Video Review Needs to Happen Now

[Kevin Sevcik]

I'm on board with Sean that I don't think video review is going to significantly reduce complaints unless it's perfect. Whatever circumstances you confine it to, complaints will just move to the boundaries outside of them, and you'll get the new class of complaints about bad video calls. All for the low price of yet another key volunteer and an av system.

Also, I feel I should point out that whatever video replay solution you guys come up with must be implemented at every regional and district. You're planning on enshrining this in the rules, and inconsistency between events seems like just as big a complaint generator as bad calls. I think this means that any solutions that depend on streaming are non starters. Also I suggest you either need a match timer in view or a scoring overlay, or you're going to be missing at the end of auton and during the endgame.

[JeremyLansing]

Originally Posted by gblake

Unless/until we have some experimental results to digest, I think our current posts are only making 11-12 years of "rubble" bounce.

I'm willing to stop bouncing the rubble, and wait for the results.
Blake

This about sums up where I'm at. I cannot wait to see the results of some offseason events that choose to implement this. Until we have some actual data, most of these arguments are speculative. Let's see how replay would work.

[Ryan Dognaux]

Originally Posted by Kevin Sevcik

I think this means that any solutions that depend on streaming are non starters. Also I suggest you either need a match timer in view or a scoring overlay, or you're going to be missing at the end of auton and during the endgame.

The solution we're using doesn't require for you to stream. Streaming and recording are two separate functions that can happen simultaneously or you can just record the matches if you don't have internet access. This is a great point though, not every event has the internet to stream out and any solution would need to be able to operate without it.

The scoring overlay you're talking about is just the real time score graphics that the field provides. We use a cheap Monoprice Video Grabber just to get that part to overlay on top of the GoPro's HD view. The real time graphics don't look as good as everything else, but at least you have them.

Originally Posted by EmileH

1) It costs a lot of money to have the events record the matches.

To record the events that are already being webcasted (which are a majority of the events today), it literally requires 0 more cost to record matches. You already have a PC, you're already feeding video into some piece of software, you can just record that same feed. The low cost setup we use would run you $500 if you already had a decent PC and $1000 if you didn't. For FIRST that's not a lot of money.

[EricH]

Folks, I'm suddenly reminded of something.

When was the first Red Card or Yellow Card issued for actions taken on the field?



Anybody who says 2010, you're right for OFFICIAL events, and wrong otherwise. It was actually IRI 2004. That's right, 5 or so seasons of testing in an offseason... And eventually, it worked.


I think, if video review is going to become a reality (note: I'd rather it didn't), that it should be thoroughly tested. Each event to try it should try something different, or slightly different from other events, and post a detailed whitepaper on setup, rules of use, actual usage, and what calls were changed. As I noted earlier, I'd be willing (if cleared by the planning committee and other interested parties) to apply it at an offseason near me, how it works TBD (I've had some ideas in the past and kind of tossed them around a bit on CD).

[rich2202]

Originally Posted by Ryan Dognaux

Scenario A: The head referee makes what they thought was the correct call based on the evidence at the time, but later finds out the call was actually wrong. The call was match deciding and one alliance goes home denied of an event win.

Scenario B: The head referee makes what they think is the correct call. One alliance uses their once per elimination tournament challenge card requiring the head referee to look at the match with the captains from the alliance to explain his reason for the call. This also gives ample opportunity for the challenging alliance to provide specific rules or Q&A's to support their challenge. A match replay is issued if the head ref confirms & agree with the challenge's claims or the results stand as called if the head referee doesn't seem evidence to overturn the call.

There are definitely times where something like this could have came in handy. Look back to San Diego last year where the incorrect call was made and was admitted to be wrong once the Q&A was produced. Maybe having that challenge would have given them a moment to find the Q&A and take another look at the situation?

My point is it's worth taking a look at and talking about. This could give referees another tool to help them out when making very tough calls. Why wouldn't we want to help our referees out?

What about scenario C: red alliance successfully challenges to get extra points to wint the match. Then blue alliance challenges a different call to get extra points to re-win the match.

[dodar]

Originally Posted by rich2202

What about scenario C: red alliance successfully challenges to get extra points to wint the match. Then blue alliance challenges a different call to get extra points to re-win the match.

If both are valid, and both alliances have challenges to use, then that is perfectly fine.

[Wayne Doenges]

The game is pretty intense for the number of refs they use.
Would it be possible to have separate refs AND scorers?
Just my $.05

[MikLast]

Originally Posted by EricH

I think, if video review is going to become a reality, that it should be thoroughly tested. Each event to try it should try something different, or slightly different from other events, and post a detailed whitepaper on setup, rules of use, actual usage, and what calls were changed. As I noted earlier, I'd be willing (if cleared by the planning committee and other interested parties) to apply it at an offseason near me, how it works TBD (I've had some ideas in the past and kind of tossed them around a bit on CD).

Well i have some good news: This thread is all about offseason events doing just this. http://www.chiefdelphi.com/forums/sh...d.php?t=145690

[Kevin Sevcik]

Originally Posted by EricH

I think, if video review is going to become a reality (note: I'd rather it didn't), that it should be thoroughly tested. Each event to try it should try something different, or slightly different from other events, and post a detailed whitepaper on setup, rules of use, actual usage, and what calls were changed. As I noted earlier, I'd be willing (if cleared by the planning committee and other interested parties) to apply it at an offseason near me, how it works TBD (I've had some ideas in the past and kind of tossed them around a bit on CD).

Don't forgot a survey of participants. Preferably anonymous to maximize the potential negative feedback.

Also, it probably won't be necessary, but y'all probably should go all in on getting video review at every offseason. At some point after some refinement, you're going to want to dump your system on an event that's completely unfamiliar with it to see how bad it's going to be for the first year being run by a crew with minimal training on it.

[Ginger Power]

Originally Posted by Wayne Doenges

Just my $.05

Inflation?

[Sperkowsky]

Hi all!

So, I wanted to wait to post these videos and I think its calm enough to do so at this point.

Sorry for the poor video quality. I had these videos recorded so, I could go back and watch did not know they would be involved in this sort of thing.

QF-1 Was pretty clean

But QF-2 was bad. To start the big 2 things are the fact that our alliance very clearly crossed the cvr twice directly in front of 2 refs and that the red robot hit one of our blue robots during the last 20 second of the match.
If those point values were recorded correctly there would have been 35 points added to the score allowing us to win.
Here is the video for those who would like to see.
https://www.youtube.com/watch?v=qHnJ...ature=youtu.be

QF-3 Had less clear cut issues but, you can see those aforementioned intent to tip fouls. Those pins at those angles deserve more penalty IMO. Even if there was just one more 5pt foul we would have won the match going into semis.
Here is the video for QF-3.
https://youtu.be/m2MWxJEBwpk

Some of the fouls we originally thought happened were not present but, you can clearly see enough in both matches to warrant us a win. 395 has not made CMP since 2011, we have never made it, and team 1546 has not made it since their rookie year in 2006. 395 is the only team that is even competing in a second regional. These are not elite teams who consistently make champs we are talking about here these are small teams who fight hard year after year in hopes of winning and its a shame to see all of that hard work destroyed by some bad calls. We are on the CMP waitlist and if I am not mistaken its possible to still get invited through that but regardless I am going to champs personally so see everyone there just wish my team could be there with me.

[EricH]

I'd like to make a quick commentary here. Just to pick on the linked videos, this is why audience videos probably shouldn't be allowed to be used for review. Zoom took time to adjust, people were in the way (down in front!), and the camera tended to follow only one robot (and, BTW, that last is fully understandable, you want to record your team's performance and I can't say I blame you there!). That being said, if the part that was needed did happen to be clear and unobstructed, it could be worth it to use that.


As far as rules go... Just remember that pins are legal up until 5 seconds (and 6' separation or chasing by the pinned robot), and tipping a robot unintentionally isn't illegal per se but a strategy to tip a robot is. Just something to think about.

[P.J.]

Originally Posted by connoc1

I would like to hear from any referee of Stronghold so far.

No referee is going to touch this issue with a ten foot pole.

Source: Am referee. Check my signature.

[Tristan Lall]

I think it's worth discussing exactly what a successful replay challenge should be able to accomplish, in order to more clearly distinguish between good and bad policies.

It's probably inherently infeasible to unwind a match just because the participants' reactions to a bad call were different than they would have been had the correct call been made. That's worse than judging intent: it's almost complete subjectivity, and trying to speculate about what constitutes a correct outcome is an exercise in conjecture. Instead of expecting to unwind the match, teams should play on, with the understanding that the call on the field might be reversed on appeal. It's up to the teams to choose how they complete the match based on incomplete information, but the fact that in hindsight they should have chosen differently is of no consequence.

By contrast, a replay challenge might be warranted in situations where the nature of a discrete event is unclear, and where that event is supposed to have a defined result when determining the outcome of the match. For example, did the robot complete a game task for which points are supposed to be awarded? The resolution is to either award points or not, and that takes effect at the end of a match, no matter what else happened.

But what about games where score depends on intra-match conditions? Those fall somewhere in between, and probably need to be considered as part of the game design process. One possible resolution is to credit the points mistakenly not awarded, but not unwind the gameplay that resulted from those points not being scored at their proper time. (Is that equitable? I guess it depends on the game.) Another resolution might be to trigger a rematch for certain missed calls: perhaps they're so fundamental to the game that to miss one is to threaten the legitimacy of the event. (In fact, that's the sort of situation where a strong replay rule is beneficial: most of the audience already knows the equitable outcome, so why not give the referees the opportunity to get with the program?)

So ultimately, having a replay system doesn't need to mean that everyone will always get rematches and slow the event to a crawl. Instead, it should be tailored to the situations for which it is usually advantageous (and then applied consistently), and should be made unavailable when it would most often be detrimental.

[Kevin Sevcik]

Originally Posted by Tristan Lall

By contrast, a replay challenge might be warranted in situations where the nature of a discrete event is unclear, and where that event is supposed to have a defined result when determining the outcome of the match. For example, did the robot complete a game task for which points are supposed to be awarded? The resolution is to either award points or not, and that takes effect at the end of a match, no matter what else happened.

By way of example here, you have 2006, where the alliance with the highest autonomous score goes on offense first, which was considered a significant advantage. Or 2004, where knocking a ball off in auton got you your balls 40 seconds sooner.