A quick one!

[Thunderheads]

This is from the other day also, but I didn't post it , as I had others to list, ...but .......

Check out this link and drop down to the second ejection, .... DJ Reyburn ...

http://umpireejections.blogspot.com/

Read the description, ...then click on the video attached.....

http://mlb.mlb.com/video/play.jsp?content_id=18955503

The PitchF/x site they use for this doesn't seem right ....that call strike 3 looks a lot further outside than the graph shows, doesn't it ??

Anyhow ....

Thoughts??

[mstaylor]

That's a no brainer.

[Mr Umpire]

I think it looks worse b/c of the camera angle. I believe it was as close as the F/X shows and too close to take like that.

[Umpete]

According to Brooks Baseball, which is responsible for producing the graph:

"The outside edges of the strikezone are then drawn to specifications corresponding to an average umpire's strikezone."

The box is not the "actual" strikezone," but a very close approximation based on the varied heights of batters and the average umpire's strike zone. Then, of course, there is the "special rule" that UEFL uses to determine if the pitch was "close enough" to be called a strike even if it appears out of the strike zone on the graph.

Also remember that camera angles and lens magnification also distort what you see on TV, and that the strike zone is actually three dimensional, and................(on and on and on).

Great pitch by a good pitcher that was close enough to be called a strike.

[grayhawk]

Looks good to me:

[sdix00]

I am gonna use this pic when anyone wants to argue for the Pitch Tracker

[Lindsay]

When we reviewed this play at the Umpire Ejection Fantasy League, we credited Reyburn with a correct call. Here's why: When determining Quality of Correctness for balls/strikes calls, we consider a few dynamics. In regards to this particular sequence, there are two key considerations to keep in mind. Several months ago, we did a feature called "Ending the Game with Science" regarding the pitch f/x technology. I hope you enjoy sports science... and math.

First, we must determine the width of the working horizontal planar strike zone. We know the plate is 17" in diameter, or 8.5" on either side (radial value, Rule 1.05). We also know a regulation MLB baseball must have a circumference no greater than 9.25 inches (Rule 1.09). Circumference = pi * diameter; therefore, diameter = circumference / pi; diameter = 9.25 / 3.14159 ... = a baseball's maximum diameter is 2.944 inches. Since there are two sides of home plate that any part of the ball may pass through and still be within the strike zone, the working horizontal planar strike zone is 17" + 2.944" + 2.944", or 22.880 inches wide. The radius, therefore, of the working zone is 11.440 inches, which converts to approximately 0.953 feet (pitch f/x charts use a horizontal unit of feet).

Pitch f/x carries with it a margin of error of approximately one inch, per the manufacturer (this may explain why the pitch f/x graph might look "off"). One inch is 0.0833 feet. By adding and subtracting 0.0833 feet to the observed px [horizontal] value, we create a confidence interval (of 100%): we know the pitch definitely was somewhere within this range. Computing the confidence interval (CI) for Reyburn's strike three pitch yields a lower bound of 0.894 ft and an upper bound of 1.060 ft (Pitch f/x generated a px [horizontal] value of 0.977 ft). Per our calculation above, the range of "definite strike" includes values between -0.953 feet and 0.953 feet, where negative values refer to the graph's left side [right handed batter's side] and positive values refer to the graph's right side [left handed batter's side]. Subtracting 0.953 from 0.894 yields approx. absolute value 0.059. Subtracting 0.953 from 1.060 yields approx. absolute value 0.107. Using this information, we find exactly what percent chance the pitch was actually a strike and what percent chance the pitch was actually a ball. For this pitch, it comes out to 36% chance strike & 64% chance ball.

For us at the UEFL, a 36/64 split means there is a significant chance of the pitch being either a strike or a ball. Under the Kulpa Rule, we call this "borderline." For us, a determination of "borderline" routinely reflects the call on the field as correct. Generally speaking, we consider any legal pitch between 0 and 0.900 to always be a strike, between 0.900 and 1.000 to be borderline, and above 1.000 to be a ball (if there was no swing, etc.). Absolute strike (meaning a 100% probability the pitch is a strike) occurs when px decreases to 0.8697, while absolute ball (meaning a 100% probability the pitch is a ball) occurs when px increases to 1.0363. .900 is a 18/82 split, while 1.000 is a 22/78 split: we give more credence to the called ball; regardless, a CI which contains both sides of the .953 "strike zone border" is usually considered borderline, and therefore, correct. So that's the long story of why Reyburn got this call right.

[dumbdumb]

If you look at the pitch f/x graph and take a ruler and go from the +1 to the -1 which was the strike zone width for inside and out last year, you will pretty much know what is a strike or is not a strike on the corners when they are technically outside the slimmed down version that was brought in a notch and used this year IMHO. The technical statistics, and scientific reasoning has been explained by Gil. But, going from the -1 to +1 makes it easier for me to figure out.

[mstaylor]

When we reviewed this play at the Umpire Ejection Fantasy League, we credited Reyburn with a correct call. Here's why: When determining Quality of Correctness for balls/strikes calls, we consider a few dynamics. In regards to this particular sequence, there are two key considerations to keep in mind. Several months ago, we did a feature called "Ending the Game with Science" regarding the pitch f/x technology. I hope you enjoy sports science... and math.

First, we must determine the width of the working horizontal planar strike zone. We know the plate is 17" in diameter, or 8.5" on either side (radial value, Rule 1.05). We also know a regulation MLB baseball must have a circumference no greater than 9.25 inches (Rule 1.09). Circumference = pi * diameter; therefore, diameter = circumference / pi; diameter = 9.25 / 3.14159 ... = a baseball's maximum diameter is 2.944 inches. Since there are two sides of home plate that any part of the ball may pass through and still be within the strike zone, the working horizontal planar strike zone is 17" + 2.944" + 2.944", or 22.880 inches wide. The radius, therefore, of the working zone is 11.440 inches, which converts to approximately 0.953 feet (pitch f/x charts use a horizontal unit of feet).

Pitch f/x carries with it a margin of error of approximately one inch, per the manufacturer (this may explain why the pitch f/x graph might look "off"). One inch is 0.0833 feet. By adding and subtracting 0.0833 feet to the observed px [horizontal] value, we create a confidence interval (of 100%): we know the pitch definitely was somewhere within this range. Computing the confidence interval (CI) for Reyburn's strike three pitch yields a lower bound of 0.894 ft and an upper bound of 1.060 ft (Pitch f/x generated a px [horizontal] value of 0.977 ft). Per our calculation above, the range of "definite strike" includes values between -0.953 feet and 0.953 feet, where negative values refer to the graph's left side [right handed batter's side] and positive values refer to the graph's right side [left handed batter's side]. Subtracting 0.953 from 0.894 yields approx. absolute value 0.059. Subtracting 0.953 from 1.060 yields approx. absolute value 0.107. Using this information, we find exactly what percent chance the pitch was actually a strike and what percent chance the pitch was actually a ball. For this pitch, it comes out to 36% chance strike & 64% chance ball.

For us at the UEFL, a 36/64 split means there is a significant chance of the pitch being either a strike or a ball. Under the Kulpa Rule, we call this "borderline." For us, a determination of "borderline" routinely reflects the call on the field as correct. Generally speaking, we consider any legal pitch between 0 and 0.900 to always be a strike, between 0.900 and 1.000 to be borderline, and above 1.000 to be a ball (if there was no swing, etc.). Absolute strike (meaning a 100% probability the pitch is a strike) occurs when px decreases to 0.8697, while absolute ball (meaning a 100% probability the pitch is a ball) occurs when px increases to 1.0363. .900 is a 18/82 split, while 1.000 is a 22/78 split: we give more credence to the called ball; regardless, a CI which contains both sides of the .953 "strike zone border" is usually considered borderline, and therefore, correct. So that's the long story of why Reyburn got this call right.

Welcome tho the site, Gil. I must admit I never really got anything out of the graphs but it does make better sense with the math.

[Lindsay]

Welcome tho the site, Gil. I must admit I never really got anything out of the graphs but it does make better sense with the math.

Thanks for the welcome. I started to see "umpire-empire.com" show up as one of the leading "referring websites" in our site stats and thought I should come on over!

[Thunderheads]

Welcome tho the site, Gil. I must admit I never really got anything out of the graphs but it does make better sense with the math.

Thanks for the welcome. I started to see "umpire-empire.com" show up as one of the leading "referring websites" in our site stats and thought I should come on over!

WELCOME! Come on over often Gil ....

Also, ... I check out your site every morning to see who tossed who!! :)

[UmpJM]

Thunder,

"Who tossed whom ?"

Children can read this site.

JM

[sdix00]

Are we allowed to correct grammar, syntax and style now?

Release the hounds!

[Thunderheads]

Thunder,

"Who tossed whom ?"

Children can read this site.

JM

or for God's sake .................:rolleyes:

[UmpPhil]

When we reviewed this play at the Umpire Ejection Fantasy League, we credited Reyburn with a correct call. Here's why: When determining Quality of Correctness for balls/strikes calls, we consider a few dynamics. In regards to this particular sequence, there are two key considerations to keep in mind. Several months ago, we did a feature called "Ending the Game with Science" regarding the pitch f/x technology. I hope you enjoy sports science... and math.

First, we must determine the width of the working horizontal planar strike zone. We know the plate is 17" in diameter, or 8.5" on either side (radial value, Rule 1.05). We also know a regulation MLB baseball must have a circumference no greater than 9.25 inches (Rule 1.09). Circumference = pi * diameter; therefore, diameter = circumference / pi; diameter = 9.25 / 3.14159 ... = a baseball's maximum diameter is 2.944 inches. Since there are two sides of home plate that any part of the ball may pass through and still be within the strike zone, the working horizontal planar strike zone is 17" + 2.944" + 2.944", or 22.880 inches wide. The radius, therefore, of the working zone is 11.440 inches, which converts to approximately 0.953 feet (pitch f/x charts use a horizontal unit of feet).

Pitch f/x carries with it a margin of error of approximately one inch, per the manufacturer (this may explain why the pitch f/x graph might look "off"). One inch is 0.0833 feet. By adding and subtracting 0.0833 feet to the observed px [horizontal] value, we create a confidence interval (of 100%): we know the pitch definitely was somewhere within this range. Computing the confidence interval (CI) for Reyburn's strike three pitch yields a lower bound of 0.894 ft and an upper bound of 1.060 ft (Pitch f/x generated a px [horizontal] value of 0.977 ft). Per our calculation above, the range of "definite strike" includes values between -0.953 feet and 0.953 feet, where negative values refer to the graph's left side [right handed batter's side] and positive values refer to the graph's right side [left handed batter's side]. Subtracting 0.953 from 0.894 yields approx. absolute value 0.059. Subtracting 0.953 from 1.060 yields approx. absolute value 0.107. Using this information, we find exactly what percent chance the pitch was actually a strike and what percent chance the pitch was actually a ball. For this pitch, it comes out to 36% chance strike & 64% chance ball.

For us at the UEFL, a 36/64 split means there is a significant chance of the pitch being either a strike or a ball. Under the Kulpa Rule, we call this "borderline." For us, a determination of "borderline" routinely reflects the call on the field as correct. Generally speaking, we consider any legal pitch between 0 and 0.900 to always be a strike, between 0.900 and 1.000 to be borderline, and above 1.000 to be a ball (if there was no swing, etc.). Absolute strike (meaning a 100% probability the pitch is a strike) occurs when px decreases to 0.8697, while absolute ball (meaning a 100% probability the pitch is a ball) occurs when px increases to 1.0363. .900 is a 18/82 split, while 1.000 is a 22/78 split: we give more credence to the called ball; regardless, a CI which contains both sides of the .953 "strike zone border" is usually considered borderline, and therefore, correct. So that's the long story of why Reyburn got this call right.

You have just blinded me with science!

[Umpire in Chief]

I am gonna use this pic when anyone wants to argue for the Pitch Tracker

Wow!

I will too I'm so sick of those Fox Box, K-Zone...

Welcome tho the site, Gil. I must admit I never really got anything out of the graphs but it does make better sense with the math.

+1 Welcome! Glad to have you here.

Thanks for the welcome. I started to see "umpire-empire.com" show up as one of the leading "referring websites" in our site stats and thought I should come on over!

Thanks for noticing

[hbk314]

Looks good to me:

The pitch over in the box was actually the third pitch. The called strike three was pitch number four of the AB. Note the 1-2 count as the pitch is going in.

[hbk314]

If you want a good example of how useless they are, look at this. All three of these are for the same pitch. Pitch #5 of the AB.

Boston's "K-Zone":

Baltimore's "K-Zone":

PitchFX:

[dumbdumb]

However, the system used to grade the umpires is called Zone Evaluation.

Here is an article about the system

http://www.nytimes.com/2009/04/01/sports/baseball/01umpires.html

Also the old grievance filed in 2001 about wanting umpires to speed up games by calling more strikes prior to these pitch evaluation systems to speed games up

World Umpires Assoc. pitch count grievance

news.findlaw.com/hdocs/docs/mlb/wuamlb71401.pdf

[sdix00]

If you want a good example of how useless they are, look at this. All three of these are for the same pitch. Pitch #5 of the AB.

Boston's "K-Zone":

Baltimore's "K-Zone":

PitchFX:

Hahaha. Fascinating.

I found a fourth. It's obviously wrong, though.

[Rich Ives]

When we reviewed this play at the Umpire Ejection Fantasy League, we credited Reyburn with a correct call. Here's why: When determining Quality of Correctness for balls/strikes calls, we consider a few dynamics. In regards to this particular sequence, there are two key considerations to keep in mind. Several months ago, we did a feature called "Ending the Game with Science" regarding the pitch f/x technology. I hope you enjoy sports science... and math.

First, we must determine the width of the working horizontal planar strike zone. We know the plate is 17" in diameter, or 8.5" on either side (radial value, Rule 1.05). We also know a regulation MLB baseball must have a circumference no greater than 9.25 inches (Rule 1.09). Circumference = pi * diameter; therefore, diameter = circumference / pi; diameter = 9.25 / 3.14159 ... = a baseball's maximum diameter is 2.944 inches. Since there are two sides of home plate that any part of the ball may pass through and still be within the strike zone, the working horizontal planar strike zone is 17" + 2.944" + 2.944", or 22.880 inches wide. The radius, therefore, of the working zone is 11.440 inches, which converts to approximately 0.953 feet (pitch f/x charts use a horizontal unit of feet).

Pitch f/x carries with it a margin of error of approximately one inch, per the manufacturer (this may explain why the pitch f/x graph might look "off"). One inch is 0.0833 feet. By adding and subtracting 0.0833 feet to the observed px [horizontal] value, we create a confidence interval (of 100%): we know the pitch definitely was somewhere within this range. Computing the confidence interval (CI) for Reyburn's strike three pitch yields a lower bound of 0.894 ft and an upper bound of 1.060 ft (Pitch f/x generated a px [horizontal] value of 0.977 ft). Per our calculation above, the range of "definite strike" includes values between -0.953 feet and 0.953 feet, where negative values refer to the graph's left side [right handed batter's side] and positive values refer to the graph's right side [left handed batter's side]. Subtracting 0.953 from 0.894 yields approx. absolute value 0.059. Subtracting 0.953 from 1.060 yields approx. absolute value 0.107. Using this information, we find exactly what percent chance the pitch was actually a strike and what percent chance the pitch was actually a ball. For this pitch, it comes out to 36% chance strike & 64% chance ball.

For us at the UEFL, a 36/64 split means there is a significant chance of the pitch being either a strike or a ball. Under the Kulpa Rule, we call this "borderline." For us, a determination of "borderline" routinely reflects the call on the field as correct. Generally speaking, we consider any legal pitch between 0 and 0.900 to always be a strike, between 0.900 and 1.000 to be borderline, and above 1.000 to be a ball (if there was no swing, etc.). Absolute strike (meaning a 100% probability the pitch is a strike) occurs when px decreases to 0.8697, while absolute ball (meaning a 100% probability the pitch is a ball) occurs when px increases to 1.0363. .900 is a 18/82 split, while 1.000 is a 22/78 split: we give more credence to the called ball; regardless, a CI which contains both sides of the .953 "strike zone border" is usually considered borderline, and therefore, correct. So that's the long story of why Reyburn got this call right.

So if the margin of error is 1" that means it is +- 1" so it could just as well be marking the ball 1" too close as well as 1" too far.

[zm1283]

I have a friend (Who is also a coach BTW) that claims that on the boxes that the networks put on screen (K-Zone, etc), the ball has to be entirely inside the box for it to be called a strike. I asked him if he really thinks that a pitch that touches the lines of that box never touched any of the three-dimensional strike zone. Oddly enough he can't tell me where he has heard this from. Probably from an announcer on TV.

[grayhawk]

Looks good to me:

The pitch over in the box was actually the third pitch. The called strike three was pitch number four of the AB. Note the 1-2 count as the pitch is going in.

I wasn't commenting on the tracker. I was showing that the pitch hit the outside corner - look where the ball is on the freeze frame.

[Umpire in Chief]

I have a friend (Who is also a coach BTW) that claims that on the boxes that the networks put on screen (K-Zone, etc), the ball has to be entirely inside the box for it to be called a strike. I asked him if he really thinks that a pitch that touches the lines of that box never touched any of the three-dimensional strike zone. Oddly enough he can't tell me where he has heard this from. Probably from an announcer on TV.

Great point the zone is 3D. I haven't seen it many times but I have seen it a few where the ball catches the back corner, and I've gotten it.

[mjr_2013]

Seriously? I'm the only one that caught this?

In the description of that video on MLB.com, his last name is spelled "Reybum"... Of all the possible misspellings, really? Reybum?

Real mature, MLB! :clap: