[FRC Blog] Replays and Montreal Finals

Posted on the FRC Blog on March 6, 2018: Replays and Montreal Finals | FIRST

Replays and Montreal Finals
Written by Frank Merrick.

Over the weekend, I had a decision to make about a replay based on a unique match situation.

I received a call about a finals match at the Festival de Robotique - Montreal Regional. After the match, the scores continued to increment even though the match was over. Clearly, there was an arena fault (or in this year’s parlance “Arcade Fault”). The question was “replay or no replay?” Usually this question is handled at the events themselves, but sometimes the question gets escalated, as in this case.

The replay rules are in Section 10.8 of the manual. Three conditions must be met before a replay takes place:

  1. There must be an arcade fault
  2. The arcade fault must have affected the outcome of the match
  3. The affected alliance must want a replay

While it’s common in our community, and even sometimes among staffers and key volunteers, to use the term ‘arena fault’ as a stand-in for ‘replay,’ the game manual itself makes a clear distinction. I’m sure I’ve done it myself sometimes. It can be tempting shorthand.

In the Montreal situation, condition one was obviously met. The field should not have continued to add scores. However, my understanding at the time was that there were no indications that any scoring issues were occurring before the end of the match. Field personnel did not notice any issues, and we did not have any information from teams suggesting there was a problem. Based on the belief at the time, that the arena fault was restricted to after the match was over, I did not believe replay condition two had been met.

In addition, using video of scores displayed at t=0, we felt we could clearly determine that the Red Alliance had won the match and by a substantial margin. The rules say video will not be reviewed at the event, and it wasn’t. Just like other high school sports, the call on the field is the call on the field. We don’t review video emailed to us by teams or at the request of local FIRST management. However, under extreme circumstances, like this one, when we are contacted about something very unusual, video can sometimes help provide context on information being reported in the field. Video in this case was used very narrowly, only to determine the score showing at the end of the match.

Based on all the information I had at the time, I gave the instruction that there be no replay, and that the Red Alliance be declared the winner.

As we looked more into this situation after the event though, we have come to realize that there were additional issues with scoring during the match, which we think could have affected the outcome.

Because of the additional issues found, we can’t say with certainty what the score should have been at t=0, so you won’t see any update to the match data or final score on the FRC Events page.

I believed I was making the fairest decision possible at the time, and I apologize to our community and especially to the teams involved in the Montreal finals, that we couldn’t provide a clean match for you.

However, the call will stand. The declared winners of the event are the winners, and their victory will remain in the record books. Fortunately, all six teams in the finals matches earned their way to FIRST Championship at the event, which hopefully is some consolation.

Also, we are working to understand what caused this issue in the first place. This condition does seem rare. However, we are instructing our FTAs that if they see this condition again, they should assume that it is an arena fault and likely affected match outcome, but that they should contact HQ support. Assuming the HQ team agrees the outcome was affected and the affected alliance wants a replay, a replay would then occur.

I’m sorry again for this problem.

Frank

I’m glad we finally got a official response for this, and I fully believe that Frank and FIRST staff did the best they could in this situation. However, I am curious; does anyone know what the other scoring issues that occurred in this match were?

I appreciate the transparency, but I’m also worried about the implications of there being other, potentially unnoticed scoring issues. Also, no mention of the fact that this error (apparently) happened at other events.

Utah QF1-1 (or maybe 2) was replayed for this same reason. I’m interested in understanding why it was replayed at Utah and not at Montreal.

I saw this issue on a webcast on Saturday. I think it might have been Utah but I can’t say for certain. The score kept counting for about 3 minutes after the match had ended.

Utah QF1-2 has video evidence of something similar. I say “apparently” because I can’t definitively say it was the same bug as Montreal, but it sure looked similar.

It happened in a qualification match at Great Northern on Saturday morning (Match 79, I believe), and the decision to replay it was made on-site by the Head Ref and FTA.

Correct, this happened at Utah QF1-2. A brief reaction video from the ground level: https://photos.app.goo.gl/Tt7FKxASqGNDv94B3

-Mike

One of our qualifying matches (79) at the Great North Regional was replayed due to an Arcade fault. Our Boost button was not working and the score continued to increase after t=0. The FTAs contacted headquarters and were told to replay the match. (according to Jon the LRI) eventually decided to replay the match.

We replayed the match and had the same outcome. At the time we just assumed the boost button wasn’t working, but I agree with briansmist’s comment of not being sure there weren’t other matches with issues that went unnoticed.

Presumably because the local field staff felt it met the criteria for a replay and did not see the need to escalate to HQ.

I like that FIRST admitted the mistake, but I would have really liked them to say that this is the final time FIRST HQ would be reviewing match video/images. Not only did they get it wrong this time, but now teams know it is a possibility so they will plead with Refs to send video/images to FIRST HQ for review.

I very much disagree with this. I know of multiple occasions throughout the years where the event would have made the wrong call, but a phone call to HQ corrected the situation, sometimes in major, event defining ways. The ability to contact an external arbiter is important for correcting these issues and ensuring some degree of fairness across events, even if sometimes mistakes are made. It’s not like it was a secret anyway.

I totally agree with contacting HQ to make tough decisions, but think using video replay just opens up a whole new can of worms.

To quote:
“In addition, using video of scores displayed at t=0, we felt we could clearly determine that the Red Alliance had won the match and by a substantial margin. The rules say video will not be reviewed at the event, and it wasn’t. Just like other high school sports, the call on the field is the call on the field. We don’t review video emailed to us by teams or at the request of local FIRST management. However, under extreme circumstances, like this one, when we are contacted about something very unusual, video can sometimes help provide context on information being reported in the field. Video in this case was used very narrowly, only to determine the score showing at the end of the match.”

The implementation of the replay rule is correct, and the result feels right, i.e. it worked as expected (without knowing the scoreboard was counting incorrectly during the match). Double points for Gryffindor.

The explanation about using video in extreme circumstances… not thrilled with that. Understandably, it makes sense to see what the score was at the end of the match. But, narrowly using the video to see what the scoreboard read at one point in time, without investigating HOW the score got to that point, is an inadequate use of resources. To discover that there may have indeed been a scoring malfunction during the match, no matter how small, after already making the decision to use video evidence despite all historical precedence dictating otherwise, is disappointing. Give me back my points, Gryffindor.

Happy this didn’t make or break any team’s qualification for Worlds, and I appreciate that this decision needs to be made hastily, but breaking tradition hastily without precedent is poor management. Hope it’s a lesson learned.

This was Finals 3. (Oh wait, Montreal. Finale 3.) There is no other match to play, except the most minuscule chance of an Overtime match or replay. The entire venue is there sitting around, the event is effectively stalled, and you can only do the Macarena so many times. Sometimes, you just have to triage it (What was the score at T=0? Was anything noticeably wrong during the match?), make a determination (and I do think the buck stopped at the right person in this case), and do the heavier investigative work when thousands of people aren’t staring at you.

Would agree. Contact HQ, but video review is a new scenario from on high is new territory.

From a competitors standpoint, its nice to hear that HQ reviewed it using all means available and came to a conclusion that was carried out.

That being said, it will be interesting to see how this plays out this year since the door was opened. I’d put money down that at least one alliance will sit arguing in the question box demanding that HQ review video of a critical call that is season ending.

Its not a door I want to see get opened if its not accessible or consistently implemented which this blog post doesn’t address.

I understand that the intentions were good. I appreciate that, so I am satisfied that best efforts were made.

But I’m guessing teams have appealed to refs hundreds, if not thousands of times with footage that would clearly show that there was a field error (signal lights blinking, wrong set-up, etc, etc.). And despite the certainty of what is shown in each of those videos, those videos categorically could not be used as evidence.

I don’t really understand why this situation is extreme. If nobody for sure knew that nothing went wrong with the scoring system in that match, then it should be replayed.

If the decision not to replay was made because “well, the red alliance put up a hell of a lot more cubes than the other alliance, and we just need a snapshot with zero seconds on the clock,” then I am not satisfied with that. The process, in my opinion, failed. (And the reasons why i would be dissatisfied with that are enough for its own thread).

Sometimes best efforts don’t yield the best results. In this case, the result is not the outcome of the match, but the precedent we have set for defining extreme circumstances in the future. And I hope it’s a lesson learned.

I think this situation is distinctly different from watching video replays of match play. The video was just used as a means of determining what the field was reporting to the scoreboard at the time, and was not used to, say, re-interpret a referee’s call of a live event or anything like that. Imagine the scoring software kept a text log of what it put on the screen - this is no worse than reviewing that.

I suppose I could get on board with this explanation.

What’s that? Video replay used to validate something that happened during an event? Hmm… interesting.

Someone should page GBlake and get his thoughts on this :smiley:

It’s cool that we’re almost to the level where this could easily be implemented in situations like this - does every event have a FirstInspires Twitch cast now? If so, what are we waiting for?