Forward pass problem solved

[Grandpa]

In complicated matters such as deciding if a players hands are passing the ball backwards it is always useful to refer back to Jacobellis v. Ohio, 1964 whereby Judge Potter described his threshold test for whether something was porn... 'I know it when I see it'.

But can he distinguish between backwards porn and forwards porn?

[Raggs]

The simplest test is just was it launched forward from the player (taking his speed in to account). Is it launched to go ahead of him or not?

If you were a line judge running at the same speed as the player, does it end up further ahead of you or not.

Problem with that one is we have seen passes that are just momentum, being call forward, if the player takes a tackle a split second after the pass. Since their momentum stops and the ball keeps going. A technical solution would remove that.

[Ymx]

Oh absolutely. I’m all for a tech solution for this one.

It’s just too hard to judge and requires refs to be in the right position and looking for it.

Simple high precision gps chip, calibrated to the corner posts would do the trick.
Identify its position and velocity at any time
Identify a pass release

[Tichtheid]

To be honest I don't think the forward pass is a big deal now that the argument for forward motion of the ball carrier has been accepted by the authorities.

There are much bigger issues in the game where better consistency in refereeing is really needed. Contact to the head is one such topic and the breakdown is always going to cause controversy

[Raggs]

To be honest I don't think the forward pass is a big deal now that the argument for forward motion of the ball carrier has been accepted by the authorities.

There are much bigger issues in the game where better consistency in refereeing is really needed. Contact to the head is one such topic and the breakdown is always going to cause controversy

If the tech is available why not use it though?

There's tech being developed to record gforces to the skull as well.

As for consistency, I blame the southern hemisphere. We were red hot in the North ( at least home Nations) but it was clear the sh weren't singing from the same hymn sheet. As someone else put it, we either solve it on the pitch or it'll get solved off the pitch at a fast greater cost.

[JM2K6]

Oh absolutely. I’m all for a tech solution for this one.

It’s just too hard to judge and requires refs to be in the right position and looking for it.

Simple high precision gps chip, calibrated to the corner posts would do the trick.
Identify its position and velocity at any time
Identify a pass release

Not enough. GPS alone won't do it, and it just isn't as accurate as people think. It's not magic!

[Ymx]

Oh absolutely. I’m all for a tech solution for this one.

It’s just too hard to judge and requires refs to be in the right position and looking for it.

Simple high precision gps chip, calibrated to the corner posts would do the trick.
Identify its position and velocity at any time
Identify a pass release

Not enough. GPS alone won't do it, and it just isn't as accurate as people think. It's not magic!

That’s kind of why I was suggesting corner flag calibration. Rather than using orbiting transmitting satellites.

But it might be that a good solution is with AI processing multi angle video.

What’s the system in the OP using?

[Ymx]

But we will need to ban baldys from rugby to use AI

https://www.theverge.com/tldr/2020/11/3 ... r-mistakes

[JM2K6]

Oh absolutely. I’m all for a tech solution for this one.

It’s just too hard to judge and requires refs to be in the right position and looking for it.

Simple high precision gps chip, calibrated to the corner posts would do the trick.
Identify its position and velocity at any time
Identify a pass release

Not enough. GPS alone won't do it, and it just isn't as accurate as people think. It's not magic!

That’s kind of why I was suggesting corner flag calibration. Rather than using orbiting transmitting satellites.

But it might be that a good solution is with AI processing multi angle video.

What’s the system in the OP using?

The pseudolite solution is (afaik) very, very expensive but yes, more accurate - cm-level at best but still much better than standard GPS. Would it be enough, do we think?

From the OP

To track the flight of the ball, the Steeden is implanted with a microchip that features an accelerometer, a gyroscope, a magnetometer, a pressure sensor and a temperature sensor, many of the technologies found in a normal mobile phone. Then there’s also additional technology that tracks the ball like a radar. There’s also a second method the company uses to track ball flight that the company wants to keep confidential for now.

So multiple relevant techs and then ~secret sauce~, hence me being a bit cautious about this.

[Raggs]

Surely the vast majority of the "heavy" lifting in terms of determining forwards/backwards wouldn't be done by gps/flag etc at all, but simply by accelerometers?

[Slick]

To be honest I don't think the forward pass is a big deal now that the argument for forward motion of the ball carrier has been accepted by the authorities.

There are much bigger issues in the game where better consistency in refereeing is really needed. Contact to the head is one such topic and the breakdown is always going to cause controversy

If the tech is available why not use it though?

There's tech being developed to record gforces to the skull as well.

As for consistency, I blame the southern hemisphere. We were red hot in the North ( at least home Nations) but it was clear the sh weren't singing from the same hymn sheet. As someone else put it, we either solve it on the pitch or it'll get solved off the pitch at a fast greater cost.

Well that's it really, seems a lot of cost for an issue that isn't really that big

[Blake]

You wouldn't use GPS but the concept is similar...triangulation; but instead of satellites in space and a receiver in a phone, you put a signal generator in the ball and multiple sensors around the pitch. With a setup like this you can have a very high degree of accuracy.

Which takes us back to this...

Great.

Who is going to pay for the technical infrastructure and extra staff required to install and use this technology?

It will be similar to Hawk-Eye in cricket. There is some investment the stadium operator can make and once done the technology is available at that ground.
It is seriously not prohibitively expensive. You need a couple of special balls with a transmitter embedded, a couple of sensors around the field to catch those transmissions, and some hardware and software to crunch the values.

If stadiums can afford Spidercams or goal-posts that light up on conversions, they can afford this.

[Grandpa]

You wouldn't use GPS but the concept is similar...triangulation; but instead of satellites in space and a receiver in a phone, you put a signal generator in the ball and multiple sensors around the pitch. With a setup like this you can have a very high degree of accuracy.

Which takes us back to this...

Great.

Who is going to pay for the technical infrastructure and extra staff required to install and use this technology?

It will be similar to Hawk-Eye in cricket. There is some investment the stadium operator can make and once done the technology is available at that ground.
It is seriously not prohibitively expensive. You need a couple of special balls with a transmitter embedded, a couple of sensors around the field to catch those transmissions, and some hardware and software to crunch the values.

If stadiums can afford Spidercams or goal-posts that light up on conversions, they can afford this.

Still think the problem here is that the hands are more important than the ball... it really doesn't matter what the ball is doing... forward, back, sideways, up or down... pretty irrelevant... it's all about the hands..

[Blake]

The simplest test is just was it launched forward from the player (taking his speed in to account). Is it launched to go ahead of him or not?

If you were a line judge running at the same speed as the player, does it end up further ahead of you or not.

Problem with that one is we have seen passes that are just momentum, being call forward, if the player takes a tackle a split second after the pass. Since their momentum stops and the ball keeps going. A technical solution would remove that.

Exactly this. Since only the ball is being measured in this solution, what happens to the player or the player's hands are completely irrelevant.
If the program knows which way play is progressing, and the sensor in the ball is pinged frequently enough by enough receivers around the filed to triangulate within mm of accuracy, you should be able to flag a forward pass pretty much every time. That's the easy part.

As I stated last time, it's the change in direction of play that will complicate matters. Things like kicks, turnovers and charge-downs. A trigger of some kind (manual or AI) is needed to "inform" the system:
a. Who is in possession
b. Which direction of play is now "forward"

[Blake]

Still think the problem here is that the hands are more important than the ball... it really doesn't matter what the ball is doing... forward, back, sideways, up or down... pretty irrelevant... it's all about the hands..

It really isn't. I can promise you.
As I said previously, the hand movement guideline is just a tool to hack our brain and prevent us from looking at the ball or the lines on the ground as the ball travels.
It's a quick shorthand to assist referees in making the correct call 98% of the time (when they can see the hands at point of release).

If you can can quantify the forward running velocity of the play (and thus the ball) at time of release of the pass, you are set.
If the player's hands were flat or angled backwards during the release of the pass, the forward velocity cannot increase.
If it does, the pass was forward and the only way that would be possible is if the player's hands were angled forward at point of release.

[Blake]

Not enough. GPS alone won't do it, and it just isn't as accurate as people think. It's not magic!

That’s kind of why I was suggesting corner flag calibration. Rather than using orbiting transmitting satellites.

But it might be that a good solution is with AI processing multi angle video.

What’s the system in the OP using?

The pseudolite solution is (afaik) very, very expensive but yes, more accurate - cm-level at best but still much better than standard GPS. Would it be enough, do we think?

From the OP

To track the flight of the ball, the Steeden is implanted with a microchip that features an accelerometer, a gyroscope, a magnetometer, a pressure sensor and a temperature sensor, many of the technologies found in a normal mobile phone. Then there’s also additional technology that tracks the ball like a radar. There’s also a second method the company uses to track ball flight that the company wants to keep confidential for now.

So multiple relevant techs and then ~secret sauce~, hence me being a bit cautious about this.

Interesting. But it seems like a weird implementation to do it that way around IMO...cramming all the tech into the ball, when you can put the tech outside?
Must have had issues with the transmitter in the ball?

[Mahoney]

Still think the problem here is that the hands are more important than the ball... it really doesn't matter what the ball is doing... forward, back, sideways, up or down... pretty irrelevant... it's all about the hands..

I think the laws use the direction of the hands as a proxy for the direction in which the ball is accelerated by the pass. Which can very much be calculated by the change in velocity of the ball.

(But my solution is still by far the simplest...)

[JM2K6]

Still think the problem here is that the hands are more important than the ball... it really doesn't matter what the ball is doing... forward, back, sideways, up or down... pretty irrelevant... it's all about the hands..

It really isn't. I can promise you.
As I said previously, the hand movement guideline is just a tool to hack our brain and prevent us from looking at the ball or the lines on the ground as the ball travels.
It's a quick shorthand to assist referees in making the correct call 98% of the time (when they can see the hands at point of release).

If you can can quantify the forward running velocity of the play (and thus the ball) at time of release of the pass, you are set.
If the player's hands were flat or angled backwards during the release of the pass, the forward velocity cannot increase.
If it does, the pass was forward and the only way that would be possible is if the player's hands were angled forward at point of release.

This is correct. We had a discussion about this on the Quins board and I was right and Mahoney was wrong (I think that's how it went), but that's essentially what I eventually boiled it down to as a mechanical explanation of a forward pass.

Interesting. But it seems like a weird implementation to do it that way around IMO...cramming all the tech into the ball, when you can put the tech outside?
Must have had issues with the transmitter in the ball?

What tech would you put outside? You mentioned Hawkeye before - it needs a lot of cameras and an unobstructed view of the ball, which it can guarantee in cricket/tennis but not in rugby. Accelerometer, magnetometer etc would only make sense inside the ball.

[Mahoney]

We had a discussion about this on the Quins board and I was right and Mahoney was wrong

Leaving aside the a priori unlikelihood of me being wrong, I don't post on the Quins board. Besides, we seem to be in agreement?

[JM2K6]

We had a discussion about this on the Quins board and I was right and Mahoney was wrong

Leaving aside the a priori unlikelihood of me being wrong, I don't post on the Quins board. Besides, we seem to be in agreement?

Hmm, must have been another boredie I was thinking of then. Minor detail anyway, sounds like you read what I wrote and and stole it for yourself

[Blake]

Interesting. But it seems like a weird implementation to do it that way around IMO...cramming all the tech into the ball, when you can put the tech outside?
Must have had issues with the transmitter in the ball?

What tech would you put outside? You mentioned Hawkeye before - it needs a lot of cameras and an unobstructed view of the ball, which it can guarantee in cricket/tennis but not in rugby. Accelerometer, magnetometer etc would only make sense inside the ball.

I only mentioned Hawk-Eye in the context of how the tech would be paid for i.e some stadiums would have it and some wouldn’t and the stadium would pay for it. Agree that the actual tech would be completely different.

I’m convinced the best implementation would just be one or two transmitters and small power supply in the ball with no sensors whatsover. Then a number of sensors around the field mounted at very precise locations.

The transmitters in the ball would ping the sensors thousands of times per second and by measuring the time between pings you should be able to calculate the exact position and speed of the ball with every ping.

But maybe I’m overlooking something what caused these other guys to go the other direction. Something practical like the fact that you would have to finely calibrate every venue for my design to work while loading the ball with tech is a more complicated engineering challenge but it’s easier to apply at any venue.

[JM2K6]

Interesting. But it seems like a weird implementation to do it that way around IMO...cramming all the tech into the ball, when you can put the tech outside?
Must have had issues with the transmitter in the ball?

What tech would you put outside? You mentioned Hawkeye before - it needs a lot of cameras and an unobstructed view of the ball, which it can guarantee in cricket/tennis but not in rugby. Accelerometer, magnetometer etc would only make sense inside the ball.

I only mentioned Hawk-Eye in the context of how the tech would be paid for i.e some stadiums would have it and some wouldn’t and the stadium would pay for it. Agree that the actual tech would be completely different.

I’m convinced the best implementation would just be one or two transmitters and small power supply in the ball with no sensors whatsover. Then a number of sensors around the field mounted at very precise locations.

The transmitters in the ball would ping the sensors thousands of times per second and by measuring the time between pings you should be able to calculate the exact position and speed of the ball with every ping.

But maybe I’m overlooking something what caused these other guys to go the other direction. Something practical like the fact that you would have to finely calibrate every venue for my design to work while loading the ball with tech is a more complicated engineering challenge but it’s easier to apply at any venue.

Not sure what the power requirements would be in order to transmit that far to so many locations, or the size & weight of it, and the transmitter would have to definitely not move within the ball itself (or break when the ball gets deformed when kicked). The processing required on the other end would be pretty impressive. But it's not impossible. Thing is, they've been trying to do this for a while (e.g. https://www.dailymail.co.uk/sciencetech ... d-try.html which is looking just at tryline stuff) and they don't seem to have cracked it yet - forward passes represent a much higher level of complexity.

[Kawazaki]

Not sure what the power requirements would be in order to transmit that far to so many locations, or the size & weight of it, and the transmitter would have to definitely not move within the ball itself (or break when the ball gets deformed when kicked). The processing required on the other end would be pretty impressive. But it's not impossible. Thing is, they've been trying to do this for a while (e.g. https://www.dailymail.co.uk/sciencetech ... d-try.html which is looking just at tryline stuff) and they don't seem to have cracked it yet - forward passes represent a much higher level of complexity.

I think the locator in the ball is passive and doesn't require power, much like the Recco devices in ski wear. It would be quite easy to suspend one of those inside the ball bladder. However, I've no idea if it could also transmit the orientation of the ball.

[JM2K6]

Not sure what the power requirements would be in order to transmit that far to so many locations, or the size & weight of it, and the transmitter would have to definitely not move within the ball itself (or break when the ball gets deformed when kicked). The processing required on the other end would be pretty impressive. But it's not impossible. Thing is, they've been trying to do this for a while (e.g. https://www.dailymail.co.uk/sciencetech ... d-try.html which is looking just at tryline stuff) and they don't seem to have cracked it yet - forward passes represent a much higher level of complexity.

I think the locator in the ball is passive and doesn't require power, much like the Recco devices in ski wear. It would be quite easy to suspend one of those inside the ball bladder. However, I've no idea if it could also transmit the orientation of the ball.

Blake is suggesting a transmitter in the ball.