Catcher Framing Pt. 2

[absintheofmalaise]

Back in 2011 I started a thread about catcher framing based on an article on BPro by Mike Fast, who is currently employed by the Astros. Yesterday, BPro published a new article by Harry Pavlidis and Dan Brooks yesterday. They have come out with a new model, Regressed Probabilistic Model. Bad news for AL East teams, McCann isn't only a very good hitting catcher, he is also excellent at framing pitches. Here's a brief explanation from the article:

 
Since the beginning of the PITCHf/x era, researchers have calculated framing in several different ways. We are presenting a new method that we will call the "Regressed Probabilistic Model" of framing (RPM for short). In brief, RPM works by calculating the combined probability (and associated run value) that each pitch will be called a strike; summing those probabilities (and run values) across opportunities; attributing those values to a player (catcher or pitcher); and regressing "career" values to the mean.
We will freely admit: If you haven't seen the results of previous framing studies, it can be tough to wrap your mind around the size of the impact of a good or bad framing catcher. These effect sizes are not out of line with what has been reported in the past, but they're still obscenely large. Everyone agrees that Mike Trout was either a deserving MVP or a deserving runner-up in each of the past two seasons, which the stats say were worth close to 10 wins apiece. Our data suggest that over the past five years, the teams that have employed good framers like Jonathan Lucroy, Brian McCann, and Jose Molina have received essentially "free" MVP-caliber seasons from framing alone. (Each of those catchers has been worth about two extra wins per season over that span). This is a staggering amount of value. Add in the fact that these wins are almost assuredly not properly priced into the free agent market, and the difference between having a good framing catcher or a bad framing catcher can make or break a cost-conscious team.

 

 

[Just a bit outside]

As a former catcher I love reading about framing.  My first thought is how lucky to be a pitcher on the Braves when they had McCann and Ross as the catching tandem.
 
Looking at their player cards they were also good at blocking balls as well as framing.  Combined they saved over 39 runs every year from 2009-2012 with a high of 53 runs saved in 2012!
 
Edit:  Messed up earlier.  Ross did not go to Atlanta until 2009.

 

[Sprowl]

Neither Pierzynski nor Saltalamacchia show up on the leader or trailer boards. Pierzynski is slightly below average on most metrics, while Salty rated very well in 2012, but close to average in other years. David Ross and Jose Molina show up on just about every leader board. It would be interesting to compare RPM Framing Runs with catcher age -- is framing a skill that takes longer to master, and should be expected to peak in a catcher's 30s? If so, there might be additional costs to going with a Swihart/Vazquez platoon during their early major-league years.

 

[Jnai]

These are being more completely built as we speak, but all the data is here:
http://www.brooksbaseball.net/c_landing.php?player=150229
 
(search for any catcher in the top right)
 
For AJP:

A.J. Pierzynski has caught 112,539 pitches that have were tracked by the PITCHf/x system between 2007 and 2013, including pitches caught in the MLB Regular Season, the MLB Postseason and Spring Training. In 2013, in 6,314 framing opportunities (i.e., taken pitches), he was a league average framer, which resulted in 28.88 fewer strikes than would be expected from a league average catcher. This was of below average value, resulting in 4.64 lost runs (worth approximately -3 million dollars).

 
and Salty:

Jarrod Saltalamacchia has caught 69,929 pitches that have were tracked by the PITCHf/x system between 2007 and 2013, including pitches caught in the MLB Regular Season, the MLB Postseason and Spring Training. In 2013, in 6,418 framing opportunities (i.e., taken pitches), he was a league average framer, which resulted in 13.77 extra strikes than would be expected from a league average catcher. This was of average value, resulting in 0.56 saved runs (worth essentially no additional value).

 

[Jnai]

And, I noticed a typo in the text, which has been fixed. =)

 

[PaulinMyrBch]

Good thread. I'm always surprised we don't talk about catcher data in more detail around here. It's one thing we can actually time at home with a stopwatch. We can get pops and backpicks on a stopwatch and have pretty good angles to all the frames and sticks. Occasionally we'll see a comment in the game thread that the catcher got that strike, but when Tek quit throwing people out, the discussion was always dead legs/arm/etc, but posting his pop time is pretty easy with a DVR and a stopwatch.

 

[The Gray Eagle]

Lavarnway rates as a slightly above average framer from 2010 to 2013. Considering he is still relatively young for a catcher and his defense has reportedly improved over the last couple of years, that is promising, even if he's nothing special at it yet.

 

[OttoC]

I must admit that the first thing I thought of when I started looking at the article was whether it could be adapted to passed balls/wild pitches and I was going to contact Dan when I came across the side bar on that very subject.

However, I get "file not found" when I click on Blocking and Defense for Pierzynski: http://www.brooksbaseball.net/c_landing.php?player=150229

 

[Jnai]

I must admit that the first thing I thought of when I started looking at the article was whether it could be adapted to passed balls/wild pitches and I was going to contact Dan when I came across the side bar on that very subject.

However, I get "file not found" when I click on Blocking and Defense for Pierzynski: http://www.brooksbaseball.net/c_landing.php?player=150229

 
Still being built. =( BP sortables has them.

 

[VORP Speed]

So Jose Molina is basically Mozart of pitch framing. McCann is good, but in 2013 Molina was twice as good. 24 runs saved vs 12.
According to the dollar value calculation Molina's pitch framing was worth 16m. He's making 2.25m in 2014.
Hanigan saved almost 10 runs in 2013 (in a somewhat reduced # of opportunities). Worth 8m according to the calculation. He's making 2.75m in 2014. So the Rays catching platoon is worth 24m and saving ~35 runs and costing them 5m? That is a staggering amount of excess value if the pitch framing metrics are indeed accurate.

 

[jon abbey]

Yep, I had quite a few fights on here about how valuable Molina was when he was on the Yankees from 2007-2009, I was a big fan despite his anemic hitting. It's great to see the catcher defensive metrics moving forward at long last, someone should alert Molina's agent.

 

[VORP Speed]

Yep, I had quite a few fights on here about how valuable Molina was when he was on the Yankees from 2007-2009, I was a big fan despite his anemic hitting. It's great to see the catcher defensive metrics moving forward at long last, someone should alert Molina's agent.

Maybe that's why the Rays signed Molina for 2 years this offseason, despite his age, and Hanigan for 4.....because they knew Jnai was going to come out with this stuff and screw up the market inefficiency.

 

[Frisbetarian]

I'm surprised this topic has had such little activity. I think Dan's (our own Jnai) and Harry's catcher framing and blocked pitches study is among the best conceived and implemented baseball research projects I have ever seen. They accounted for pitcher contribution, made umpire adjustments, utilized a with or without you (WOWY) analysis, have a solid year to year correlation, and even regressed (slightly) to the mean. This is a pretty exciting and meaningful new metric that deserves more reads and comments. I'm moving this topic to the main board in the hopes more members will see it and chime in.
 
I have a few questions/comments. First, I know what you are trying to do by combining your stat with WARP/WAR, but I don't think it really works. WAR and WARP are based on average values, while the framing stat is context (pitch count) dependent. Also, WAR/WARP are (mostly) not regressed while this metric is. Finally, WAR/WARP already have issues with differing accuracy on their component stats, and adding another with uncertain (at this point) precision just muddles an imperfect stat further. 
 
I'm not sure yet how I feel about having this count context dependent. I understand pitches are called differently in different counts, but worry that using the run values based on pitch count could lead to over or understated totals. It's great you show both framing runs above and below average using your method and an average value on the BP website. I need to look at those numbers and think more about this, but would appreciate any comments you may have on the subject.
 
Are you as confident in the blocking/wild pitch data/stats as the framing info? How much are those numbers regressed?
 
Obviously, the -50.6 run framing stat for Ryan Doumit in 2008 really sticks out. I see from the BP stat page it was even higher, -55.6 runs, using an average run value, so context did not hurt him. For Doumit that season, 383 of the 2485 taken pitches that should have been called strikes were called balls because of his lack of catching skill, over 15%. That seems really high to me, and it does not include clear cut obvious strikes. Do you know how many no brain strikes there were in his 1988 predicted strike total. Obviously, his missed strike percentage would be much higher looking at just "borderline" pitches.
 
Which leads to my final comment, which I hope makes sense. You guys are using an all or nothing approach to balls and strikes called in this study, and because of that have had some pretty high above and below expected runs saved via pitch framing results. Would it be possible to find the percentage of times a pitch is called a strike in a certain zone, say + or - .3" on a ball 21" high (depending on batter height) and + or - .3" on a ball 12" left of the center of the plate. If that is possible, and the pitch described above is called a strike 38% of the time, then a catcher who receives it as a strike would get credit for .62 extra strikes and one that received it as a ball would get a -.38. Again, I have no idea if this makes sense or is even possible, but the really high run totals do strike me as a potential issue.
 
I hope the above does not seem critical. I am incredibly impressed by this study, and thank Dan and Harry for making it public. I really hope you guys will consider presenting your research at this summer's Sabermetrics, Scouting, and the Science of Baseball conference. 

 

[Jnai]

Which leads to my final comment, which I hope makes sense. You guys are using an all or nothing approach to balls and strikes called in this study, and because of that have had some pretty high above and below expected runs saved via pitch framing results. Would it be possible to find the percentage of times a pitch is called a strike in a certain zone, say + or - .3" on a ball 21" high (depending on batter height) and + or - .3" on a ball 12" left of the center of the plate. If that is possible, and the pitch described above is called a strike 38% of the time, then a catcher who receives it as a strike would get credit for .62 extra strikes and one that received it as a ball would get a -.38. Again, I have no idea if this makes sense or is even possible, but the really high run totals do strike me as a potential issue.
 
 

 
I'll respond to the other stuff in a bit, but just to make it clear, this is exactly what we do. We create a probability map based on the location and of the pitch, handedness of the batter, handedness of the pitcher, height of the batter (we normalize it), and count, and then credit or debit the battery based on the difference in probability from the actual pitch outcome and the actual call that was made.

 

[Frisbetarian]

 
I'll respond to the other stuff in a bit, but just to make it clear, this is exactly what we do. We create a probability map based on the location and of the pitch, handedness of the batter, handedness of the pitcher, height of the batter (we normalize it), and count, and then credit or debit the battery based on the difference in probability from the actual pitch outcome and the actual call that was made.

 
Thanks. That's explained right at the beginning of the article, too. Mea culpa. I was looking for a Doumit explanation, but maybe he's just that bad.

 

[Idabomb333]

This is awesome, thanks to the authors and to Abs/Fris for sharing here.  
 
I'm not at all sure that this is a reasonable concept, but the section of the article about Yadier Molina being good and also having to "give back" some of his value to the pitchers got me thinking.  Does it seem possible that there's a synergy effect there?  Suppose a pitcher throws a lot of strikes but hasn't really built up the reputation with umpires yet, and then he combines with a catcher who's good at framing.  He gets even more strike calls now, and maybe that adds to his reputation and then he gets more pitcher-based strike calls too, etc.  If the same pitcher combines often with a catcher bad at framing, maybe he never develops the reputation as a strike thrower, and never gets the pitcher-based strike calls.
 
I don't know enough statistics to know whether the section explaining why the pitcher adjustment is good accounts for this sort of thing.  I would think that if the pitcher effect for good catchers is usually negative and the pitcher effect for bad catchers is usually positive, that would have stuck out to you guys and been handled.  Just thought I'd throw the idea out there in case it helps.
 
I'd be interested to look for other things like that, especially given the inclusion of context.  For example, does there seem to be any particular skill (or loss of skill) for anyone or a subset of people in framing 0-2 pitches?  Are some catchers particularly good at framing sliders?  Something like that could add to synergy with pitchers too - have the guy who frames sliders catch the starter who throws a lot of sliders and the guy who frames curves catch the starter who throws that more, etc.  I have no idea whether there's even really enough data to measure those ideas with any confidence, but I hope you guys are planning on trying to get even deeper with questions like that if you think you can.

 

[Idabomb333]

Another thought: I imagine there are pitchers who throw more or fewer borderline pitches.  Maybe Koji throws so many undeniable strikes that catcher framing for him is essentially irrelevant, and then you have a Carlos Santana type catch him without the same reservations.

 

[ALiveH]

The whole discussion of catcher framing highlights how consistently bad umpires are at calling balls & strikes.
 
I think this proves without a shadow of a doubt that balls-strikes should be called in real-time by some mechanical non-human system.
 
Ironically if that happened that would kill the value of  catcher framing.

 

[Reverend]

That's not what irony is. That would be simple observation, assessment, execution and causation resulting in intended change.

 

[Mighty Joe Young]

I find the whole catcher framing skill completely fascinating .. but , I'm really not sure I believe the magnitude of some of the reported effects. This kind of reminds me of Voros McCracken's work - which I think major league FO's take seriously , but have never really completely bought into. For obvious reasons - its rather difficult to buy into the concept that a ball in play hit off of Pedro Martinez has as much chance of being a hit as one off of some scrub up from AAA for an emergency start. (As I understand it, the latest incarnation of DIPS DO subscribe some skill to the pitcher)
 
The proof that Catcher Framing is not generally accepted as an important skill would seem to be in the salary scales of catchers who are good at this. If Molina was really a 6 Win player one would think he could command upwards to 20m a year - instead he's practically an NRI. Players of his ilk had better hope that it becomes widely accepted - AND that robot umps never grace a major league ballpark.

 

[Saints Rest]

This is awesome, thanks to the authors and to Abs/Fris for sharing here.  
 
I'm not at all sure that this is a reasonable concept, but the section of the article about Yadier Molina being good and also having to "give back" some of his value to the pitchers got me thinking.  Does it seem possible that there's a synergy effect there?  Suppose a pitcher throws a lot of strikes but hasn't really built up the reputation with umpires yet, and then he combines with a catcher who's good at framing.  He gets even more strike calls now, and maybe that adds to his reputation and then he gets more pitcher-based strike calls too, etc.  If the same pitcher combines often with a catcher bad at framing, maybe he never develops the reputation as a strike thrower, and never gets the pitcher-based strike calls.
 
I don't know enough statistics to know whether the section explaining why the pitcher adjustment is good accounts for this sort of thing.  I would think that if the pitcher effect for good catchers is usually negative and the pitcher effect for bad catchers is usually positive, that would have stuck out to you guys and been handled.  Just thought I'd throw the idea out there in case it helps.
 
I'd be interested to look for other things like that, especially given the inclusion of context.  For example, does there seem to be any particular skill (or loss of skill) for anyone or a subset of people in framing 0-2 pitches?  Are some catchers particularly good at framing sliders?  Something like that could add to synergy with pitchers too - have the guy who frames sliders catch the starter who throws a lot of sliders and the guy who frames curves catch the starter who throws that more, etc.  I have no idea whether there's even really enough data to measure those ideas with any confidence, but I hope you guys are planning on trying to get even deeper with questions like that if you think you can.

It seems to me that another potential synergy -- I'll camm it the Greg Maddux effect -- is that if you are a pitcher with good to excellent control with a catcher who is excellent at framing pitches, in tandem you would be able to expand the strike zone quite significantly and create an even larger donut hole.  Conversely, a patcher who never get the borderline strike will have to shrink the strike zone and reduce the size of the donut hole.  It's not hard to see how either of these extremes could directly affect a pitcher's success rates, not only in terms of BB and K, but also in terms of LD%, BABIP, etc. 

 

[Mighty Joe Young]

Another interesting angle on this:
 
The more this "skill" is publicized and generally recognized as valid the less it's likely to be seen. Umpires read this stuff too. If you're behind the plate and Molina is catching how much of a tendency is there to start squeezing the pitcher? They know he's good at making balls look like strikes - how long before the Umpire is making an adjustment as well? Not to mention the batter swinging at more borderline pitches. Both of these factors should, and probably will, contribute to a decline in the observed skill.

 

[ToeKneeArmAss]

LTTP, but wanted to add to the praise for the work done here by Dan and Harry.
 
Another idea to explore, assuming the data is available in sufficiently large sample sizes: I wonder whether certain umpires are prone to mis-call certain pitches in certain situations?  For instance, does Umpire X tend to call breaking balls from LHPs to RHHs on 0-2 counts when they are near the outside middle of the plate (ie 'backdoor breaking balls')?
 
Knowing whether these tendencies exist would yield value in two ways - it would allow the analysis above to be further adjusted for the situation (ie, not all pitches within +/- 1 inch of the outside middle border are created equal, so the % call strikes/called balls could be contextually varied), AND it would be of value to teams to know when umpire/batter-hitter-matchup/count situations lend themselves to calling certain pitch types/locations.
 
Great thread.

 

[ToeKneeArmAss]

And a question - does the analysis done to date factor in the value of stealing a strike depending on the count?  I.e., since a stolen strike on 3-2 yields a tremendous differential in outcome value, does the analysis give that more value in terms of ultimate wins created through framing than a strike stolen on a 3-0 count (to cite two presumed extremes)?

 

[Reverend]

Jesus Christ you're smart.

 

[Just a bit outside]

And a question - does the analysis done to date factor in the value of stealing a strike depending on the count?  I.e., since a stolen strike on 3-2 yields a tremendous differential in outcome value, does the analysis give that more value in terms of ultimate wins created through framing than a strike stolen on a 3-0 count (to cite two presumed extremes)?

Yes. This is in the original article.

 

[Hagios]

 

Another interesting angle on this:
 
Umpires read this stuff too. If you're behind the plate and Molina is catching how much of a tendency is there to start squeezing the pitcher? They know he's good at making balls look like strikes - how long before the Umpire is making an adjustment as well? Not to mention the batter swinging at more borderline pitches. Both of these factors should, and probably will, contribute to a decline in the observed skill.

My understanding is that good pitch framers are guys who keep their gloves set up well ahead of time at the right spot with little movement. They may subtly shade towards the center of the plate, but they aren't doing much. I think it's more that the bad pitch framers hurt themselves by stabbing at borderline pitches and then obviously snapping their glove back (or leaving it way out of the zone - damned if you do, damned if you don't). In other words, good pitch framers are basically "plain vanilla" and bad pitch framers are the oddballs. I don't think umpires are going to be inclined to give the oddballs much credit.

I do wonder if players will start changing their approach based on a catcher's framing.

 

[Frisbetarian]

What I find most surprising about this is the huge run value a few framed/blown calls can create over the course of a season. The study has the value of single pitch, depending on count, between .08 (0-0 count) and .590 (3-2 count) runs, with the average value just over .14 runs. Looking at the 2013 offensive numbers in MLB after each pitch count, this seems reasonable.
 
[tablegrid= 2013 Offensive Stats By Count ]Split PA BA OBP SLG OPS First Pitch 20499 0.336 0.341 0.540 0.882 After 1-0 73403 0.269 0.378 0.433 0.811 After 2-0 25491 0.279 0.495 0.458 0.953 After 3-0 8499 0.289 0.739 0.491 1.230 After 0-1 90971 0.223 0.263 0.338 0.601 After 1-1 72634 0.239 0.307 0.372 0.679 After 2-1 37476 0.255 0.387 0.412 0.800 After 3-1 15303 0.275 0.577 0.467 1.044 After 0-2 36575 0.167 0.196 0.248 0.444 After 1-2 51774 0.179 0.228 0.271 0.499 After 2-2 42026 0.193 0.289 0.301 0.590 Batter Ahead 62486 0.297 0.463 0.494 0.957 Even Count 62075 0.270 0.275 0.421 0.696 Pitcher Ahead 60312 0.202 0.210 0.294 0.504  [/tablegrid] 
But as we saw with Ryan Doumit's 2008 number, 50.6 runs below average, it can add up fast over a number of games. Doumit played just 909 innings behind the plate that season, 101 games, and missed 383 pitches, just under 4 per game. In that limited time, he, according to the study, cost his team ~ 5 wins. Looking at it another way, if a batter recieved the benefit of 2 calls per game, less than 1 every other at bat, he would, on average, be 42 runs better than another batter who didn't get any calls over 150 games. For reference, David Ortiz led the Red Sox in batting runs above average at +35.3, and no other Red Sox player was over +20. In fact, only 5 MLB players in 2013 generated more than 42 batting runs above an average player. 
 
If these numbers are even close to accurate, and I see no reason why they are wrong (and also understand the ball and strike calls should tend to even out a bit over the course of a season), then I agree with Alive that an automated ball and strike system is overdue in MLB. 

 

[Idabomb333]

What I find most surprising about this is the huge run value a few framed/blown calls can create over the course of a season. The study has the value of single pitch, depending on count, between .08 (0-0 count) and .590 (3-2 count) runs, with the average value just over .14 runs. Looking at the 2013 offensive numbers in MLB after each pitch count, this seems reasonable...
<Snip>
 
Looking at it another way, if a batter relieved the benefit of 2 calls per game, less than 1 every other at bat, he would, on average, be 42 runs better than another batter who didn't get any calls over 150 games. For reference, David Ortiz led the Red Sox in batting runs above average at +35.3, and no other Red Sox player was over +20. In fact, only 5 MLB players in 2013 generated more than 42 batting runs above an average player. 
 
If these numbers are even close to accurate, and I see no reason why they are wrong (and also understand the ball and strike calls should tend to even out a bit over the course of a season), then I agree with Alive that an automated ball and strike system is overdue in MLB. 

Great point.  I wonder how many runs were affected by replay on home runs in 2013.  Catcher framing runs clearly dwarfed that number, whatever it was.  I wonder if all of the replay in 2014 will even approach the impact of correctness in strike zone calls.
 
The obvious difference is that the framing issue comes up a heck of a lot more often, with smaller individual case impact.  A replay system for balls and strikes wouldn't work.  But automation, or at least supplementing umpires with Pitch F/X data, should be a bigger deal relative to replay in the discussions about how to make sure the game on the field corresponds to the game in the rule book.

 

[Hairps]

Great work, and an interesting thread. A few meandering thoughts come to mind, from someone who is definitely uneducated in this area...
 
- I started to wonder the other day how much of framing could be attributed to where the catcher sets up. I was watching an NC State vs. Florida State game a few weekends ago, and you couldn't help but notice the late break and tailing of Carlos Rodon's pitches. The color guy, a former college coach, opined that because a number of the pitches had such crazy late movement that they looked to be crossing thru the strike zone but then darting down and away so hard that the catcher had to consistently move his mitt to get it, costing Rodon a number of otherwise called strikes in the process. Then, a few innings later, on a steal attempt, that same catcher -- NC State's Brett Austin -- went to throw out the runner and even though the batter did not move out of the way there was no apparent batter interference because he didn't get in the way of Austin's throwing motion. And then it seem to sort of strike the announcer and me at the same time...Austin is consistently set up way too far back in the strike zone!
 
Anywho, if/when we get to the point of looking at ways to teach framing, I wonder if this would be a simple enough place to start.
 
- Seems to me it will be very tricky to properly assign the credit for framing whenever folks get to the stage of rolling it into a catcher's overall WAR, or whatever version of a unified player value you want to use. Especially, as it relates to the pitcher involved. If a catcher is getting credit for making a non-strike a strike, how would you then appropriately "penalize" the pitcher for throwing a non-strike for which he was credited a strike. Of course, maybe this has already been appropriately decided and discussed -- I no longer have my BPro subscription. SORRY!
 
I've got a bunch of other random thoughts but clearly I need to dive into this and the other catcher framing research in greater detail first. Damn, SoSH is pretty amazing sometimes.

 

[SumnerH]

 And then it seem to sort of strike the announcer and me at the same time...Austin is consistently set up way too far back in the strike zone!

 
The catcher can't set up in the strike zone, they'd be called for interference all the time.  The batter has the right to a legal swing, which means the catcher must be far enough behind the zone not to interfere with them.

 

[Hairps]

 
The catcher can't set up in the strike zone, they'd be called for interference all the time.  The batter has the right to a legal swing, which means the catcher must be far enough behind the zone not to interfere with them.

 
Well, shit, you know what I mean. Not *in* the strike zone, but behind the strike zone whatever you call the place where the catcher squats. He was squatting much farther back in that area-not-called-the-stike-zone than most catchers, and farther back than that former catching coach had seen before.

 

[wolfe_boston]

Am I the only person that believes that the concept of pitch framing is a bunch of media generated crap? There are too many variable to make any accurate statistical analysis as the pitcher, umpire, pitch selection and execution and batters are different every day. For example, how do you reconcile a pitcher who has a reputation for throwing strikes, a batter known to rarely take strikes and catcher who is suposedly is a good framer. If I'm an umpire, I'm going to look at the location of the pitch; I'm not going to care where the catcher catches the ball. Many moons ago, when I was in Little League, there were kids who would "frame" pitches. I thought it was stupid then, too.

 

[Laser Show]

Am I the only person that believes that the concept of pitch framing is a bunch of media generated crap? There are too many variable to make any accurate statistical analysis as the pitcher, umpire, pitch selection and execution and batters are different every day. For example, how do you reconcile a pitcher who has a reputation for throwing strikes, a batter known to rarely take strikes and catcher who is suposedly is a good framer. If I'm an umpire, I'm going to look at the location of the pitch; I'm not going to care where the catcher catches the ball. Many moons ago, when I was in Little League, there were kids who would "frame" pitches. I thought it was stupid then, too.

 
As an umpire, you'll have a tiny fraction of a second to look at that location of the pitch and then call it. Framing is the catcher's attempt to influence limited human vision - the umpire sees where the ball is caught much longer than where it crossed the plate, thus influencing his decision, consciously or not.
 
Not the most elegant explanation, but still, pitch framing is legit and not even remotely media bs.

 

[WenZink]

If pitch-framing is a bunch of baloney, then how come the most common missed strike call is when a pitcher misses his spot horribly and a catcher has to reach across to make the catch?  I've seen catchers set up off the outside corner, then lunge back over the zone, and what was supposed to be a ball out of the zone, ends in the zone for a called ball.  (Admittedly. the lunge by the catcher may be an alert to the umpire that something went wrong.)

 

[wolfe_boston]

As an umpire, you'll have a tiny fraction of a second to look at that location of the pitch and then call it. Framing is the catcher's attempt to influence limited human vision - the umpire sees where the ball is caught much longer than where it crossed the plate, thus influencing his decision, consciously or not.
 
Not the most elegant explanation, but still, pitch framing is legit and not even remotely media bs.

You're saying there is subliminal effect caused by the catcher's actions. If it exits, how do you measure it? Does it happen 10 times a game? Often, the announcers will talk about how the umpire is being consistent with his strike zone. How is this possible with different catchers?

As to the fraction of a second, the umpire has had thousands of reps to develop a sense of the relationship of the "blur" to the plate, which never moves. A batter has to use the same fraction of a second to put his bat in the way of the pitch which would seem harder than just identifyingnthe location. I also believe that an umpire is blocked by the catcher's body from seeing where a low pitch is being caught.


Finally, are there articles out there where umpires admit that there are certain catchers that cause them to alter their strike zones?

 

[Skip]

If you watched the Brewers this year you'd see the importance of pitch-framing. Lucroy is the king of it and all of their pitchers are benefitting big time from it.

 

[derekson]

How could anyone watch last night's game and see the difference in strikes that McCann was getting called by framing them well from the ones Pierzynski wasn't framing well and wasn't getting called and then say pitch framing is a myth? Last night was pretty much a perfect demonstration of the value of a catcher who frames well.

 

[rymflaherty]

How could anyone watch last night's game and see the difference in strikes that McCann was getting called by framing them well from the ones Pierzynski wasn't framing well and wasn't getting called and then say pitch framing is a myth? Last night was pretty much a perfect demonstration of the value of a catcher who frames well.

 
Last night's umpire, Wolcott, has been in the majors less than a year.
I wonder if the importance of framing becomes even more exaggerated with less experienced Umpires?  Purely speculation on my part, but it's not a stretch to think a less experienced umpire may be influenced by that sort of thing more....And it certainly would help explain last night.

 

[wolfe_boston]

How could anyone watch last night's game and see the difference in strikes that McCann was getting called by framing them well from the ones Pierzynski wasn't framing well and wasn't getting called and then say pitch framing is a myth? Last night was pretty much a perfect demonstration of the value of a catcher who frames well.

OK, let's say I concede that good pitch framing has a positive effect on the number of called strikes. Does it ever affect the outcome of a game? Based on game thread posts, AJP, among other things, absolutely sucks at pitch framing. In our 1/8 of a season sample, has he cost us a game? If so, can we expect him to cost us 8 games over the season?

If I have Catcher A who has a higher projected WAR than Catcher B, but is considered to be worse at pitch framing, how much of a gap in projected WAR do I need before I should feel comfortable in choosing Catcher A?

Baseball has so many meaningful performance measurement statistics that I don't see the benefit off pontificating over something can't even remotely be quantified.

 

[Just a bit outside]

OK, let's say I concede that good pitch framing has a positive effect on the number of called strikes. Does it ever affect the outcome of a game? Based on game thread posts, AJP, among other things, absolutely sucks at pitch framing. In our 1/8 of a season sample, has he cost us a game? If so, can we expect him to cost us 8 games over the season?

If I have Catcher A who has a higher projected WAR than Catcher B, but is considered to be worse at pitch framing, how much of a gap in projected WAR do I need before I should feel comfortable in choosing Catcher A?

Baseball has so many meaningful performance measurement statistics that I don't see the benefit off pontificating over something can't even remotely be quantified.

Please find the thread on pitch framing. It has been quantified and it matters.

 

[Hee Sox Choi]

Pitch framing article about Ryan Hanigan:
 
http://grantland.com/the-triangle/qa-ryan-hanigan-on-the-art-of-pitch-framing/
 
I found this article to be very informative and was bummed to see him traded to the Rays.  

 

[Reverend]

Pitch framing article about Ryan Hanigan:
 
http://grantland.com/the-triangle/qa-ryan-hanigan-on-the-art-of-pitch-framing/
 
I found this article to be very informative and was bummed to see him traded to the Rays.  

 
Our very own Dan Brooks is teaming up with Harry Pavlidis over at Baseball Prospectus to develop a new model of understanding the framing issue. You can read about it in the thread here which also has a link to earlier Baseball Prospectus stuff on the matter.
 
Also: the Red Sox ought to win this evening.

 

[Sampo Gida]

You're saying there is subliminal effect caused by the catcher's actions. If it exits, how do you measure it? Does it happen 10 times a game? Often, the announcers will talk about how the umpire is being consistent with his strike zone. How is this possible with different catchers?

As to the fraction of a second, the umpire has had thousands of reps to develop a sense of the relationship of the "blur" to the plate, which never moves. A batter has to use the same fraction of a second to put his bat in the way of the pitch which would seem harder than just identifyingnthe location. I also believe that an umpire is blocked by the catcher's body from seeing where a low pitch is being caught.


Finally, are there articles out there where umpires admit that there are certain catchers that cause them to alter their strike zones?

 
 
Comparing hitters, who are young and in the 99.999% percentile of the population in being able to see pitches and react to them and umpires, some of them in their 50's,  who come from a pretty dilute talent pool is rather silly.  As for the umps being blocked by the catcher, maybe.   Umpires standing off to the side of the plate to avoid foul balls also makes it harder for them to call a pitch based only on the path of the ball.
 
I think its very likely some umpires discount that blur and recognizing they don't have the best angle they attach more weight to where they think the ball ends up.    On a pitch the  catcher catches without excess glove movement and makes it look like the pitcher hit his spot, the lack of movement carries weight.  I have seen catchers set up with their glove 3 inches off the outside edge of the plate and get that call all night long.   Ever see a case where  a pitcher misses location badly but still throws a clear strike and it gets called a ball because the catcher had to move his glove a lot?  Happens all the time.
 
I have heard anecdotal reports of catchers tipping the umpire to location by tapping there leg so they get a better call since the umpire may have trouble covering the entire strike zone visually.  I recall Francona saying he asked one umpire in all seriousness if he was actually able to see Lesters pitches, suggesting some of the worst umpires might only hear pitches from certain hard throwing pitchers.
 
How else do you explain umpires missing pitches right down the middle in some cases, or pretty close to that, or calling balls that are 3-6 inches off the plate as strikes.
 
As it is, one study I ready says umpires get calls wrong 14% of the time, another had it as low as 10%.  That's average.  If an ump is is twice as bad as the average he is a bit more accurate than flipping a coin, but not that much more.

 

[Reverend]

Seriously, where the hell were you guys when that pitcher framing thread--what might be possibly the next great inefficiency to be exploited--was posted?

 

[wolfe_boston]

Seriously, where the hell were you guys when that pitcher framing thread--what might be possibly the next great inefficiency to be exploited--was posted?

All though I have been a serious Sox fan for 40+ years I only joined SOSH a few weeks ago.

 

[wolfe_boston]

I still have the same question; is it even remotely possible to calculate a pitch framing WAR component?

 

[Reverend]

I still have the same question; is it even remotely possible to calculate a pitch framing WAR component?

 
You would calculate it through runs saved the same as you do with some of the advanced fielding stats components of WAR, except that yo wouldn't because WAR sucks.
 
Even David Cameron of FanGraphs has said WAR is problematic because it actually destroys useful data by folding it into an amorphous lump of abstracity.

 

[Reverend]

OK, he didn't say "amorphous lump of abstracity" but he did say the rest.

 

[Jnai]

Dave Cameron does not think WAR sucks. For whatever reason, "WAR Sucks" has become a thing that SoSH believes. WAR has problems, but all measurement systems have problems. It throws away useful data, but all measurement systems throw away useful data.
 
Anyway, Wolfe, I encourage you to read the article. The difficulty is the "R" part of WAR, which is difficult to measure for pitch framing because we have a relatively limited pool of players. Our "Wins" are reported as "Above Average", which is slightly different, but also useful.
 
[Article link: http://www.baseballprospectus.com/article.php?articleid=22934]