Ball & Strike Challenges

redsox102704

New Member
May 9, 2010
243
I think it's time to implement a challenge system for balls and strikes. Since every stadium (I think) has an electronic strike zone that we see on TV, here's what I was thinking of. Each team can challenge three pitches per game. If they are proven right on any of the three, one more will be awarded.

The beauty of the system is that the computer has already decided whether the pitch is a ball or strike as soon as it crosses the plate. No replay is necessary. The challenge would simply entail the home plate umpire (or crew chief) signaling (or using an earpiece to contact) a replay official. The replay official would then simply tell the umpire what the pitch computer saw. The whole process shouldn't take more than 15 seconds.

This system still allows the human element to remain, but allows for some relief on a few close pitches that can decide a game.
 

InstaFace

The Ultimate One
SoSH Member
Sep 27, 2016
21,754
Pittsburgh, PA
How is there a human element if the computer overrules the human every time the computer thinks the human is wrong? Why bother with the human any more?
Because the humans are unionized. Don't hold your breath for robot USPS mail delivery either.

Half-jokes aside, I'd argue the purpose of replay review in sports is to avoid the consequences of an obviously, objectively wrong call. We have that today in baseball: home run / foul, trap / catch, safe / out, etc. But I'm not sure borderline pitch calls rise to the level of "objectively right/wrong" very often.
 

bsj

Renegade Crazed Genius
SoSH Member
Dec 6, 2003
22,774
Central NJ SoSH Chapter
Balls and strikes should be computer called to begin with. Humans were instilled as arbiters in sport when technologies to do so we're not even imaginable. To intentionally maintain a fallible system for purposes of tradition is ridiculous.
 

nvalvo

Member
SoSH Member
Jul 16, 2005
21,478
Rogers Park
A problem with this that is rarely acknowledged is that the pitchFX cameras/computers just plain miss some small percentage of pitches. Not a ton, not even one percent. But some.

Until that changes, we aren't ready.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
A problem with this that is rarely acknowledged is that the pitchFX cameras/computers just plain miss some small percentage of pitches. Not a ton, not even one percent. But some.
It's worse than 1%, I think, depending on what a "miss" is. We really don't know what the actual accuracy of the PITCHf/x (or, this year, the TrackMan) systems are in practice. Assuming that PITCHf/x works like every other instrument, presumably there's a roughly normal distribution of accuracy, where 68% of the pitches are shown within one standard deviation of their true location. What's that standard deviation? No idea. Probably the operators know, but we don't.

My feeling, mostly a guess, is that one standard deviation is on the order of 1/2 inch, in practice. That's pretty good, but it means that about 5% of pitches are one inch off, and about 0.3% are two inches away from their true location. (Plus, occasional whacky things where PITCHf/x measures the wrong thing. Much more rare now than they used to be, but I think the PITCHf/x release point still shows Koji Uehara throwing one pitch left-handed, for example.)

One inch isn't much, but it's certainly enough to change a ball from a strike. That's going to affect about 15-20 pitches per game -- mostly harmlessly since most pitches are not within an inch of the edge of the zone, but a lot of pitches are, since that's where pitchers aim.

I think umpires are about 97% accurate (but that's based on comparing to PITCHf/x ... but I also base it on changes larger than an inch). I don't know that PITCHf/x would be more accurate than umpires.

I have no sense for how accurate the TrackMan system, which they started using this year instead of PITCHf/x, is. If the standard deviation is less than a half inch under actual ballpark conditions, then it could enter the conversation.
 

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
The other factor we have to consider is that the plot we see on the screen isn't the strike zone in that particular AB. It's an approximation of the league average strike zone. But the zone changes with every hitter because every hitter is a different height. The left and right edges should remain constant, of course, but high and low strikes will not be. So using PITCHf/x in disputes over high or low strikes may not be any more accurate than the ump on any given call.
 

charlieoscar

Member
Sep 28, 2014
1,339
Another factor to consider is that the plot seen on the screen is two-dimensional; whereas, the strike zone is three-dimensional. So where does the measurement take place...at the front edge of the plate or the back end of the strike zone?

From MLB Rules: The STRIKE ZONE is that area over home plate the upper limit of which is a horizontal line at the midpoint between the top of the shoulders and the top of the uniform pants, and the lower level is a line at the hollow beneath the kneecap. The Strike Zone shall be determined from the batter’s stance as the batter is prepared to swing at a pitched ball.

For starters, how do you accurately determine where "the hollow beneath the kneecap" is, especially when the batter is wearing baggy pants?

That aside, a sinking fastball may be thrown so the top of the ball just clips the bottom of the strike zone right at the front edge of the plate (i.e., the beginning of the strike zone) and then continue sinking so it is below the hollow of the knee when it reaches the batter's legs. It may not look like one but by definition it is one. An umpire is positioned a few feet from the from edge of the plate. How well can he, or something like PITCHf/x or TrakMan, judge the exact moment when the ball enters the strike zone?

By the same token, the umpire must judge the exact midway distance between the top of the batter's shoulders and the top of his uniform pants. Batters may well be into their swing before the ball even gets to the strike zone. If a right-handed batter checks his swing, his right shoulder probably drops and his left hip may rise, disrupting the guidelines. In this case, you have the bottom of a high pitch clipping the top of the rear of the strike zone.

On the right- and left-hand sides of the strike zone you have pitches that tail away from the zone. How easy it it to judge, humanly or electronically, if a ball clips the fron of the zone and tails away of curves into the rear of the zone?

Finally, a 100-mph pitch is in the strike zone for approximately 11.3 milliseconds (d=17" + diameter of ball).

As an aside to this, I had a tour of the PITCHf/x operation during a game at St. Louis back in 2007 and they had a person watching live video for each batter so they could place the bottom of the strike zone. This was done preparatory to building a database for each batter but if the operator chose to mark it when the batter first got to the box as opposed to when the batter took his actual batting position, there could have been differences.
 

Kenny F'ing Powers

posts way less than 18% useful shit
SoSH Member
Nov 17, 2010
14,425
The other factor we have to consider is that the plot we see on the screen isn't the strike zone in that particular AB. It's an approximation of the league average strike zone. But the zone changes with every hitter because every hitter is a different height. The left and right edges should remain constant, of course, but high and low strikes will not be. So using PITCHf/x in disputes over high or low strikes may not be any more accurate than the ump on any given call.
Put a sensor on players belts. Inspect uniforms before the game to ensure guys aren't wearing belts down to their asshole. Problem solved.
 

Rasputin

Will outlive SeanBerry
Lifetime Member
SoSH Member
Oct 4, 2001
29,419
Not here
I think umpires are about 97% accurate (but that's based on comparing to PITCHf/x ... but I also base it on changes larger than an inch). I don't know that PITCHf/x would be more accurate than umpires.
Accuracy is only part--and the lesser part--of the problem. Every single person in the game and the vast majority of the fans will tell you that consistency is more important than accuracy. The errors you speak of will clearly lead to some inconsistency but I rather suspect it would be an improvement over the bad umpires.

Personally, I would like to see an experiment where they put something akin to an RFID chip in the ball to see if it still plays like the current balls. I could be mistaken, of course, but I'd think it would improve accuracy.
 

Zososoxfan

Member
SoSH Member
Jul 30, 2009
9,208
South of North
It seems painfully obvious to me that you would just standardize the bottom and top of the strike zone. 30 inches to 60 inches or whatever makes the most sense.
 

Zososoxfan

Member
SoSH Member
Jul 30, 2009
9,208
South of North
Some things don't make sense.
I understand that and don't think it should have an impact. Make the zone as big or as small (and as low and as tall) as desired and have the players learn their way around it. I don't have a problem making short guys learn how to hit high ones (Pedey would sign up for this yesterday) and tall guys to hit low ones.
 

joe dokes

Member
SoSH Member
Jul 18, 2005
30,236
I understand that and don't think it should have an impact. Make the zone as big or as small (and as low and as tall) as desired and have the players learn their way around it. I don't have a problem making short guys learn how to hit high ones (Pedey would sign up for this yesterday) and tall guys to hit low ones.
For stickball in the schoolyard in brooklyn, the strike zone was drawn in chalk on the wall behind the batter. It didn't change for the short kids. Chalk on the ball meant strike on the black.
 

timlinin8th

Member
SoSH Member
Jun 6, 2009
1,521
The answer really lies in between: Automate the static horizontal axis of the plate, put a small LED in the plate umps mask that lights up if the ball goes over the plate (using an rfid in the ball or some other method) and let the umps call the vertical "batter height" part of the zone. The egregious 6 inch outside calls go away, but the union stays happy because the humans get to human still. Its a small step but the game needs to start making small steps instead of trying to integrate sweeping changes that go nowhere.

*edit* and go the point of ball and strike challenges, NO EFFIN WAY. Omg that would be a colossal disaster.
 

DJnVa

Dorito Dawg
SoSH Member
Dec 16, 2010
53,838
It's based on the height when the player swings. If the guy puts his chin on the ground when waiting for the pitch and stands at normal height when swinging, the latter is what counts.
It's when "preparing to swing", not when when swinging, although this could be a minuscule difference.
 

Savin Hillbilly

loves the secret sauce
SoSH Member
Jul 10, 2007
18,783
The wrong side of the bridge....
I understand that and don't think it should have an impact. Make the zone as big or as small (and as low and as tall) as desired and have the players learn their way around it. I don't have a problem making short guys learn how to hit high ones (Pedey would sign up for this yesterday) and tall guys to hit low ones.
This would effectively create a built-in advantage for players of average height at the expense of players at the extremes, and to what purpose? Just to facilitate robot-umpery? I'm pro-bot in principle, but I think this is a terrible idea. If you can't implement automated ball-strike calls in a way that works according the rulebook for players of every size, then don't do it.
 

BigChara33

New Member
May 2, 2017
79
Then someone needs to just crouch all the way until their knees touch their chest. Good luck getting a strikeout...
 

garlan5

Member
SoSH Member
May 13, 2009
2,684
Virginia
Just put a virtual reality headset on the ump so he can watch pitchfx then just have him do some funny antics like the naked gun
 

Sampo Gida

Member
SoSH Member
Aug 7, 2010
5,044
MLB seems to have taken a step back from Pitch f/x as statcast is no tas accurate or reliable according to this

https://www.google.com.tw/amp/s/fivethirtyeight.com/features/baseballs-new-pitch-tracking-system-is-just-a-bit-outside/amp/

Pitch f/x is better, but not perfectly accurate as well. So one has to determine what is the accuracy and what calls are clearly outside defined error bands, be that 1 inch, 2 inch etc. Borderline calls within thesecerror bands wont be able to be changed.

As I recall , SportsVision claims Pitch f/x is accurate to within 1 inch when properly calibrated. However it is not always properly calibrated

Calibration Drifting due to lighting changes and the stadium moving resulting in increased error during a game. My understanding is recalibration was notvpissible during the game.

Measurement Error (Calibration Error) by Stadium can be greater than 1.5 inches horizontally and 1.1 inches vertically although I believe that may have improved since the early years.

I'd like to see a system that is self correcting rather than challenge based. A wireless transmission to the ump instructing a change to his call. This eliminates some ludicrous calls where balls right down the middle are called balls and balls 3 inches outside are called strikes.

If the system is down, out of calibration or whatever then so be it. Umps call stands as it has for over 100 years.
 

Dewey'sCannon

Member
SoSH Member
Jul 18, 2005
869
Maryland
I think the first step should be to use Pitch F/X (or whatever automated system works best) as an "aid" to the umpire - if the pitch is in the zone, it triggers a vibration in the ump's ball/strike counter (do they still have those) so he has this info when he makes his call.

And part of the review process for the umps would be to track the rate of deviation fro each ump and review those calls where the ump's call differed from the system to see if the call was reasonably justifiable.
 

AB in DC

OG Football Writing
SoSH Member
Jul 10, 2002
13,616
Springfield, VA
I watched part of the Wimbledon men's final this weekend. A couple times, one of the players challenged whether a ball was inside the line or outside the line. Each challenge took 15 seconds at most. I'm not exaggerating -- challenge was made, someone punched it up on a computer, and the visuals were immediately displayed on a big monitor next to the court. If you blinked you probably wouldn't even notice it.

If baseball could do the same thing for the strike zone -- at least on the horizontal access -- would that really be such a bad thing?
 

Hoodie Sleeves

Member
SoSH Member
Nov 24, 2015
1,204
If I had to choose one that would more accurately and more frequently do so, it would not be the human crouching behind the plate.
Yeah, this.

PitchFX/TrackMan/Whatever doesn't have to be perfect. It has to be better than the umpires, and from what I've seen, it already is.


The vertical thing is nonsense - the umps make a ton of high/low strike mistakes, and the adjustments they make to tall/short batters are almost non-existent.


This whole thing sounds like the argument against robot cars - where people are arguing about how a car will handle the Train Problem - while we've got people driving around texting and eating a burrito while driving.

Don't let the best be the enemy of the better.
 
Last edited:

Sampo Gida

Member
SoSH Member
Aug 7, 2010
5,044
One of the concerns with using tech to make calls is security and also reliability issues. For example can the system be hacked and someone introduce a bias to one team or another by altering the algorithm. Gamblers or even some teams might be so motivated. How quickly can such a hack be detected? Would we ever know about it?

Also, calibrations fail and such failures may not be detected right away. We will all trust the robot when he is as wrong as a drunk and blind ump behind the plate because robots are always right, even when they are off.

Then there is the human element. Pitchers with good command and catchers with good framing skills getting more calls than those who dont, and hitters with the ability to adjust to a variable zone better than those who just swing at everything after a bad call. Do we want to eliminate that?
 

uncannymanny

Member
SoSH Member
Jan 12, 2007
9,078
To throw some cold water on the smart ball idea:

1) a quick Google says MLB uses approx 160k in-game balls per year. Whose salary are the 160k smart chips and whatever they need to make them impact-resilient coming out of? Certainly not either union, so it's likely that MLB itself would have to care enough to spend whatever it costs to make those balls.

2) RFID/BTLE/etc are also, having programmed with both, not capable of sub-inch accuracy, especially moving at 90+ mph. They would also be used in an environment full of other devices. We've seen how reliable the NFL's wireless headsets are.

3) Any ball tracking tech would have to be installed in the plate and, although it's possible, I'm not sure how accurate separating balls, hands and bats crossing the plate would be.

It's far more likely that a better camera-tracking system is the answer, which brings us right back to improving the current systems.

(Maybe I'm forgetting some tech, forgive me if I missed something obvious)
 
Last edited:

soxhop411

news aggravator
SoSH Member
Dec 4, 2009
46,274
I think tonight's Yankees/Red Sox game was called perfectly behind home plate



 

NJ_Sox_Fan

Member
SoSH Member
Jan 2, 2006
10,736
NJ
I just wish they'd get rid of the umps who can't call strikes and balls correctly. Holbrook, for example.
 

catomatic

thinks gen turgidson is super mean!!!
SoSH Member
Jul 16, 2005
3,390
Park Slope, Brooklyn
Give the umps Google Glasses and real time Pitch FX/Questec info on the lens so they can keep a small measure of their dignity. Was at the Stade Fasciste this evening and even if this Sox edition came up small every time there was a RISP and the Yanks took full advantage of extra pitches, tonight's umpiring was atrocious and no doubt affected the outcome.
 

Pitt the Elder

Member
SoSH Member
Sep 7, 2013
4,417
Thanks for posting this. I was at the game tonight along the third base side and had a hard time gauging balls and strikes but I got the sense that the ump was squeezing the zone given how often Sandy froze his glove for a few seconds on close calls. Seemed to really change the complexion of the game, particularly because Sale wasn't missing bats like be usually does. And that doesn't even include the Reed / Sanchez at bat where it looked like he through him strike 3 multiple times.

That said, it's just one game and tomorrow is another day. Another park and another ump, and Sale goes 7 with 10k and only a couple runs.
 

geoduck no quahog

not particularly consistent
Lifetime Member
SoSH Member
Nov 8, 2002
13,024
Seattle, WA
An NCAA ump I know says he and his cohorts can't really judge high and low accurately. The system they use is based on the catcher's knees being generally at the bottom of the zone and his mask at the top. Interesting enough, since he mentioned that, I've noticed some similarity between the real zone and the catcher's stance. So they're looking at where the ball travels and watching the glove action relative to the crouch.
Dirty little secret I guess.

Me? I'd be much happier if the K Zones disappeared from the screen. I preferred when you couldn't get upset 85 times a game.
 

Snodgrass'Muff

oppresses WARmongers
SoSH Member
Mar 11, 2008
27,644
Roanoke, VA
Me? I'd be much happier if the K Zones disappeared from the screen. I preferred when you couldn't get upset 85 times a game.
I stopped getting upset so often when I accepted that the top and the bottom of the zone fluctuate from batter to batter so close calls there aren't necessarily being shown accurately on the screen, and that the system itself isn't 100% accurate from the get go, so even pitches on the inside or outside of the plate who looks like obvious balls or strikes might not be shown correctly.

Sure, the egregious ones are still gonna get us rankled, but those would likely be catching our eye with or without the strike zone plot.
 

Al Zarilla

Member
SoSH Member
Dec 8, 2005
58,867
San Andreas Fault
A problem with this that is rarely acknowledged is that the pitchFX cameras/computers just plain miss some small percentage of pitches. Not a ton, not even one percent. But some.

Until that changes, we aren't ready.
I'm sure there are some air traffic control images where the radar or computers or some other part of the system are in error, but it's better than having some people standing at runways deciding which planes should land when. What pitchFX has right now has to be far better than Holbrook was last night. Baseball (sports?) is in the dark ages when it comes to using available technology. Yes, there are nuances like distinguishing Judge from Altuve, but that's pretty simple to work out without starting a whole Manhattan Project.