MASN: A Primer on Pitch Framing

[Can_of_corn]

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

That looks to be straight bean counting. This is how many balls the computer says were strikes, this is how many balls the ump called strikes. Useful data but it doesn't show if anything the Catcher did caused it.

[weams]

That looks to be straight bean counting. This is how many balls the computer says were strikes, this is how many balls the ump called strikes. Useful data but it doesn't show if anything the Catcher did caused it.

Well, to do otherwise would be like attempting to discern the scoring technique of the East German judge. No?

[Can_of_corn]

Well, to do otherwise would be like attempting to discern the scoring technique of the East German judge. No?

The problem is we already know, for a fact, that other factors influence if a pitch is called a ball or a strike. To ignore that information and give sole credit to the catcher is wrong.

[weams]

The problem is we already know, for a fact, that other factors influence if a pitch is called a ball or a strike. To ignore that information and give sole credit to the catcher is wrong.

It's why I think the endeavor itself is wrong. Totes.

[srock]

The problem is we already know, for a fact, that other factors influence if a pitch is called a ball or a strike. To ignore that information and give sole credit to the catcher is wrong.

This is great raw data but probably need lots more. The next step is to apply meaning and that is the hard part.

There may be no meaning at all which is fine, one less variable to worry about.

Nobody can realy know if at this point if there is anything useful in the data

Sent from my iPhone using Tapatalk

[El Gordo]

Anybody ever talked to umpires about pitch framing? They might be reluctant to acknowledge it, but they probably know more about it's impact on their game.

[Can_of_corn]

Anybody ever talked to umpires about pitch framing? They might be reluctant to acknowledge it, but they probably know more about it's impact on their game.

Good question. Would probably be a better idea to ask a retired ump about it, more likely to get a straight answer.

[weams]

Good question. Would probably be a better idea to ask a retired ump about it, more likely to get a straight answer.

Do they get a union pension?

[Can_of_corn]

Do they get a union pension?

I am sure they do. I am also sure it wouldn't be at risk for discussing pitch framing.

[srock]

Pitch framing is intended to influence (trick) the umpire from calling a ball a strike and making sure a strike is called a strike. No umpire, active or retired, will admit to getting influenced by it.

We all know it works. Otherwise, why do umps call this pitch a ball. Catcher puts in inside target. Pitcher misses target by 6-9 inches but it still catches plenty of plate on the outside half. However, catcher has to reach way over across his body to catch the ball. That pitch is called a ball at least 75% of the time.

Extreme example.

What's harder is framing the corners when the pitcher hits the target.

There are so many variables to deal with. For example, is it framing if the ump is consistently calling the outside pitches 3 inches off the plate? That's taking advantage not framing.

Most likely simply correlating the balls called strikes and strikes called balls to a specific catcher is adequate given a large enough sample size. Call it framing and don't sweat the details. But it would take a pretty big sample size to convince me of much.

It would be interesting to break it out by pitcher/catcher combo and even by catcher/pitcher/umpire combo. I just doubt it's meaningful.

Sent from my iPhone using Tapatalk

[CA-ORIOLE]

Extreme example.

What's harder is framing the corners when the pitcher hits the target.

There are so many variables to deal with. For example, is it framing if the ump is consistently calling the outside pitches 3 inches off the plate? That's taking advantage not framing.

Most likely simply correlating the balls called strikes and strikes called balls to a specific catcher is adequate given a large enough sample size. Call it framing and don't sweat the details. But it would take a pretty big sample size to convince me of much.

It would be interesting to break it out by pitcher/catcher combo and even by catcher/pitcher/umpire combo. I just doubt it's meaningful.

Sent from my iPhone using Tapatalk

Pitchers are adjusted for from my understanding. Since they rotate around (I'd assume fairly evenly), I'm not sure there's a real need to adjust for umpires.

As for the sample sizes, they are huge and accounted for over many years of data. Some catchers consistently and clearly do better than others and do it on different teams and/or different staffs. I personally think you can see good ones and bad ones and they generally correlate to the numbers.

I think it's a valid point that teams could look at certain types of pitches/locations/counts/umpires etc. and adapt a strategy to optimize favorable calls. If so, the question I'd ask is why aren't we doing it better?

[El Gordo]

Some catchers get more OZ pitchers called strikes than others, regardless of pitcher, type of pitch, umpire, or team? I doubt they can filter out all that noise. IMO framing is about as valid as CERA. Too many variables to give much credence to either.

[CA-ORIOLE]

Some catchers get more OZ pitchers called strikes than others, regardless of pitcher, type of pitch, umpire, or team? I doubt they can filter out all that noise. IMO framing is about as valid as CERA. Too many variables to give much credence to either.

CERA and pitch framing is not even a close analogy.

[srock]

Some catchers get more OZ pitchers called strikes than others, regardless of pitcher, type of pitch, umpire, or team? I doubt they can filter out all that noise. IMO framing is about as valid as CERA. Too many variables to give much credence to either.

Actually it's just the opposite. All that is needed is to correlate catchers to balls called strikes and strikes called balls then compare the catchers over some number of data points. It seems the the stats people think there is signal doing this.

The hard part is taking each variable into consideration to determine if any of them are meaningful or if they are just noise. I suspect they are, but I do think it's an interesting area to investigate.

Sent from my iPhone using Tapatalk

[El Gordo]

CERA and pitch framing is not even a close analogy.

Care to explain why?