[BostonCard]

OK, here’s the situation:

Western Michigan up 12-7 on Michigan; Michigan tries for a Hail Mary that goes out the end zone as time seemingly expires.  “After further review” the refs add one second back to the clock, Michigan tries again, scores a touchdown and wins 13-12.

https://x.com/awfulannouncing/status/209...86880?s=61

Honestly, I find it pretty indefensible, but perhaps I’m missing something in the rules.

BC

[BostonCard]

OK, here’s the situation:

Western Michigan up 12-7 on Michigan; Michigan tries for a Hail Mary that goes out the end zone as time seemingly expires.  “After further review” the refs add one second back to the clock, Michigan tries again, scores a touchdown and wins 13-12.

https://x.com/awfulannouncing/status/209...86880?s=61

Honestly, I find it pretty indefensible, but perhaps I’m missing something in the rules.

BC

[BostonCard]

Here’s the explanation:

https://x.com/bigten/status/2096460445213217097?s=61

BC

[BostonCard]

Here’s the explanation:

https://x.com/bigten/status/2096460445213217097?s=61

BC

[M T]

Here’s the explanation:

https://x.com/bigten/status/2096460445213217097?s=61

BC

I can believe that the time shown in a scorebox or overlaid on the TV is not what the game clock shows.  I often see that in WBB games where you see both the physical clock and the time put on the screen by the broadcaster.  

But in this case, both the Big Ten review system and the broadcast claim to be showing the stadium clock, each also showing the play.
The Big Ten indicates the two separate cameras were synced some 4 or 5 hours earlier.   The broadcaster is using (live) video processing to insert the clock into the broadcast feed.

Eighty-five minutes prior to the game, all broadcast cameras used for the replay system, including the camera on the game clock, are synced using a strobe system to ensure all cameras are aligned.

That doesn't clarify why what they saw are "aligned" (I would say "synced") at the end of the game. I sorta suspect that they are feeding each camera to some digital recording device, and each view we see at the end shows the same position (frame number) in the two recordings. But what if a few frames got lost during transmission to the recording device?

For the inset, I could imagine that the inset image is a frame behind or maybe ahead due to processing the two into one image, but I don't know.  (But in this case, the Big Ten seems to be asserting that the inset clock was several frames ahead of the main image.)  Certainly it would be easy to demonstrate the accuracy of the inset by turning the main camera to the stadium clock while the inset is also showing the ticking clock.

Likewise, the replay system could also view a ticking clock after the end of (or during) the game with all cameras to confirm whether or not there is a discrepancy.

I would have more confidence in the Big Ten review system if all the cameras had a common physical sync signal inserting a frame count to all cameras, and the frame number was shown in all views.  Indeed, why not insert the game clock time onto every frame as meta information?  Or, almost equivalently, have each frame contain the GPS time in, say, microseconds.

Heck, the recording system should be able to record the time the signal arrives from each camera.  But notice that is vulnerable to errors.  I am sure all this is done over a network to the recording system and there are delays, out-of-order arrivals, and lost packets or retransmits that can happen on such systems.  (Such delays also apply to the broadcast main image and inset image.)

I wouldn't be surprised if some people in the stadium recorded it with their phones so that the stadium clock and the touch were both in the same view.  It will be interesting if someone comes out with such a view.  But considering those in the stadium are almost entirely Michigan fans, I doubt they will look at it closely.

I certainly expect someone (ESPN?) will do a detailed error analysis of each system to understand how apparent errors crept into at least one of them.  Annotating each frame with its GPS time as the frame is processed by the camera is probably the best way to avoid such errors.

If I were a betting man, I'd bet that the broadcaster was more likely to have smaller errors.  (A lot can happen in 5 hours.)

[M T]

Here’s the explanation:

https://x.com/bigten/status/2096460445213217097?s=61

BC

I can believe that the time shown in a scorebox or overlaid on the TV is not what the game clock shows.  I often see that in WBB games where you see both the physical clock and the time put on the screen by the broadcaster.  

But in this case, both the Big Ten review system and the broadcast claim to be showing the stadium clock, each also showing the play.
The Big Ten indicates the two separate cameras were synced some 4 or 5 hours earlier.   The broadcaster is using (live) video processing to insert the clock into the broadcast feed.

Eighty-five minutes prior to the game, all broadcast cameras used for the replay system, including the camera on the game clock, are synced using a strobe system to ensure all cameras are aligned.

That doesn't clarify why what they saw are "aligned" (I would say "synced") at the end of the game. I sorta suspect that they are feeding each camera to some digital recording device, and each view we see at the end shows the same position (frame number) in the two recordings. But what if a few frames got lost during transmission to the recording device?

For the inset, I could imagine that the inset image is a frame behind or maybe ahead due to processing the two into one image, but I don't know.  (But in this case, the Big Ten seems to be asserting that the inset clock was several frames ahead of the main image.)  Certainly it would be easy to demonstrate the accuracy of the inset by turning the main camera to the stadium clock while the inset is also showing the ticking clock.

Likewise, the replay system could also view a ticking clock after the end of (or during) the game with all cameras to confirm whether or not there is a discrepancy.

I would have more confidence in the Big Ten review system if all the cameras had a common physical sync signal inserting a frame count to all cameras, and the frame number was shown in all views.  Indeed, why not insert the game clock time onto every frame as meta information?  Or, almost equivalently, have each frame contain the GPS time in, say, microseconds.

Heck, the recording system should be able to record the time the signal arrives from each camera.  But notice that is vulnerable to errors.  I am sure all this is done over a network to the recording system and there are delays, out-of-order arrivals, and lost packets or retransmits that can happen on such systems.  (Such delays also apply to the broadcast main image and inset image.)

I wouldn't be surprised if some people in the stadium recorded it with their phones so that the stadium clock and the touch were both in the same view.  It will be interesting if someone comes out with such a view.  But considering those in the stadium are almost entirely Michigan fans, I doubt they will look at it closely.

I certainly expect someone (ESPN?) will do a detailed error analysis of each system to understand how apparent errors crept into at least one of them.  Annotating each frame with its GPS time as the frame is processed by the camera is probably the best way to avoid such errors.

If I were a betting man, I'd bet that the broadcaster was more likely to have smaller errors.  (A lot can happen in 5 hours.)