcher that pulls pitches will have those borderline pitches called balls, the rational from umpires: if you are pulling the pitch either you suck at catching or you are trying to fool me, but it is very obvious to an umpire the difference. And the only reason to pull a pitch is because the catcher doesn’t think it was a strike and if he thinks it is a ball I am inclined to believe him.
        As for looking at the glove, umpires are trained to set up so that we can track the ball from pitchers hand to catchers glove without moving our head, as head movement is like taking a picture with a cheap camera while moving it, it goes out of focus and you can’t see what you are looking at.

        I will come back here to post more from an umpire’s POV, I have just seen so much of this stuff and the conjecture on what umpire’s see and do, and the why behind it. That info is out there to help analyze some of these sabermetrics, you just need an interested umpire willing to give some insight.


        Vote -1 Vote +1

  8. DrBGiantsfan says:

    I still can’t get past the notion that a pitcher with poor command is going to be much harder to frame than one who puts the pitch right where you set the glove. Do we know that the variance from pitcher-to-pitcher is less than the variance from catcher-to-catcher? It also might be interesting to go back and look at whether catcher numbers changed when they changed teams.

    Vote -1 Vote +1

    • Pirates Hurdles says:

      Agreed, I think this is the heart of the argument that top framing catchers aren’t really adding 30 runs a year. Framing is important, but there is almost certainly a pitcher component.

      Vote -1 Vote +1

    • William Hargis says:

      Watching Glavine pitch to a spot was infuriating because I didnt like the ever expanding strike zone some umpires provided to him. If you can acknowledge that Glavine got a larger strike zone, its not hard to envision a savy catcher like Molina that is able to manipulate an umpire to calling a strike.

      Vote -1 Vote +1

      • Nathan says:

        I hated that 6 inches off the plate strike call he and Maddox always got. And those two assholes could hit it every time.

        Vote -1 Vote +1

    • Balthazar says:

      You’re framing the concept a bit incorrectly, Dr. (HADDA say it.) Framing is not about where the catcher sets his glove before the pitch; framing is about how stable a catcher _keeps_ his glove as he receives the pitch into it.

      Pitchers with bad command won’t throw as many strikes, and won’t get as many close calls from an umpire. That said, it is just as possible for a catcher to ‘win a strike’ on a call under those circumstances as any other. It just won’t matter as much because the when a guy is missing a lot the value of any given additional strike called is going to be less in aggregate.

      All of the talk about ‘just a few more strikes’ misses the issue that some strike calls are more important than others. Getting the call to have your pitcher go ahead 1-2 really matters in the outcome of an at-bat, for instance. It’s unlikely that a catcher can reliably influence the specific moment he ‘wins a strike,’ the effect of framing is more an aggregate effect as I understand it. One shouldn’t assume that it is a random effect, however. Since getting the strike call is always a positive, I’ll take a guy who is ever a little more positive, other things being roughly equal. I definitely think framing matters more than how good an arm a catcher has (unless a catcher is absolutely awful, in the > 25% caught range).

      Vote -1 Vote +1

  9. Rauce says:

    I’d like to read up on how pitch-framing data is gathered? Any suggestions?

    Vote -1 Vote +1

  10. jmpmk2 says:

    Considering the Rockies were also 6th in pitching WAR last season, might any regression be counteracted by a commensurate improvement in pitch framing?

    Vote -1 Vote +1

  11. ttnorm says:

    I wonder if anyone has studied whether framing is projectable from year to year. I know umpires read this stuff and laugh. I doubt that any catcher near the top of the chart is going to get a break from an ump ever. I wouldn’t be surprised if the Supervisor of Umps discusses these results with his team.

    Vote -1 Vote +1

  12. Payroll says:

    Let me get this right. The Twins dump Doumit, pick up Suzuki, and project to frame worse next year?

    Vote -1 Vote +1

    • Balthazar says:

      Suzuki is likely to catch more than Doumit did, and Suzuki is absolutely awful at framing. Yes, the Twins got worse . . . .

      Vote -1 Vote +1

  13. JZ says:

    Do deep backstops have an effect on framing? I remember in little league feeling more pressure to keep balls in front of me when there were really deep backstops.

    Vote -1 Vote +1

  14. zeke4au says:

    I’m probably missing something obvious, but what made the Yankees go down so much with the addition of McCann, while the Braves also went down so much for losing him (and for having more Laird)?

    Vote -1 Vote +1