Video Review Needs to Happen Now

[JohnFogarty]

You'd like me to go through every part of your argument? I have loads of time today.

Originally Posted by Lil' Lavery

Video review opens a can of worms that should be left closed. Beyond simple time delays, it poses....what level of evidence is needed, what is the timing/procedure for calling video reviews*, how many video reviews are allowed

Time delays are a valid concern. I'll use the Orlando Regional as a benchmark for a lot of things I'll talk about here since that was my most recent experience. We were 1.5 hours behind schedule pretty much all the time. Taking up the time of normal refs with this new video system I'm proposing would not work. Yet, I see the solution as simple and I'll provide my solution.

In qualifications if there is a disputed match where a team believes there was something not scored correctly we could have an additional referee who's entire job it was is to review video to sort these problems out. You don't even have to take the time of the normal match refs to do this sort of after-match verification. The video ref could take a look at the camera view footage to determine if the appropriate call was made.

The review time for videos would be kept short if a designated video review ref could not find indisputable evidence that the call was botched then there would be no changes.

In eliminations where the match scores are in my opinion even more critical to maintaining the quality of the event you can follow a similar procedure. Give at max 5 minutes to determine the call.

Originally Posted by Lil' Lavery

concerns regarding what source of video reviews should be allowed, how many videos should be allowed

As I said previously. The only video that would be reviewed would be that from an official FIRST mandated system.

Originally Posted by Lil' Lavery

A simple overhead camera? I'm assuming you mean similar to the fisheye cameras used by FiM/MAR/Indiana. Because if you mean a true overhead camera, that's simply not a realistic option at the vast majority of FRC venues (and certainly not an option without additional rigging costs on the venue side). And if you mean some other static camera (or even a true overhead), there are still plenty of disputes that will not be able to solve. Was a robot's wheel touching the outerworks at the end of autonomous, or just their bumper overhanging? Did their mechanism extend more than 15" beyond the frame perimeter? Did their robot fully stop contacting the drawbridge door momentarily? What was the game clock at that second during the overhead view? Heck, I'm watching the Tippecanoe stream right now, and there are still quite literal blind spots behind easy drawbridge/portcullis (as well as smaller ones behind the towers). No single camera is going to solve all FRC disputes.


e; To demonstrate my point, I just captured this from the NYC webstream. While the lack of quality is due to stream compression, the general point still stands. This is what typically constitutes an overhead view in FRC. Did the robot in the orange circle cross the defense? Commit a penalty?

A overhead camera like this one. https://youtu.be/PNs40CrPWUk?t=10s
A solution would have to be found. FIRST provides the field, the ref system, etc. This would have to become part of it.

It would very obvious that the camera would have to provide a high enough resolution video of the entire field to be validly able to determine calls.

Some of those calls that you just mentioned can't even be called consistently by the refs with their own eyes during the course of a match let alone an event. (i.e 15" perimeter rule, crossings, etc.) If you watched that video you'd even see that crossings were not being counted correctly even over the simple defenses.

Blind spots are valid and just like in football where sometimes a call can't be made definitively even with camera angles a call would just have to be left to stand. HOWEVER, if it can be proven with a simple system like the overhead camera like I am proposing than that alone is a drastic improvement.