Finals 1 at Glacier Peak

So I was watching Glacier Peak District finals 1, saw the blue alliance won the first match, got up went to the bathroom and by the time I got back they were announcing they’d changed the outcome of the match. I had missed the explanation as to why, but people in the chat had said they removed the balance bonus for the blue alliance, because 4918 had not climbed but was holding it level. I saw in the score breakdown that indeed the 15 points had been removed, but then I wondered, why hadn’t they removed the 15 points from red? Because they too had a robot, 4911, holding the bar making it level while they hadn’t climbed. Needless to say this concerned me greatly and left me with many questions.
How did they come to the decision to alter the score and outcome? Did they review a video or photographic evidence? Were they just going off of memory? Did they just not remember the red alliance had been in that exact situation as well? Here is the match video from YouTube: https://youtu.be/26XinRoNxf8

12 Likes

Probably.

3 Likes

The rules state that referees are not to review video or photo evidence without exception.

3 Likes

Well that’s what I don’t understand… they’re not allowed to review photos or video, but they’re allowed to review their own possibly faulty or incomplete memories, and change the outcomes of matches based on them? They didn’t review the memories of people who saw that red alliance was in the same situation? They didn’t review the scoreboard showing red was awarded the balance bonus, one hang and one park?

18 Likes

There’s a heated discussion in the Glacier Peak PNW discord server. One of the PNW FTA’s(for the event I assume) stated multiple moments of clarity. One of which was 2910 wasn’t able to resolve the issue of the prior match, when they informed the refs(during awards ceremony). Additionally they apologized for the error and that any complaints should be directed at FIRST Wa and FIRST.

This is my interpretation as always. Many others took time to chime in as well. Especially the alliance captain of Red, 2910.

1 Like

That was the FTA for this event.

I guess I’m just concerned about the review process. Idk how the people in charge came to the realization that they had awarded points wrongly to blue, but didn’t come to the realization they had done the same for red. I just want to everyone to play by the same rules and come up with a way to avoid these mistakes in the future.

2 Likes

Those were my exact concerns during the discussion. The response was an apology and that those direct concerns should be handled by FIRST Wa and FIRST. My solution for that is having teams assist in enforcing. This is in light to a possible resolution of Jack in The Bot or Spartabots that could’ve appeared, if they had known about the issue immediately after the faulty calling. Rather than during awards(they could’ve as alliance captains, addressed the refs for their issue)

2 Likes

They did. I was in the stands and I saw Spartabots talking to the ref immediately after that announcement, but the ref made body gestures like he couldn’t do anything.

1 Like


Week 1 mistakes are bound to happen, it’s unfortunate that this was such a big one. I’m sure they will take this as a learning for future events. The alliance captain of Red attempting to correct the situation is commendable.

2 Likes

Saw this same problem at FMA Hatboro Horsham this weekend too. In QM8 the blue alliance was given the points for the switch being level despite 2729 being parked and touching the switch. I spotted this from the livestream and suggested 423 send someone to the question box to dispute it. The head ref at the event agreed they shouldn’t have gotten the points and that this mistake changed the outcome of the match, but said it’s too late to change anything because they can’t use video review to change a call.* The team understood the ruling and doesn’t blame the head ref, but submitted a NMIR anyway to pass their complaint up the ladder.

It seems strange to me that the FMS would automatically assume that if the switch is within ±8° then the alliance should get the points when there is a very common case for that not to be true. The refs already have an ‘endgame’ page on their scoring tablets to mark hanging/parked/none for each robot. I would think they could easily have a ‘no parked robots touching the switch’ button and require that to be pressed to get the level points, rather than assume that’s the case and make the head ref remember to change it if it isn’t. I imagine that would save a bunch of scoring mistakes over the course of the season.

 
* As a side note, I’m against video review for changing calls that happen during the match (e.g. fouls, number of balls scored, etc). Referees aren’t perfect, and if we debated every split-second decision or missed call we’d be there all day. But for things scored based on their state at the end of the match where referees should reasonably have enough time to get them right (e.g. climbing states, scoring in 2019, etc) video review seems like a no brainer at this point.

2 Likes

I believe that the head referee panel has a level switch override button if the level is inaccurate.

2 Likes

I always thought scoring errors in a match are supposed to result in a replay, not a retroactive change to the match score. I’m recalling a moment in the 2016 NE DCMP, when an autonomous ball was not counted properly - initially volunteers changed the outcome of the match, but after a call with HQ the decision was to replay due to the scoring error rather than retroactively change the match score.

2 Likes

I think this depends upon whether it was field sensor failure or human operator failure. A misentered foul or tech foul would simply just be corrected to change the outcome appropriately. However if a field sensor is acting faulty, it may be difficult to tell to what extend it affected the match. In your example, it would be very difficult to tell whether it was just the one auto ball counted incorrectly or if the issue is bigger than that.

It depends on the year. Sometimes the GDC allows adjustments for missing scores if it can be verified. If I recall, the fuel counter in SteamWorks would be retroactively corrected if there were balls stuck in it.

As the coach and the alliance representative this is somewhat inaccurate, we did not go up with a picture but did try to say that the call should have been applied to both alliances.

What we were told was that it was ruled 4911 was not touching the bar and therefore it was a level climb. Though it appears they were this was the call on the field and it couldn’t be changed.

We would like to thank the referees for their time and work. And would like to congratulate 2910, 4911 and 4173 for their event win, you deserved it!

11 Likes

Ah ok, thanks for the info!

Gracious Professionalism in action. Also, I was really impressed with both 2976 and 2930’s bots at competition. Well done.

3 Likes

I can’t speak to why red was awarded the points, I hadn’t even noticed it until now, however, I might be able to shed some light on why the call was changed for blue.

I had the opportunity to talk to one of the refs involved with scoring the blue climb after the event. They said that what happened was each of the refs each independently scored the blue climb. Both imputed two climbed for 2930 and 2976, but a park for 4918.

Refs do not mark level vs non level, that’s FMS. They also do not handle instances like this where a robot is holding it level, that’s up to the head ref to cancel out. The ref I spoke too said that the head ref simply forgot to cancel the level, and quickly realized after the score was posted, so they reversed it.

Seems like an honest mistake. No video replay was used.

I have no idea about what happened with red, however.

2 Likes

At Durham yesterday there was a match where they updated the score after the fact because of an error…

We also had to replay a Semi-final match there after the scoring system froze mid-match. That sucked, both alliances would have had great scores (and we almost certainly would have won it)