Can someone tell me if I can trust these pitch framing numbers at all?

They seem like ... nonsense.

I mean, I have an open mind, but I don't think the "generous" reading of those numbers is to dock Jaso 4 wins. The generous reading is to say that we have virtually no confidence in the statistic, but we can probably assume that a Molina is better around the dish than a non-Molina.

His career is now long and varied enough to have achieved the sojourner box set.

Seriously, 4 wins is nuts, though he does qualify that's only if both guys caught 1000 innings. Still, that's roughly the equivalent of saying the difference between them defensively is the average Ozzie Smith season & Derek Jeter's worst season*. Not buying that, sorry.

*according to B-R's D-WAR

Count the ringzz.

I mean, I have an open mind, but I don't think the "generous" reading of those numbers is to dock Jaso 4 wins. The generous reading is to say that we have virtually no confidence in the statistic, but we can probably assume that a Molina is better around the dish than a non-Molina.

Seriously, 4 wins is nuts, though he does qualify that's only if both guys caught 1000 innings. Still, that's roughly the equivalent of saying the difference between them defensively is the average Ozzie Smith season & Derek Jeter's worst season*. Not buying that, sorry.

I have no dog in this hunt and agree it sounds nutty. So, without trying to steer to a conclusion, let me make a couple of assumptions I will pull out of... thin air.

What's a respectable guess, albeit of the wildass sort, as to what the top-notch framing of a single pitch might be worth? No idea. So, what about one hundred pitches--could framing interfere over one hundred pitches with a batter's ability to get a hit--could it add enough strikes, and take away enough balls, to elminate one hit? Sure. By which I mean, the idea of that doesn't make me cringe with disbelief. Our catcher, then, saves a hit every six innings he catches, and if he catches 120 full games, or 1080 innings, more or less, he's saving over his season around 180 hits, or around, what, 70 runs?

70 runs sounds crazy. Our artful framer would be winning 7 extra games for his team over the course of a season. That can't be. (Can it?) Could it be half that? Could The Artful Framer be saving a hit every 200 pitches, thereby adding 3-4 wins to his team's total?

Seriously, 4 wins is nuts, though he does qualify that's only if both guys caught 1000 innings. Still, that's roughly the equivalent of saying the difference between them defensively is the average Ozzie Smith season & Derek Jeter's worst season*. Not buying that, sorry.

Given that it's generally agreed that catcher is a more difficult defensive position than shortstop, why would we expect the range of defensive talent at the position to be less than SS or other infield positions? Intuition would suggest the opposite, wouldn't it?

I thought that was a joke chart until I read the comments here.

Given that it's generally agreed that catcher is a more difficult defensive position than shortstop, why would we expect the range of defensive talent at the position to be less than SS or other infield positions? Intuition would suggest the opposite, wouldn't it?

That makes sense, but the 40 runs *just on pitch-framing* seems extreme. If you asked a group of learned baseball men to rank the defensive responsibilities of catchers, I'm guessing that pitch-framing would be somewhat down the list. You also have game-calling, controlling the running game, blocking bad pitches, overall infield play (fielding bunts and such), and blocking the plate.

Of course, when you think about it, given that catchers receive thousands of pitches, it might make sense that this is their best opportunity to make an impact. But then again, I've been hearing for about 20 years that we ought not to pay attention to Catcher's ERA, that there is no real discernible ability there.

That chart is clearly not to be believed. It shows Molina worse in blocking (wild pitches/passed balls I assume).

A guy that big is like having a billboard planted behind home plate. Nothing gets past him! It just bounces off his chest, love handles, shoulders, whatev.

Given that it's generally agreed that catcher is a more difficult defensive position than shortstop, why would we expect the range of defensive talent at the position to be less than SS or other infield positions? Intuition would suggest the opposite, wouldn't it?

Right, but what this chart is suggesting is that the difference in talent between two random catchers is the same as the difference between the best shortstop ever and the 9th-worst shortstop season since 1950.

His career is now long and varied enough to have achieved the sojourner box set.

Is that a Magnolia Electric Co. reference, or is that just a thing they referenced?

Edit: Molina, yeah, just got it. Nicely done.

Shouldn't we be able to correlate pitch framing with something else?

Like BABIP? Or K rate? Or BB rate?

Something like that would go a long way for me. My position on the pitch framing stuff is that on one level it rings true that statheads have been mis-evaluating catcher defense for some time, but this stuff seems extreme.

Here is the thread on the original study of catcher framing by Mike Fast. Mike makes some comments in the discussion, the most relevant might be his last comment (#76):

I wanted to mention here that I think my understanding of what Dan's figure of 0.13 runs/pitch meant, as I expressed in #65, was not correct. I need to figure out the ultimate run impact, which I think will be less than what I said in my original article but more than 50% of what said due to some of the reasons I explained in #65, even if some of the things I said in that post were wrong. In other words, your objection in #63 has more merit than I originally realized, even if the impact is less than 2x.

There are a number of factors that I need to improve upon to tighten up the valuation in my model, and this will be one of the important ones.

I'm not aware of a follow-up by him, but I don't tend to see things that aren't linked here at BBTF.

My personal sense is that there's something here, and I'm fairly confident that Mike's study got the general order of catchers right in terms of this ability (i.e., Molina really is one of the best at framing). But I do think the range of impacts is too wide. If I had to guess, I'd at least halve it, but that's just a guess.

Don't think this ever got posted here so:

from Tango

I'm sure that's still preliminary but seems a lot of the effect found in the original analysis is really pitcher effect. But he still comes up with a guesstimate of the true talent difference between best and worst of about 4.5 wins per 125 games. Still seems too big to me but there ya go.

One thing claimed in the article (and I find this kinda hard to believe too but seems the fundamental sort of thing Tango wouldn't have wrong) is that changing a ball to a called strike is worth .12 runs. He adjusts that under the sensible notion that we're most likely talking about changing a "probable ball" into a "probable strike" and goes with .08 runs.

Like I said, it seems high to me but assuming it's roughly correct, it means that getting one extra K per game (hardly an outlandish level of framing ability) would save 1 run per 12.5 games or 10 runs per 125 games. So a 4-win difference over 125 games is Molina bringing two balls into the strike zone every game while Jaso carries two balls out of the strike zone every game. Given 150 pitches per game that's a seemingly miniscule difference in ability having a major impact on the result. (Of course of those 150 pitches, how many are borderline strikes which are not swung at?)

So really the believability of the stat would seem to come down to whether that .08 runs per framed pitch is correct/believable.

The other upshot of that article is that Derek Lowe is a thief!

My personal opinion is that if pitch framing is really worth that much, then the automated strike zone needs to be in place by the start of the 2012 season. The strike zone is there so that the batter can at least have some chance at receiving a hittable baseball. It's not there so that catchers can play a game of "three-card monte" with umpires.

I don't think we need to worry about the poor hitters being able to see enough hittable pitches. Has that been a problem in the past?

I don't think we need to worry about the poor hitters being able to see enough hittable pitches. Has that been a problem in the past?

That misses the point. The point is the manner in which the catcher catches the ball has ZERO effect on a hitter's ability to hit the pitch. So it should have the same effect on the strike zone (which was invented so that hitters could expect hittable pitches regularly). That it apparently doesn't have ZERO effect on the strike zone, represents a problem IMO. And if the difference is as large as the study says, it's a BIG problem.

Why wouldn't an effect this large show up in CERA? What would be the expected difference in CERA with this large of an effect?

I've sat behind home plate for a lot of games at a lot of levels. The way the catcher behaves has always seemed to have a large effect on how the umpire sees the ball. The spread of results of Fast's original study and the subsequent follow ups by Tango and crew don't suprise me one bit.

I was at a game last week and the catcher let one go off the tip of his glove and it hit the ump in the nuts, who promptly went down in a heap.

It was a fastball RIGHT DOWN THE MIDDLE. (I think mixed signals). Needless to say, the pitcher didn't get the call. ;)

My personal opinion is that if pitch framing is really worth that much, then the automated strike zone needs to be in place by the start of the 2012 season. The strike zone is there so that the batter can at least have some chance at receiving a hittable baseball. It's not there so that catchers can play a game of "three-card monte" with umpires.

It seems like pitch framing is exactly the kind of thing that pitch/fx should be used to quantify. Should be relatively straight-forward to compare called strikes/balls vs pitch/fx strikes/balls for given catchers. Make an adjustment for the "called zone" effect, etc.

So wait, are you people saying I'm supposed to just accept this? That your catcher's umpire-bamboozling ability is just as important as your centerfielder's overall ability?

So wait, are you people saying I'm supposed to just accept this? That your catcher's umpire-bamboozling ability is just as important as your centerfielder's overall ability?

My point exactly. If this effect is real, it's something that needs to be fixed, not scouted for.

I just don't see how it's any different from any other playing skill. Pitchers can use whatever type of delivery they want to, within the rules, even if it's "deceptive" in terms of keeping the ball from the hitter's view longer, having it come from an unusual angle, etc. Should the rules be changed to prevent that? That doesn't exactly prevent the hitter from hitting the ball--only from seeing the ball, which is only part of the ability to hit it. Hitters, in turn, can stand anywhere they like as long as it's in the batter's box and not on top of the plate. They move around to get a better view. Fielders position themselves differently when the ball is pitched, because they think it helps them get breaks. Framing pitches is like those things--a skill that can be taught and scouted for, and if it has this kind of effect, why not scout for it and teach it?

I just don't see how it's any different from any other playing skill. Pitchers can use whatever type of delivery they want to, within the rules, even if it's "deceptive" in terms of keeping the ball from the hitter's view longer, having it come from an unusual angle, etc. Should the rules be changed to prevent that?

It's based on deceiving the umpire and not the other players. The umpires should be given the tools they need to be objective when calling strikes (rammed down their throats if necessary).

Framing pitches isn't morally questionable, it's just that it's only valuable due to the limits of the current system. It can hardly be controversial to want balls and strikes called accurately, and the better that gets the less valuable framing is.


Framing pitches is like those things--a skill that can be taught and scouted for, and if it has this kind of effect, why not scout for it and teach it?

Sure, there are no robo umpires on the horizon, just don't sign a long-term contract based on framing ability.

Framing pitches is like those things--a skill that can be taught and scouted for, and if it has this kind of effect, why not scout for it and teach it?

No, it's not. I'm with Voros here. The rules define a strike zone. If a catcher has a skill which allows him to fool the umpire into calling a ball a strike, that is fundamentally different from those other skills. It is perfectly OK to fool the batter, fool the runner, deke the fielder. That is well withing the bounds of fair play. It is not (or should not) be OK to fool the umpire. The competition is supposed to be player vs player, not player vs arbiter.

why the assumption that it's all balls being called strikes. How much of it is actual strikes that were framed properly and helped the umpire see the true nature of the pitch ?

Poor framing turns strikes into balls sometimes too.

Poor framing turns strikes into balls sometimes too.

Well, that's the other side of the same coin. Same point applies.

I agree that if pitch framing is worth several wins a season, then MLB needs to start looking at rule changes. The proper analogy is something like a batter being very good at pretending to get hit by a pitch. This part of the game is undesirable but the overall effect is probably quite small. The feasible solutions would create a "medicine worse than the disease" situation. If pitch framing can be worth 3-4 wins a season, then I for one welcome our new robo-umpire overlords.

So really the believability of the stat would seem to come down to whether that .08 runs per framed pitch is correct/believable.

I'm pretty sure that's correct. All other things being equal. But maybe they're less likely to get a "framed" third strike. Or something of that sort.

As others have pointed out, what's missing is something that demonstrates that all things are in fact equal. Something like Craig Wright's matched innings study. Or something using Keith Woolner's method and see if you can pick up a K rate influence, a BB rate influence, or some kind of influence on balls in play.

Keith's methods won't detect influences under ~15 runs a full season, but we're talking an influence far too large to hide from a study of this nature.

Part of the reason for the large range between the best and worst catchers is that catchers with poor mechanics can't get a lot of strikes called as strikes. As I recall the analyses, the worst catchers don't receive the ball smoothly and their movements aren't quiet; instead they are jerking the mit around more, with more head and body movement. Perhaps it leaves a subconscious impression on the umpire that the pitcher is erratic and can't hit his target.

The interesting aspect of the pitch framing studies to me is that it seems to close the gap between what many good pitchers have said about catchers over the years and the prevalent sabermetric skepticism of those views. Roy Oswalt used to talk about how incredibly important Ausmus' receiving skills were to the pitcher's performance. For most of us, it was, how can all of this talk of a quiet or soft receiving target be that important? Yet Roger Clemens and Andy Pettitte considered Ausmus a prerequisite to signing with the Astros. At one time, the Red Sox considered signing Ausmus to try to attract Clemens back to Boston. Greg Maddux had his own catcher caddy who followed him from the Braves to the Cubs. The value associated with catcher pitch framing perhaps explains why pitchers' preferences for catchers with good receiving skills was not irrational.

As I recall the articles in the Hardball Times, the study controls for ball/strike differentials attributable to the pitcher and umpires.

It seems to me this would have more to do with a pitcher missing his target than any ability of the catcher to fool an umpire. The typically egregious call is when the catcher has to move from one side of the plate to the other to catch the ball from where he initially set up. Umpires are trained to pay attention to the plate and the batter's stance when he's in hitting position, not where the catcher's mitt is. I think this idea of framing is more "steadiness". Doubtless, the sheer girth of a Molina impedes the view of the umpire and throws some more guess work into things.

> Pitchers can use whatever type of delivery they want to, within the rules, even if it's "deceptive"

I think we need to control for the pitchers before we get completely worked up about the catchers. There is plenty of anecdotal evidence that a deceptive motion can fool an umpire just as much as it fools a batter. Greg Maddux and Tom Glavine were known for their control which helped give them the benefit of the doubt on close calls. This system is going to give all that credit to the catcher. Of course right now all the credit is currently assigned to the pitcher when some of it should go to the catcher.

It has been obvious for years that catchers are getting shortchanged by WAR calculators. This is just the first step in remedying that issue.

I think you'll see a lot of resistance from catchers about automating the strike zone judgement calls. If that happens there will be a strong incentive to trade catching ability for offense. You'll have a whole new "era" in the defensive spectrum of players. I'm also not sure you can make that change until you can ripple it down to the minor leagues. It will completely change how catchers are selected for the majors.

I am personally not interested in whether or not this study proves that we ought to have robo-umps. I just want to know if it's true! I mean, this is some serious #### here. 3 wins based on just some small fraction of a catcher's defensive duties. People should be going crazy, DIPS style.

I think we need to control for the pitchers before we get completely worked up about the catchers.

They address this on The Book thread that's linked above. A lot of it is indeed pitchers. But even when they parse out the pitcher's influence, the catchers still have a huge ability to affect balls/strikes, and thus run-scoring. It looks like a 20-30 run swing between the best and worst, not counting Ryan Doumit, who is on another planet of terribleness in this statistic.

I think you'll see a lot of resistance from catchers about automating the strike zone judgement calls. If that happens there will be a strong incentive to trade catching ability for offense. You'll have a whole new "era" in the defensive spectrum of players.

I agree with the first sentence, but not the rest. As I said above, pitch framing is only one of many duties that the catcher has, and it's not usually mentioned as one of the most important. I'm sure you can find baseball men that would not be surprised that a good framer can steal a win or two for his team, but you can also find baseball men that will tell you that Omar Vizquel saves a run every single game, or whatever. One of the interesting findings on that The Book thread is that it looks like major league backups are superior pitch-framers than are major league starters.

One of the interesting findings on that The Book thread is that it looks like major league backups are superior pitch-framers than are major league starters.

This is what we'd expect, isn't it? Aren't most backup shortstops superior to starting shortstops?

3 wins based on just some small fraction of a catcher's defensive duties.

I really disagree that pitch framing is a small fraction of a catcher's defensive duties. What we've seen over the last 10 years is a complete misunderstanding of catcher defense. I think that everyone can agree on that.

According to our best available metric, Pete Rose was worth more in his prime defensively than Johnny Bench was. I think most of us would agree that isn't true. The catcher defense numbers coming out of Mike Fast's study seemed to match better with actual MLB decisionmaking--i.e., if pitch framing didn't matter at all (our previous estimate of value) why not put anyone with a bat back there?

MLB teams have operated for a long time on an assumption of high defensive value for catchers. They have treated catchers similarly to shortstops. IMO, catcher defense has been a pretty big missing piece for some time now.

According to our best available metric, Pete Rose was worth more in his prime defensively than Johnny Bench was.

Which metric is this?

Regarding the 0.08 run value for a successfully framed pitch, it seems to me that it should be possible to stop guessing and actually assign a unique run value to every single pitch that the catcher receives. If he somehow turns a pitch that is called a ball 99.9% of the time into a strike, he should get the full 0.12 runs. But if he's turning a pitch that is called a ball 53% of the time into a strike, the value should be much less. The majority of framed pitches will be borderline pitches, pitches that are routinely called both ways. It might be a gargantuan mathematical task, but you should be able to score every single pitch. And we might arrive at a very different value.

I really disagree that pitch framing is a small fraction of a catcher's defensive duties.

If you polled baseball men, where do you think it would rank among the duties I mentioned in #9? I think it would be about halfway down the list. Maybe I'm wrong.

IMO, catcher defense has been a pretty big missing piece for some time now.

I agree with this.

why not put anyone with a bat back there?

There are actually pretty good reasons not to do so. One is that catchers play only about 80% of the games that players at other positions play, so you're missing out on a big bat there. Another is that the rigors of catching can negatively affect your offense. Plus you have to consider that leaving a guy who's not very good at pitch blocking and throwing can cost you 10-15 runs per year. Add it all up and the Blue Jays probably made the right call nearly 2 decades ago when they gave Carlos Delgado a 1B mitt.

But then again, I've been hearing for about 20 years that we ought not to pay attention to Catcher's ERA, that there is no real discernible ability there.

This is my big issue with thr "framing" study.

If a catcher was saving 20 or 30 runs a year above average, it should show up in CERA, but it doesn't.

I mean, 20 runs in 1000 innings in a 4.00 ERA run environment is almost 0.20 of ERA.

The CERA has other issues - the top catchers defensively would, logically, be catching the younger/wilder pitchers to help calm them down and build them up (see the movie Bull Durham for a simple example). However, if you do check CERA you see that Molina has done better than his teammates over the vast majority of his seasons which does suggest that something is there.

iirc I-Rod had poor CERA's early on then improved as he got older. As a 20 year old he was known to focus just on catching baserunners (important when guys like Tim Raines and Rickey Henderson were still playing) and not paying attention to anything else back there. Meanwhile Mike Piazza was the #1 catcher for a team that regularly was near the top of the league for ERA despite a spaghetti arm. A shame we don't have this data for the 90's as it would be interesting to see what it would say about those two who were viewed as one of the best ever defensively and one of the worst ever largely on their ability to throw out baserunners.

If a catcher was saving 20 or 30 runs a year above average, it should show up in CERA, but it doesn't.

Pitch framing doesn't impact batted balls. Batted balls are a huge component of ERA so you're probably in the noise. It should show up in K/BB rates. Knowing that pitchers are the primary input for K/BB, can we show that certain catchers improve K/BB rates of pitchers?

So wait, are you people saying I'm supposed to just accept this? That your catcher's umpire-bamboozling ability is just as important as your centerfielder's overall ability?

No ... or yes. I don't "believe" it either but I don't have evidence to contradict it. And, as I noted, if the run differential of a borderline ball being called a strike (or borderline strike being called a ball) is really .08 runs then the difference we're talking about is Molina/Martin stealing just two strikes a game while Jaso/Doumit are losing just two strikes a game.

You raise a good point of why this doesn't show up in CERA, assuming what we both understand about CERA not mattering is true?

It should also show up as a similar effect among batters. Presumably taking that borderline pitch vs. swinging at it (and usually missing it or hitting it weakly) has a similar level of effect. We certainly rave about plate discipline and about batters who see, oh, 4.2 vs. 3.8 pitches per PA. If 240 (mostly borderline) pitches per 600 PA is a big deal for batters why not 250 borderline pitches per 125 games caught for a C?

Baseball is a game where seemingly trivial differences in outcomes (reaching base one extra time per week) can make a "huge" difference (40 points of OBP).

Also, if you don't like the C results, check the Tango study (where he kinda addresses the CERA question but I don't really buy it). Anyway, McCann steals 2-2.5 strikes per game ... while Derek Lowe is stealing 4! Tango doesn't go into detail on the pitchers but Lowe is #1, Livan #2 and Moyer #3 over the last few seasons -- aren't these exactly the pitchers we think work the edges and get by on weaker stuff by getting the calls?

This goes to Voros' point and extends it beyond Cs -- Lowe is stealing .32 runs per 9 innings.* Granted, at this point in his career he needs all the help he can get but, wow. (Note Lowe & McCann together were stealing .5 runs per 9 innings.)

Regarding the 0.08 run value for a successfully framed pitch, it seems to me that it should be possible to stop guessing and actually assign a unique run value to every single pitch that the catcher receives. If he somehow turns a pitch that is called a ball 99.9% of the time into a strike, he should get the full 0.12 runs. But if he's turning a pitch that is called a ball 53% of the time into a strike, the value should be much less. The majority of framed pitches will be borderline pitches, pitches that are routinely called both ways. It might be a gargantuan mathematical task, but you should be able to score every single pitch. And we might arrive at a very different value.

Excellent idea. I might have missed it in the thread but, also, in a typical 140 pitch game, how many pitches are actually close enough for catcher framing to matter? Those would be pitches that (probably) aren't swung on, that are within something like two or three inches off the plate, and within two or three inches over the plate. Twenty? Less?

A question for statistics majors: Is it possible to determine the margin of error for a study without knowing in advance the results of that study? What I'm clumsily trying to say is, with instruments available in the 18th century, without knowing what techniques of measurement might eventually become available in the future, it could be said with certainty that with 18th c. tools we knew the circumference of the earth within something like 2%. Can we do the same with framing?

#41 That's why I mentioned Craig Wright's study. You control for the pitcher by looking at pitcher/catcher combinations. This is also the basic method that Keith Woolner used.

Kiko/14, my best estimate at the current time is that the effect is about 65% of what I said it was in the study I published, but it may be somewhat higher than that due to the effect preferentially happening on two-strike counts. Max Marchi found an effect of similar size to what I originally reported.

My study is here:
http://www.baseballprospectus.com/article.php?articleid=15093

Max Marchi's study is here:
http://www.hardballtimes.com/main/article/evaluating-catchers-framing-pitches-part-3/

In addition to the effect likely being somewhat smaller than I reported (due to the effect Kiko mentioned in the other thread, of calls not being flipped from 100% ball to strike, but some lower probability), the numbers that DRaysBay is reporting are the unregressed (observed) numbers. The regressed (estimate of persistent skill) for Molina is +28 runs per 120 games and for Jaso -5 runs per 120 games.

If you take my best estimate that the real effect is 65% of what I originally reported, then that's +18 runs for Molina and -3 runs for Jaso, or about two wins difference per season for framing.

tshipman/13, Greg Pope/19, and Ron J/30, snapper/40, DL/42, Walt Davis/43,
I address in the BPro article that this is seen both in K rate and BB rate and is plausible based on the catcher ERA effect from Sean Smith's study in the 2011 THT Annual. There is a catcher ERA effect that appears to be larger than the framing effect I found.

Sleepy/21,
My study was based upon PITCHf/x data.

Teixeira/32,
I adjusted in my study (as did Max in his) for the effect of pitchers, which is presumably largely their ability/tendency to hit their catcher target at the edge of the zone.

Preserved/38,
I've modeled it for an average pitch distribution, and it comes out to about 65% of 0.12 runs/call, which is around 0.08 runs. But I would like to go back and do the study over while taking into account the actual run value of each call for each catcher/pitcher. I haven't had the time to do that yet.

Btw, in the study, I talk about some of the challenges with the measurement. I would highly recommend reading it. Many of the questions in this thread were addressed in my article. Max's articles on the topic are also highly recommended.

Mike--thanks for the links. Fascinating work.

I love this work! Plus I think it's awesome to believe that all of the big discoveries haven't been made yet.

If you take my best estimate that the real effect is 65% of what I originally reported, then that's +18 runs for Molina and -3 runs for Jaso, or about two wins difference per season for framing.

Thanks, Mike! That strikes me as a reasonable number - significant, consistent with the idea that catcher defense is more important than measured by other things, but not outrageous. I can totally buy this. Your work here is excellent, thanks!

Mike, thank you for replying. I re-read the study and noticed the section on k-rates. If I'm reading that correctly, it says that the delta on K rates is only +.002.

Is that right? Because that seems like such a tiny amount to be nearly nonsignificant. Roy Halladay's K%, for example, is listed on Fangraphs as being 23.6% in 2011. So we actually wouldn't even notice a difference in the Fangraphs stats if he went from Ryan Doumit to Jose Molina?

Or am I missing something?

I really do appreciate the tremendous amount of work you did.

Thanks for your response, Mike.

the catcher ERA effect from Sean Smith's study in the 2011 THT Annual

I was not aware of this. This is the type of thing that should inspire more discussion than any Chass thread. Is the BTF community a handful of years behind the statistical vanguard?

16. Voros McCracken, Human Shield Posted: November 29, 2011 at 01:06 AM (#4002700)
My personal opinion is that if pitch framing is really worth that much, then the automated strike zone needs to be in place by the start of the 2012 season. The strike zone is there so that the batter can at least have some chance at receiving a hittable baseball. It's not there so that catchers can play a game of "three-card monte" with umpires.

Yeah, that was my thought.

However, all this analysis is great work. This will continue to be important. Excellent thread.

Mike, thank you for replying. I re-read the study and noticed the section on k-rates. If I'm reading that correctly, it says that the delta on K rates is only +.002.

Is that right? Because that seems like such a tiny amount to be nearly nonsignificant. Roy Halladay's K%, for example, is listed on Fangraphs as being 23.6% in 2011. So we actually wouldn't even notice a difference in the Fangraphs stats if he went from Ryan Doumit to Jose Molina?

Or am I missing something?

K rates increased by 0.2% on average for the pitchers caught by the catchers in the top third and decreased by 0.5% for the pitchers caught by the catchers in the bottom third. (It looks like there is a typo in the table in that the trailing 5 got dropped from the -.005 in the final cell of the table. I'll get that fixed tomorrow.) Doumit and Molina are the far extremes, so the effect from switching between those two would be more than that. But for switching between typical catchers in bottom/top third, the typical effect on pitcher K/PA rate would be to change it by 0.7%. So for Halladay, to move him from 23.6% to 24.3%, for example, which would translate to about 6-7 extra strikeouts for Halladay over the course of a season.

As far as the effect on umpires goes, I have at least three thoughts.

1. The catchers who do best in the framing metric are not really the ones who are "fooling" the umpires. They are the ones who are getting out of the way and letting the umpire do his job. Or to put it another way, helping him see where the pitch went and not distracting him from focusing on that. The ones who do poorly are the ones who are actively doing things to distract and fool the umpire into thinking that pitches were balls.

2. An automated strike zone is not something that is easy to do without a lot of associated problems. It's not some panacea. It MAY be better than what we have now, but that is no guarantee. There are times when PITCHf/x has been off by four to six inches for a whole game, or worse. Setting the upper and lower limits of the strike zone is also not something that is necessarily done very accurately. These are potentially problems that could be overcome, but before we switch strike zones so radically and possibly ruin the game, it would be wise to work those problems out, I think.

3. A much bigger effect than catcher framing is that umpires call the zone relative to catcher target. If the pitcher hits the catcher target, he's much more likely to get a strike call than if he misses it by a foot or more. This has something like a 4-5x bigger effect on the strike zone than catcher framing. It's what gave Tom Glavine his career. It's what keeps Livan Hernandez around. It's what keeps Felix Hernandez from completely dominating the league. Mariano Rivera would probably be okay without it, but he takes advantage of it better than almost anyone but Glavine. If you're looking for a strike-zone effect on the game that's huge, it's zone relative to catcher target far moreso than catcher framing.

Isn't there a chance that umpires will eventually maybe feel "hoodwinked" by some catchers and not generous enough with other pitchers, based on the data, and letting that impact their calls the following season?

Which would be interesting, to the extent that the unfortunate 'adjustment' better aligned balls and strikes calls to reality.