Swihart v. Vazquez: The Value of Framing

Savin Hillbilly

loves the secret sauce
SoSH Member
Jul 10, 2007
18,783
The wrong side of the bridge....
I think you'd be better off to just stop at the bolded - its meaningless information.
Is a half-year actually that small a sample for framing, though? A half-year for a starting catcher is roughly 60 games * league average of 145 pitches per game = over 8,000 pitches. Assuming a swing rate in the 50% range that's 4,000-ish called pitches. That seems like a lot of events. I don't know if the fact that the events accumulate over a relatively short time frame matters that much. I'd be curious to hear what our statistical gurus have to say.
 
Last edited:

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
I was curious about this, so I checked it out.

David Price pitched 170.2 innings for the Rays in 2014 before being traded to the Tigers.

Ryan Hanigan caught three of those innings.

Hanigan started 0 of Price's 23 starts, and came into several of them after Price had left the game.

The only time Hanigan caught Price in an actual game was on 5/13/14, in the 8th and 9th inning - and on 6/15/14, in the eighth inning.
Good research. Still, they were teammates and if Swihart starts and Hanigan catches his rest days, pairing him with Price makes some sense if only because that's the one day when Swihart's better bat is less likely to be needed.
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
I think you'd be better off to just stop at the bolded - its meaningless information.
Umm... limited, but not at all meaningless. And it's not as if the data conflicts what is obvious to even casual observers, that Vazquez is much better defensively than Swihart. It just attempts to quantify that difference.

Besides 4,000 pitches for Vazquez and 6,000 pitches for Swihart aren't that small as far as data samples go. They're much bigger samples than, say, MLB plate appearances. (201 for CV and 309 for BS.)

Cervelli, the best framer last year at 27 RAA, caught 8,700 pitches. The Tigers' James McCann was the worst, -17 RA, and caught 7,900 pitches.

http://www.statcorner.com/CatcherReport.php

Think on that for a minute, a 44-run difference between Cervelli and J.McCann from framing only. That's about the same as the offensive run production difference between David Ortiz and Cameron Maybin.
 

chrisfont9

Member
SoSH Member
Of the two one assumes CV to have much more need of AAA development time. He has to prove he has recovered from the TJ surgery and that his defensive chops are still intact. More to the point he has to prove that his bat will play in MLB. 150 Abs in Pawtucket should do the trick - at least from the TJ recovery viewpoint. Quite frankly he could spend the entire year there developing the offensive part of his game. People forget how bad - as in JBJ bad - his 2014 actually was.

I expect CV will be in Pawtucket until June - then he comes up with Hannigan traded.
Well, except this is the guy the team penciled in as the starter a year ago, based on what he did in 2014 at the ML level and previously in the minors. Prior to his injury, the team clearly had designs on starting with Vazquez and slowing down Swihart's movement toward the majors. So some of us may think that Vazquez belongs in the minors moreso than Swihart, I have some doubt that the organization feels that way. I suppose they could change their mind if they see enough in Swihart's progression last year in all phases of the game, but wouldn't just assume this is the case.
 

PaulinMyrBch

Don't touch his dog food
Lifetime Member
SoSH Member
Dec 10, 2003
8,316
MYRTLE BEACH!!!!
It's a useless debate because so much depends on the recovery of Vazquez and where he is medically in spring training and opening day. But if he's 100% healthy and where he was the millisecond before the tendon popped, he's the starting catcher. Because between the kids, he's the only major league catcher (defensively) at this point in their careers, and pitching is important.

I'd much rather roll the dice with the elite defensive guy and hope the bat gets better, than the other way around. And the Swihart bat isn't elite yet. Pitching is important. When they both have between 500-750 at bats, we can get a better picture on the bats. Right now we know about the D and that is going to drive the decision IMO.

If healthy, I expect about a 60-40 split CV-BS
 

absintheofmalaise

too many flowers
Dope
SoSH Member
Mar 16, 2005
24,749
The gran facenda
Umm... limited, but not at all meaningless. And it's not as if the data conflicts what is obvious to even casual observers, that Vazquez is much better defensively than Swihart. It just attempts to quantify that difference.

Besides 4,000 pitches for Vazquez and 6,000 pitches for Swihart aren't that small as far as data samples go. They're much bigger samples than, say, MLB plate appearances. (201 for CV and 309 for BS.)

Cervelli, the best framer last year at 27 RAA, caught 8,700 pitches. The Tigers' James McCann was the worst, -17 RA, and caught 7,900 pitches.

http://www.statcorner.com/CatcherReport.php

Think on that for a minute, a 44-run difference between Cervelli and J.McCann from framing only. That's about the same as the offensive run production difference between David Ortiz and Cameron Maybin.
Those 4000 caught ball sample sizes are not the actual sample. The sample size is actually much smaller. You need to look at caught balls that are around the edges of the de facto strike zone, not at every caught ball.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
Those 4000 caught ball sample sizes are not the actual sample. The sample size is actually much smaller. You need to look at caught balls that are around the edges of the de facto strike zone, not at every caught ball.
... and that were called, not swung at.
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
Those 4000 caught ball sample sizes are not the actual sample. The sample size is actually much smaller. You need to look at caught balls that are around the edges of the de facto strike zone, not at every caught ball.
Good point. Borderline calls are about 10% of all pitches, so one season's worth is about 800 pitches, a larger sample size than a full season of plate appearances.
 
Last edited:

Zincman

Member
SoSH Member
Jul 31, 2006
444
New London
The difference in framing skills between Vazquez and Swihart is not likely to be static. That is to say that Swihart has greater room for growth in this area whereas Vazquez is not likely to become a great deal better, albeit he is already very good. Baseball Prospectus recently has a nice piece on radical improvement in framing and given Swihart's makeup and athleticism, this is an area in which he is likely to improve.

http://www.baseballprospectus.com/article.php?articleid=28021
 

PrometheusWakefield

Member
SoSH Member
May 25, 2009
10,507
Boston, MA
The difference in framing skills between Vazquez and Swihart is not likely to be static. That is to say that Swihart has greater room for growth in this area whereas Vazquez is not likely to become a great deal better, albeit he is already very good. Baseball Prospectus recently has a nice piece on radical improvement in framing and given Swihart's makeup and athleticism, this is an area in which he is likely to improve.

http://www.baseballprospectus.com/article.php?articleid=28021
Well, yes, it's easier to improve at something if you suck at it. But he could improve considerably at it and still not be good at it. And this skill turns out to be hugely important to the value of the defensive position he is playing.

And I want a catcher who is not only good at it, I want a catcher who is amazing at it. Because in addition to the valuation that iayork talks about in the .com article linked above (in which he values a strike as opposed to a ball at .13, which means that catchers who get +2 strikes per game are worth a quarter run per game and thus 30+ runs per season), there's the cumulative secondary impacts on the whole pitching staff.

I agree we don't need to make a decision now but I'd like to see Swihart try out a number of positions this spring to maximize our ability to keep his bat in the lineup. Looks to me like we might have some needs at third base.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
Baseball Prospectus recently has a nice piece on radical improvement in framing and given Swihart's makeup and athleticism, this is an area in which he is likely to improve.

http://www.baseballprospectus.com/article.php?articleid=28021
I hadn't seen that article, thanks for pointing it out. The other example I know of where a catcher abruptly and dramatically improved his framing is Saltalamacchia, who went from a moderately bad framer pre-2011, to quite a good one in 2011, gaining around 1.5 strikes per game from his previous level. He's not the most encouraging example, though, since he promptly went into a framing tailspin, dropping year after year until he was legendarily terrible in 2014 and 2015.

I agree that Swihart has a reasonable chance to improve, given that he's young, still learning, athletic, and working for a team that clearly values framing a lot. And it's important to point out that he's already not bad -- he's basically league-average, or close to it, already.

One other point: I don't know if this is a real trend, but I have the impression that the most effective framers are having slightly less impact year after year. Maybe it's just because Molina was such an outlier early on, but I suspect that umpires are gradually figuring out framers. Umpire ball and strike calls have improved quite dramatically since they started to be graded via PITCHf/x (see the chart below showing the percent of out-of-zone calls since PITCHf/x was introduced). If that's a trend that continues, which of course it may not, the value of an elite pitch framer may drop along with it.
 
Last edited:

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
Good point. Borderline calls are about 10% of all pitches, so one season's worth is about 800 pitches, a larger sample size than a full season of plate appearances.
How many pitches we are talking about has nothing to do with plate plate appearances when talking about stabilization. A batter's strike out rate stabilizes after a different number of PA's than their batting average, for example. 800 might be enough for pitch framing to stabalize, but that number compared to PA's tells us absolutely nothing.
 

Zincman

Member
SoSH Member
Jul 31, 2006
444
New London
Well, yes, it's easier to improve at something if you suck at it. But he could improve considerably at it and still not be good at it. And this skill turns out to be hugely important to the value of the defensive position he is playing.
I'm just an old-timer lurking about these pages but I can still recognize a"straw man argument" even when it comes from such a distinguished source as Prometheus. According to Dr York and others, Swihart doesn't suck at it, he was basically about league average and, in fact, improved a bit as the season went on according to the BP article I posted.

Won't the question eventually be does Swihart's projected offensive prowess outweigh the difference between his and Vazquez' defense. I believe it will because I believe Swihart will improve greatly and that the significance of framing will continue to dissipate. FWIW we need to stop talking about moving Swihart. The bat is not likely to play at first base and no small sample size will prove if he's a MLB 3b long term
 

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
Won't the question eventually be does Swihart's projected offensive prowess outweigh the difference between his and Vazquez' defense.
It may already. 188 points of OPS is a lot to overcome. Of course, it's possible that Swihart's second half is not representative of where he is right now as a hitter, but at his age it's not a bad bet to make. OPS supposedly has a .95 correlation to run scoring.

I believe it will because I believe Swihart will improve greatly and that the significance of framing will continue to dissipate. FWIW we need to stop talking about moving Swihart. The bat is not likely to play at first base and no small sample size will prove if he's a MLB 3b long term
If his second half is representative of where he is as a hitter, his bat would certainly play at 1st. His 2nd half .805 OPS would have placed him 26th in the majors which is an every day starter, though on the back end of that list. If, at 22 years old he was already at that level, there's a pretty good chance he's going to hit well enough to be an average 1st baseman at the plate over the long haul. Now add in defense, and we're probably talking about a comfortably above average player, as his athleticism would probably translate to solid range, his arm would probably be better than most at the position and as a catcher, I imagine he would scoop throws really well.

He is far less valuable as an above average first baseman than he would be as an above average, or well above average catcher, but I think all indications are that his floor at first base would be pretty high. That said, I'd rather keep him behind the plate, and think a long term tandem of Swihart and Vazquez is the best option. If Vazquez is recovered enough this year and forces the issue, deal Hanigan at the deadline. Having great depth at catcher is nothing but a good thing.
 

grimshaw

Member
SoSH Member
May 16, 2007
4,299
Portland
Getting him the at bats wherever you can has to be close to equal importance at this stage of his development since there is nothing that needs drastic improvement defensively. Obviously the pitches in the dirt are somewhat of an issue, but you can only take so many balls in the gut during practice. The trial by fire portion is pretty much over and he is adequate but with plenty of things to improve on. Having Vazquez in the dugout every day is going to complement that even more.

He is young enough where it isn't an old dog/new tricks situation (unless he doesn't take to it mentally), so it isn't the end of the world if he's pretty meh elsewhere. He needs the offensive reps so they can determine his ceiling more quickly and figure out where his bat can play, as much if not more than refining his pitch framing and receiving abilities.
 

Rovin Romine

Johnny Rico
Lifetime Member
SoSH Member
Jul 14, 2005
25,906
Miami (oh, Miami!)
It may already. 188 points of OPS is a lot to overcome. Of course, it's possible that Swihart's second half is not representative of where he is right now as a hitter, but at his age it's not a bad bet to make. OPS supposedly has a .95 correlation to run scoring.
I'm not sure how to ask this question. Does worth of the OPS gap change if we're talking about the bottom guy being average or well-below average with the bat? Meaning, if Vaz is adequate offensively (i.e., there's no extra negative synergistic worth in his being an "easy out" in the lineup), does Swihart have to be better by the same amount he's currently projected to be, or does he have to be much much better? On the other hand, if Vaz is very bad offensively, is there an extra hidden synergistic penalty for playing that kind of a bat? Meaning a more modest OPS gap between the players would be more significant?
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,693
Halifax, Nova Scotia , Canada
I think the problem people have with pitch framing is not the frequency of stolen strikes (or stolen balls) - which is pretty demonstrable and definitely seems to be a skill - but the actual magnitude of the effect in terms of Runs Saved. The link cited above states a run value of .13 per stolen strike. Sadly it doesn't really state where that number comes from - nor do any of the subsequently linked articles from what I could see. I'm not saying I disagree with it - I just don't trust it as an actual fact.

In 2008, Dan Turkenkopf estimated that each switched call is worth about 0.13 runs on average. If that’s about right, then having elite framers like Molina or Lucroy on staff, instead of average catchers, adds between a quarter and a third of a run per game over an average catcher, or the equivalent of about an extra 20-35 runs of value over an entire season from framing alone.
and

I also tweaked my run value of a "mistake" thanks to a comment from Iblemetrician that pointed out that missed pitches are probably more likely to occur at some counts rather than others. The value I'm using for these results is .133 runs per mistake, which is substantially lower than the original value of .161.
That simply doesn't pass the smell test. I suppose it's somewhat like DIPS theory - which initially seemed crazy - then it didn't - and now is settling in somewhere in between.

CV was (is?) an outstanding defensive catcher - what that means in terms of value - vs. easily measured offensive value is an open question IMO. He may contribute 2 WAR on framing alone - then again he may not. And as has been pointed out by many, it seems to be a diminishing phenomenon. Whereas as Swihart's skills are much easier to quantify.

And ultimately - If Pitch Framing is as advertised why doesn't the market reflect this?

/ENDRANT
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
I think the problem people have with pitch framing is not the frequency of stolen strikes (or stolen balls) - which is pretty demonstrable and definitely seems to be a skill - but the actual magnitude of the effect in terms of Runs Saved. The link cited above states a run value of .13 per stolen strike. Sadly it doesn't really state where that number comes from - nor do any of the subsequently linked articles from what I could see. I'm not saying I disagree with it - I just don't trust it as an actual fact.
For someone with the time and inclination, it shouldn't be hard to figure out using linear weights. The difference between an 0-1 and 1-0 count is about .200 OPS. The difference between "strike three" and "ball four" is huge, about 2/3rds of a run, on average, IIRC. (A walk is worth .35 runs. An out -.30 runs. Or something close to that.) Do the math for all the counts in between 0-0 and 3-2 for every on-base state and you're close to knowing the value of a "stolen" strike or "lost" ball.


And ultimately - If Pitch Framing is as advertised why doesn't the market reflect this?
Seriously? If OBP is as advertised why didn't the market reflect it for over 100 years? The entire point of sabermetrics and baseball data analysis is to find "hidden value" unrecognized by the market. Framing has been valued by teams, coaches and pitchers for a very long time. What's new is that technology, PITCHf/x, makes it possible to more accurately measure the value of framing and determine its value, which appears to be a lot higher than previously thought.
 
Aug 22, 2014
61
pitch framing is nice but there's a chance concentrating too much on it might make us underrate even harder to pin down skills such as "gamecalling".
 

ALiveH

Member
SoSH Member
Apr 23, 2010
1,142
To me, the more interesting question than Vasquez OR Swihart is how to maximize the value of Vasquez AND Swihart. In 2016, Vasquez needs some AAA reps to re-prove himself coming off injury. And we have a very solid veteran, Hanigan that there's no rush to displace.

For 2017+, Hanigan is probably gone and we have to carry 2 catcher's on the roster. Catcher is a brutal position, so let's figure out a way to keep them both fresh and maximize Vasquez's D versus Swihart's O. For example, I could see Vasquez playing 2/3 and Swihart playing 1/3 of the games at catcher. That way we get the benefit of Vasquez's D over a whole season and he stays relatively fresh & rested. Meanwhile, Swihart can be a part-time sub at 1B/3B/DH. Then, there is the fact that catcher is a notoriously injury-prone position, so there is a lot of value to not forcing a catcher to play through minor injuries and/or having a very solid everyday catcher available if V or S go down for an extended period.
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,693
Halifax, Nova Scotia , Canada
For someone with the time and inclination, it shouldn't be hard to figure out using linear weights. The difference between an 0-1 and 1-0 count is about .200 OPS. The difference between "strike three" and "ball four" is huge, about 2/3rds of a run, on average, IIRC. (A walk is worth .35 runs. An out -.30 runs. Or something close to that.) Do the math for all the counts in between 0-0 and 3-2 for every on-base state and you're close to knowing the value of a "stolen" strike or "lost" ball.




Seriously? If OBP is as advertised why didn't the market reflect it for over 100 years? The entire point of sabermetrics and baseball data analysis is to find "hidden value" unrecognized by the market. Framing has been valued by teams, coaches and pitchers for a very long time. What's new is that technology, PITCHf/x, makes it possible to more accurately measure the value of framing and determine its value, which appears to be a lot higher than previously thought.
Seriously? OBP was NOT generally recognized as a huge component of offensive value (Branch Rickey notwithstanding) until the early 2000s - or at least until Bill James started to have some influence on traditional baseball thinking.

At this point most MLB franchises are pretty savvy to modern sabremetric methods. Given that assumption (I suppose there's still a few luddite clubs out there) wouldn't that drive up the market for the Jose Molinas and Ryan Hannigans of the world if the magnitude of the effect is universally recognized?
 
Jun 27, 2006
66
I'm not sure if there is any way to measure it, but would the confidence the pitcher has in his catcher greatly effect his confidence in throwing borderline pitches? Meaning he would be less likely to give in in certain situations. Thus greatly effecting the ability for the batter to make good contact. I would think most pitchers if given the choice would rather pitch to vaquez.
 

geoduck no quahog

not particularly consistent
Lifetime Member
SoSH Member
Nov 8, 2002
13,024
Seattle, WA
I'm surprised that Ian's example charts seem to show more variation east-west than north-south. Maybe that's a function of the particular umpire chosen, or a misreading on my part.

Logic says that an umpire should see the east-west crossing of the plate much better than the north-south. After all, he's looking down on the pitch and sees the ball and the plate in the same view. For height, he needs to rely on where the catcher catches the ball (obviously because he's not looking from the side) and assume the track of the ball when it's at the front or rear of the plate.

My point? In theory, pitch framing should only matter for height because umpires should almost always get the horizontal correctly. And yet, that doesn't seem to be the case. It's tough for me to read the dynamic charts so maybe someone could help me out.

My theory? Catchers should be able to steal high and low strikes much easier than any others. It's not fair, but unless the umpire's wearing an earpiece that blips when a pitch is too high or low, it's the way things are.
 

shaggydog2000

Member
SoSH Member
Apr 5, 2007
11,922
I'm surprised that Ian's example charts seem to show more variation east-west than north-south. Maybe that's a function of the particular umpire chosen, or a misreading on my part.

Logic says that an umpire should see the east-west crossing of the plate much better than the north-south. After all, he's looking down on the pitch and sees the ball and the plate in the same view. For height, he needs to rely on where the catcher catches the ball (obviously because he's not looking from the side) and assume the track of the ball when it's at the front or rear of the plate.

My point? In theory, pitch framing should only matter for height because umpires should almost always get the horizontal correctly. And yet, that doesn't seem to be the case. It's tough for me to read the dynamic charts so maybe someone could help me out.

My theory? Catchers should be able to steal high and low strikes much easier than any others. It's not fair, but unless the umpire's wearing an earpiece that blips when a pitch is too high or low, it's the way things are.
Umpires are blocked by catchers. Depending on where they set up, they can be blocked on the inside or outside edge of the strike zone.
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
Umpires are blocked by catchers. Depending on where they set up, they can be blocked on the inside or outside edge of the strike zone.
Umpires almost always set up between the catcher and batter. (See this fangraphs article. It's about CF camera views, but you can see the umps in relation to catchers. http://www.fangraphs.com/blogs/ranking-baseballs-center-field-camera-shots-2015-update/ ) So you'd think they'd see inside pitches clearly and possibly have trouble with outside pitches, especially low and outside pitches.
 

PrometheusWakefield

Member
SoSH Member
May 25, 2009
10,507
Boston, MA
I hadn't seen that article, thanks for pointing it out. The other example I know of where a catcher abruptly and dramatically improved his framing is Saltalamacchia, who went from a moderately bad framer pre-2011, to quite a good one in 2011, gaining around 1.5 strikes per game from his previous level. He's not the most encouraging example, though, since he promptly went into a framing tailspin, dropping year after year until he was legendarily terrible in 2014 and 2015.

I agree that Swihart has a reasonable chance to improve, given that he's young, still learning, athletic, and working for a team that clearly values framing a lot. And it's important to point out that he's already not bad -- he's basically league-average, or close to it, already.

One other point: I don't know if this is a real trend, but I have the impression that the most effective framers are having slightly less impact year after year. Maybe it's just because Molina was such an outlier early on, but I suspect that umpires are gradually figuring out framers. Umpire ball and strike calls have improved quite dramatically since they started to be graded via PITCHf/x (see the chart below showing the percent of out-of-zone calls since PITCHf/x was introduced). If that's a trend that continues, which of course it may not, the value of an elite pitch framer may drop along with it.
The other thing that strikes me as complicated about pitch framing as a statistic is that like other defensive statistics it's really only a relative value - we aren't even really tracking the objective measurement of a catcher from year to year, we're only tracking how good the catcher is compared to the rest of the league. And since the secret is out that this is really important the rest of the league is almost certainly getting a lot better at this very quickly, not only because teams are now going to be heavily training their guys in how to do this better but also because teams are giving real playing time to guys like Rene Rivera and his 33 RC+.

Swihart doesn't just have to improve his game to be average or above average on these numbers; he has to improve his game relative to the 31 other catchers in MLB who are also going to be working hard to improve their performance on this metric or who are about to take over playing time from catchers who can't keep up. And yeah, Swihart is a good athlete but a lot of these other guys are good athletes too, and it ain't a magic trick - good framing is really just an extension doing the things that make you a good receiver in general.
 

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
So... your argument is that this skill isn't so difficult to learn that the other 30 teams would have much trouble catching up, but that somehow means Swihart isn't likely to close the gap with Vazquez going forward? o_O
 

Rasputin

Will outlive SeanBerry
Lifetime Member
SoSH Member
Oct 4, 2001
29,594
Not here
So... your argument is that this skill isn't so difficult to learn that the other 30 teams would have much trouble catching up, but that somehow means Swihart isn't likely to close the gap with Vazquez going forward? o_O
It's a perfectly reasonable argument, and you know it. Other teams aren't going to improve by having their players improve, but by playing different players, which means that if Swihart wants to improve relative to the league, he's got to improve a whole heck of a lot.
 

absintheofmalaise

too many flowers
Dope
SoSH Member
Mar 16, 2005
24,749
The gran facenda
The other thing that strikes me as complicated about pitch framing as a statistic is that like other defensive statistics it's really only a relative value - we aren't even really tracking the objective measurement of a catcher from year to year, we're only tracking how good the catcher is compared to the rest of the league. And since the secret is out that this is really important the rest of the league is almost certainly getting a lot better at this very quickly, not only because teams are now going to be heavily training their guys in how to do this better but also because teams are giving real playing time to guys like Rene Rivera and his 33 RC+.

Swihart doesn't just have to improve his game to be average or above average on these numbers; he has to improve his game relative to the 31 other catchers in MLB who are also going to be working hard to improve their performance on this metric or who are about to take over playing time from catchers who can't keep up. And yeah, Swihart is a good athlete but a lot of these other guys are good athletes too, and it ain't a magic trick - good framing is really just an extension doing the things that make you a good receiver in general.
How would you measure framing except against other catchers in the league?
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
How would you measure framing except against other catchers in the league?
One can measure "raw" framing, i.e. how many balls a catcher turns into strikes and how many strikes he turns into balls. In fact, that is measured:

http://www.statcorner.com/CatcherReport.php

But when placing a value on framing ability, it's useful to state the effect in terms of "runs above average," a stat similar in concept to WAR.

Why use a relative stat instead of an absolute stat? Dunno. I suppose it's more useful to say "Buster Posey saves 30 runs more/120 games than James McCann" than it is to say "Buster Posey saves 'X' runs with his framing." Framing effects are such a new concept that for most people that "'X' runs/120 games" would be kind of meaningless without putting that number in context vis-a-vis other catchers.

Edit: In other words, while everyone knows what a .402 OBP and .590 SLG means, few people have any idea what 13.7 zBall% and 10.7 oStrike% means, so those get translated into 26.7 runs above average.
 
Last edited:

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
It's a perfectly reasonable argument, and you know it. Other teams aren't going to improve by having their players improve, but by playing different players, which means that if Swihart wants to improve relative to the league, he's got to improve a whole heck of a lot.
No, it's not reasonable since playing great framers who have terrible bats means that the value gap between them and Swihart is already minimized, if not outpaced by what Blake can do with the stick. He also mentioned that other teams will be teaching their catchers to "do this better" which suggests exactly what I found puzzling. If other teams will be teaching their guys to better frame pitches, it's apparently not that hard and Swihart, who has proven to be very capable of adapting and learning new skills, should be able to improve going forward at at least the same pace his contemporaries are improving. If it is really hard and can't be taught that easily, then the rest of the league won't be catching up since those great framers who can't hit still won't be able to hit, and the good bats who can't frame won't be closing the gap with Swihart with the glove.

His position is nonsensical.
 

The X Man Cometh

New Member
Dec 13, 2013
390
I don't think anyone deprecates game-calling and pitcher-managing, but no one has managed to put believable numbers on those yet, so it's hard to say how much they're worth.
I definitely see people on internet forums using framing runs as a proxy for defensive aptitude, simply because its something they can measure and latch onto. I won't name names :)

Regarding the thread itself. Something which is worth mentioning is Blake's platoon splits.

As RHH vs. LHP (91 PA)
8.8% walk rate
20.9% strikeout rate
.283 BABIP
65 wRC+

As LHH vs. RHP (218 PA)
4.6% walk rate
26.6% strikeout rate
.390 BABIP
102 wRC+

So on one hand, you have a reasonable expectation that his BABIP comes down. But isn't there's also the possibility that he narrows his split to at least somewhat mitigate that? His walk and K rates on the right side are better than the left.
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
No, it's not reasonable since playing great framers who have terrible bats means...
Seems like a straw man. No one is talking about playing great framers with terrible bats over lousy framers with terrific bats.

The idea of forgoing offensive value for defensive value isn't exactly a new. Ozzie Smith's career OPS was .666. Mark Belanger's was .580.
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,693
Halifax, Nova Scotia , Canada
Seems like a straw man. No one is talking about playing great framers with terrible bats over lousy framers with terrific bats.

The idea of forgoing offensive value for defensive value isn't exactly a new. Ozzie Smith's career OPS was .666. Mark Belanger's was .580.
Well .. In this case we are talking about playing great framers with terrible bats over average framers with terrific bats.

Another factor favouring Swihart .. To reach his potential .. Which seems to be that of an elite offensive catcher .. He needs as many ABs as possible. Playing three times a week is a waste.

Regardless of this debate I suspect we will see what the RedSox think come June or thereabouts.
 

Zincman

Member
SoSH Member
Jul 31, 2006
444
New London
I hope that everyone understands that framing is not some new tactic recently discovered by the sabermetric community. We have been teaching framing skills for over 30 years and good coaches at every level have been training catchers to frame properly for years. Frankly, MLB didn't pay a great deal of attention to it and in my years with the scouting community it was occasionally mentioned but rarely emphasized. My guess is that MLB had no real way of measuring it and the result was that they probably underestimated the impact, perhaps vastly, until recently.

Other teams aren't going to improve by having their players improve, but by playing different players, which means that if Swihart wants to improve relative to the league, he's got to improve a whole heck of a lot.
I don't think this is true, since improvement of framing skills is very teachable. Check out the Derek Norris story I posted in my original post. It is likely that there will always be a bell curve of framers, since some will have more skill than others, but the marginal difference will continue to dissipate as framing skills become more and more emphasized and particularly as umpires enter the equation.

Moreover, Prometheus continues to cling to the notion that Swihart sucks at framing and the implication continues that he has a ton of ground to make up on the rest of the league. Metrics dispute that as Swihart was about league average and continued to improve. The difference is going to get more and more narrow with passing time.
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
Well .. In this case we are talking about playing great framers with terrible bats over average framers with terrific bats.
Only if you think Swihart's .312 wOBA is terrific and Vazquez's .277 wOBA is terrible. (Average is .304.)

Anyway, I am not talking about playing CV over Swihart. I like them both and think the offensive value vs. defensive value questions are interesting.



Another factor favouring Swihart .. To reach his potential .. Which seems to be that of an elite offensive catcher .. He needs as many ABs as possible. Playing three times a week is a waste.
I wouldn't say "a waste," but good point. But why is Swihart's offensive value inextricably linked to being a catcher any more than, say, Betts's offensive value was chained to second base?
 

Rasputin

Will outlive SeanBerry
Lifetime Member
SoSH Member
Oct 4, 2001
29,594
Not here
No, it's not reasonable since playing great framers who have terrible bats means that the value gap between them and Swihart is already minimized, if not outpaced by what Blake can do with the stick. He also mentioned that other teams will be teaching their catchers to "do this better" which suggests exactly what I found puzzling. If other teams will be teaching their guys to better frame pitches, it's apparently not that hard and Swihart, who has proven to be very capable of adapting and learning new skills, should be able to improve going forward at at least the same pace his contemporaries are improving. If it is really hard and can't be taught that easily, then the rest of the league won't be catching up since those great framers who can't hit still won't be able to hit, and the good bats who can't frame won't be closing the gap with Swihart with the glove.

His position is nonsensical.
I was only commenting on the specifics of Swihart improving v other teams. I probably shouldn't have waded in without reading more of the thread, but frankly I find the whole argument rather silly. We have them both. We're going to play them both, and with any luck, we're going to keep them both until they get too expensive.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
I think the problem people have with pitch framing is not the frequency of stolen strikes (or stolen balls) - which is pretty demonstrable and definitely seems to be a skill - but the actual magnitude of the effect in terms of Runs Saved. The link cited above states a run value of .13 per stolen strike. Sadly it doesn't really state where that number comes from - nor do any of the subsequently linked articles from what I could see. I'm not saying I disagree with it - I just don't trust it as an actual fact.
A more recent, and nicely detailed, explanation of the value of a strike in terms of runs is Framing and Blocking Pitches: A Regressed, Probabilistic Model, by Harry Pavlidis and Dan Brooks. They offer a much more complex model, but their simplistic value is ~0.14 runs per pitch. The fact that it's so close to the 2008 value that was calculated quite differently is pretty encouraging.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
I'm surprised that Ian's example charts seem to show more variation east-west than north-south. Maybe that's a function of the particular umpire chosen, or a misreading on my part.
Most catchers seem to frame best either at the top or the bottom of the zone, though rarely both (Jose Molina and Vaquez being among the rare exceptions who get strikes all around the zone). Here's Brian McCann, for example, getting lots of extra low strikes while losing some at the top:

Salty's 2011 is the most dramatic example of north/south framing I've seen; he pulled a huge number of extra strikes from the top of the zone, while losing a few at the bottom. In subsequent years, he continued to lose strikes from the bottom while not getting enough at the top to compensate.
(If there's a simple way to change the size of the images in the new system, someone please let me know.)
 

Jordu

Member
SoSH Member
Apr 30, 2003
9,157
Brookline
I hope that everyone understands that framing is not some new tactic recently discovered by the sabermetric community. We have been teaching framing skills for over 30 years and good coaches at every level have been training catchers to frame properly for years. Frankly, MLB didn't pay a great deal of attention to it and in my years with the scouting community it was occasionally mentioned but rarely emphasized...
How did you scout a catcher for game-calling?
 

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
Seems like a straw man. No one is talking about playing great framers with terrible bats over lousy framers with terrific bats.

The idea of forgoing offensive value for defensive value isn't exactly a new. Ozzie Smith's career OPS was .666. Mark Belanger's was .580.
Except this guy.

Well, yes, it's easier to improve at something if you suck at it. But he could improve considerably at it and still not be good at it. And this skill turns out to be hugely important to the value of the defensive position he is playing.

And I want a catcher who is not only good at it, I want a catcher who is amazing at it. Because in addition to the valuation that iayork talks about in the .com article linked above (in which he values a strike as opposed to a ball at .13, which means that catchers who get +2 strikes per game are worth a quarter run per game and thus 30+ runs per season), there's the cumulative secondary impacts on the whole pitching staff.

I agree we don't need to make a decision now but I'd like to see Swihart try out a number of positions this spring to maximize our ability to keep his bat in the lineup. Looks to me like we might have some needs at third base.
The other thing that strikes me as complicated about pitch framing as a statistic is that like other defensive statistics it's really only a relative value - we aren't even really tracking the objective measurement of a catcher from year to year, we're only tracking how good the catcher is compared to the rest of the league. And since the secret is out that this is really important the rest of the league is almost certainly getting a lot better at this very quickly, not only because teams are now going to be heavily training their guys in how to do this better but also because teams are giving real playing time to guys like Rene Rivera and his 33 RC+.

Swihart doesn't just have to improve his game to be average or above average on these numbers; he has to improve his game relative to the 31 other catchers in MLB who are also going to be working hard to improve their performance on this metric or who are about to take over playing time from catchers who can't keep up. And yeah, Swihart is a good athlete but a lot of these other guys are good athletes too, and it ain't a magic trick - good framing is really just an extension doing the things that make you a good receiver in general.
The Rivera's of the world getting more playing time does not make Swihart less valuable as a catcher because the Rivera's of the world can't hold a candle to Swihart at the plate. Pitch framing is one component of the overall package and overall value is what matters. So, again, either this skill isn't too hard to pick up and Swihart should, like the rest of the league, improve going forward, or it is hard, and players like Rivera are going to have to make up a huge gap offensively with their framing skills.

Of course, since PW is starting from the position that Swihart is a bad framer, he clearly doesn't know what he's talking about in the first place, so it shouldn't be shocking that his follow up post was an exercise in contradiction.
 

Super Nomario

Member
SoSH Member
Nov 5, 2000
14,062
Mansfield MA
In general, the numbers are surprisingly consistent year to year, but there are definitely error bars. I don't know of anyone who has tried to calculate the variance year to year, but my gut feeling is that 3/4 of a strike per game is in the right ballpark. That is, if Vaz got an extra 2 strikes per game in 2014, it's very realistic to expect him to be somewhere between 1.25 and 2.75 strikes per game in 2016. (Two strikes per game translates to about 30 runs per season.)

I wrote a bunch of articles about framing on the .com over the season. I think this one (which covers Vazquez's framing) links to most of them, if you want to see how I generated the numbers. It's a slightly different approach than StatCorner or Baseball Prospectus, but everyone ends up with pretty similar numbers, which is reassuring that it's a real thing.
Is the 1.25 to 2.75 range your estimate for Vazquez's "true" talent level or where you think he's likely to be in 2016 (assuming health)? If the former, wouldn't normal regression patterns tell us he's more likely to be at the bottom of that range (since presumably 1.5-strikes-per-game guys are much more common than 2.5-strikes-per-game guys)?
 

iayork

Member
SoSH Member
Apr 6, 2006
639
Is the 1.25 to 2.75 range your estimate for Vazquez's "true" talent level or where you think he's likely to be in 2016 (assuming health)? If the former, wouldn't normal regression patterns tell us he's more likely to be at the bottom of that range (since presumably 1.5-strikes-per-game guys are much more common than 2.5-strikes-per-game guys)?
Yeah, maybe. Looking over catchers over the past few years, there's probably some regression going on, but there are also long-term trends (probably aging) that are hard to separate out.
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,693
Halifax, Nova Scotia , Canada
A more recent, and nicely detailed, explanation of the value of a strike in terms of runs is Framing and Blocking Pitches: A Regressed, Probabilistic Model, by Harry Pavlidis and Dan Brooks. They offer a much more complex model, but their simplistic value is ~0.14 runs per pitch. The fact that it's so close to the 2008 value that was calculated quite differently is pretty encouraging.
You see .. This is the essence of the problem. My brain tells me you are probably correct. That the magnitude is more or less real. The difficulty is reconciling that with decades and decades of watching baseball -often with an analytical approach. How could a couple of strikes a game be the difference between a 3.00 and a 3.25 era ..

It was very much like DIPS theory. I found it absurd when it told me there was no difference between Pedro Martinez and Pete Shourek on a ball in play.
 

Plympton91

bubble burster
SoSH Member
Oct 19, 2008
12,408
Umm... limited, but not at all meaningless. And it's not as if the data conflicts what is obvious to even casual observers, that Vazquez is much better defensively than Swihart. It just attempts to quantify that difference.

Besides 4,000 pitches for Vazquez and 6,000 pitches for Swihart aren't that small as far as data samples go. They're much bigger samples than, say, MLB plate appearances. (201 for CV and 309 for BS.)

Cervelli, the best framer last year at 27 RAA, caught 8,700 pitches. The Tigers' James McCann was the worst, -17 RA, and caught 7,900 pitches.

http://www.statcorner.com/CatcherReport.php

Think on that for a minute, a 44-run difference between Cervelli and J.McCann from framing only. That's about the same as the offensive run production difference between David Ortiz and Cameron Maybin.
Right. Which is as good a reason to call bullshit on the statistic as it is to go sign Francisco Cervelli for $20 million per season.
 

Plympton91

bubble burster
SoSH Member
Oct 19, 2008
12,408
I hadn't seen that article, thanks for pointing it out. The other example I know of where a catcher abruptly and dramatically improved his framing is Saltalamacchia, who went from a moderately bad framer pre-2011, to quite a good one in 2011, gaining around 1.5 strikes per game from his previous level. He's not the most encouraging example, though, since he promptly went into a framing tailspin, dropping year after year until he was legendarily terrible in 2014 and 2015.

]
 
Last edited:

Plympton91

bubble burster
SoSH Member
Oct 19, 2008
12,408
Or maybe the technology for measuring pitch framing skills is highly variable, or pitch framing skills are heavily dependent on the stuff and command of the pitching staff (hell of a lot easier to frame late career Curt Schilling than early career Randy Johnson) or the skill itself isn't stable.

If pitch framing were as important as the spreadsheets are telling us, Jose Molina would have gotten multiyear contact offers at $15 million a year. He didn't. I'm guessing the spreadsheets are getting out wrong and not 30 front offices full of people who are paid to figure out what matters and what doesn't and who is and isn't good at what matters.

Another way to check the statistic against reality is to see how much offense teams are willing to give up for a great framing catcher. If they're increasingly willing to bench David Ortiz for Cameron Maybin or if catchers who hit like Kevin Cash and Sandy Leon but frame like Christian Vazquez start getting 100 starts a year for pennant winners, then I'd start giving the stat more credence.
 

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
Just because front offices value something, that doesn't mean the market place will reflect that. If no one starts paying significantly more for great pitch framers, there won't be a reason to start paying significantly more for great pitch framers. Especially if that skill isn't coming attached to a solid bat. Why drive up the market if you don't have to? In cases where a great pitch framer also has a good stick (Russell Martin, for example) it's hard to see where that contract splits between his various skill sets, which muddies the waters a bit.
 

dhappy42

Straw Man
Oct 27, 2013
15,826
Michigan
Right. Which is as good a reason to call bullshit on the statistic as it is to go sign Francisco Cervelli for $20 million per season.
Go ahead and call bullshit on framing and RAA, but if you want to be taken seriously, you should probably back it up with something other than your feelings.

I'm no fan of Cervelli (who is a free agent in 2017), but fangraphs shows him at 3.8 WAR last year and projects him at 1.9 WAR this year. (That's 0.8 WAR more than Swihart.) How much is 2 WAR worth? The Red Sox just traded four valuable prospects for a 1.2-2.2 WAR reliever who they'll pay $12.25M this year.

Incidentally, Cervelli's age 23 batting average was .298, Swihart's was .274 (.303 after in the second half of the season, though.)