Zone versus man-to-robot refereeing styles

[alicen]

Originally Posted by Andrew Schreiber

Im very aware of this. Hence why I said "UNLIKE Refs". I was merely providing a point of reference to other folks for groups that do use zone scouting and why it was effective for that group. I don't propose refs do this in any way shape or form. Frankly judges miss a ton of stuff but because they are only watching to validate claims made by teams they can afford to without it impacting their job.

my mistake. I misinterpreted what you wrote!
I don't think there is an easy solution to making ref calls more reliable.
At least not with the current number of refs most events are able to get.

Of course, all of these statements can be changed easily, but you'll lose other things. Want an exciting complicated game? - refs will be less reliable
Want reliable refs? - game will be more like Recycle Rush
Want both? - get more volunteers, or be okay with a much simpler game

[philso]

Originally Posted by Caleb Sykes

[/list]Drawbacks include:

• May require more referees. Most events presently have 5 or 6 referees, while this arrangement would require 6 or 7 referees.
  
• Referees would need to move around more, which is often difficult because the area near the field is quite cramped at some venues.
  
• Viewing angles may be sub-optimal. With one ref on each side of each zone, at least one ref will always have a good viewing perspective on anything that happens.
  
• Inputting fouls and crossings could potentially be difficult if 2+ referees need access to the same input panel. However, 2 referees watching 2 robots and inputting into 1 panel could still be superior to the current system, where 1 referee watches 2 robots and inputs into 1 panel.

Originally Posted by Donut

These two pretty much make a change like that a non-starter. I would run into my fellow refs if I had to follow a team around the field, looking at a zone allows you to focus on one area and certain penalties for that area.

Constantly running up and down one side of the field would be difficult enough but if the robot cycling by going from one corner to the opposite corner, it would be almost impossible for the Referee to keep up with it and their view of the robot would be worse than it is now. If the Refs have to go behind the Driver Stations to get to the other side of the field, they will lose sight of the field and could get tripped up by all the activity happening behind the Driver Station (this would have been really bad in 2014 and 2015). It would also take a Referee who was very physically fit to keep running back and forth all day long. This would greatly reduce the pool of potential volunteers.

For this to work, the Referees would have to be stationed on tall chairs (like those at tennis tournaments) so that they have an adequate view of the field without running around.

[Bryan Herbst]

How to best utilize referees is certainly an interesting discussion, and there is no "right" answer.

First it is critical to understand that what works well one year will not work with other years. For example, take the proposal that each ref is assigned to a single robot and apply it to Recycle Rush. Robot A makes a stack of 6 totes plus a recycling container, then 2 minutes later Robot B knocks the stack over. Ref A (assigned to Robot A) properly records the stack. When the stack is knocked over, does Ref A have to remember that the stack was created by "his" robot? Does Ref B just have a button to decrement the score if "his" robot knocks a stack over? What if the stack just falls over on its own? Clearly robot-based ref assignment would have been problematic.

As evanperryg touched on, UX is also a huge concern (and one that the folks at HQ actually think about a lot). Buttons need to be big enough that the refs can easily hit the right one without hunting and pecking. There also needs to be sufficient information on screen to verify the current state of the match.

If refs were to be assigned to individual robots, here's a rough list of things that they would need to see on screen (just during teleop):

• The number and color of the robot they the ref is assigned to (because this changes in every one of the 60+ matches)
• Fouls and technical fouls assigned to that robot
• Representation of each of the 5 defenses- what the defense is, how many times the robot has crossed it
• Scaling/challenging status

If we are operating under the assumption that the refs would also be moving around the field to see "their" robot, you have to multiply this information for each panel each ref might run over to (let's be nice and assume that three refs per side stay on their side of the field, so only this x3). This also makes my first bullet a bit more complicated.

Not saying it's impossible or a bad idea, just that it is complicated.

If anyone is around for setup day at an event, ask your FTA if you can check out the ref panels while they are testing the field. It will give you pretty good perspective on how they currently work.

[Caleb Sykes]

Originally Posted by DonRotolo

Read what EricH wrote, and do that. It's clear you've never volunteered as a referee.

Try being one and then you could have the credibility to comment.

I don't believe that being a referee is a necessary criterion to comment on how referees operate, but apparently you do. That is fine, just be careful that you don't comment on how US presidents operate until you have been a US president.


If I got something drastically wrong in my analysis of how the referees currently operate, I would appreciate someone explaining what I missed.

[alicen]

Originally Posted by Caleb Sykes

I don't believe that being a referee is a necessary criteria to comment on how referees operate, but apparently you do. That is fine, just be careful that you don't comment on how US presidents operate until you have been a US president.


If I got something drastically wrong in my analysis of how the referees currently operate, I would appreciate someone explaining what I missed.

I think what they're referring to is the fact that for things like auto scoring and end game scoring, the ref panels across from each other have to match, that way you don't have doubled scoring or incorrect scoring.
The only thing that multiple refs on a single panel would work for is most fouls.

[alicen]

A couple notes --
- if red 1 pins blue 1, who puts in the foul? red 1 ref or blue 1 ref?
- since refs have been heavily involved in scoring in recent years, how do you propose they do verification checks? (exe: refs have to agree/match on challenges/scales this year for the score to be finalized)
- moving around to follow a robot is the biggest issue I see with this (tablets could be a solution, but they're not exactly comfortable to hold in a useful position for 8+ hours a day)

other than that, I readily agree with changing up the ref style, or maybe creating a position that is the scorekeeper on the field, so refs can focus on penalties and not things like crossings

[Donut]

Originally Posted by Caleb Sykes

• May require more referees. Most events presently have 5 or 6 referees, while this arrangement would require 6 or 7 referees.
  
• Referees would need to move around more, which is often difficult because the area near the field is quite cramped at some venues.

These two pretty much make a change like that a non-starter. I would run into my fellow refs if I had to follow a team around the field, looking at a zone allows you to focus on one area and certain penalties for that area. Also the two events I have been a referee at the last two years had to pull in at least one referee last second (two weeks or less notice) just to get to 6, we'd be more short with 7 needed.

On the note of tablets, I'm concerned about wireless interference given how many comms issues already occur for teams, and a wired tablet would be a tripping hazard. I would however like to see the time sensitivity on the current screen buttons adjusted, as there were times where I and others have attempted to push a button but not had it register. Not sure if FIRST uses the default settings for button sensitivity or has actually tested the timing settings but it is a pretty easy change to make for next year's displays.

[Caleb Sykes]

Originally Posted by Caleb Sykes

[*]Inputting fouls and crossings could potentially be difficult if 2+ referees need access to the same input panel. However, 2 referees watching 2 robots and inputting into 1 panel could still be superior to the current system, where 1 referee watches 2 robots and inputs into 1 panel.[/list]

I thought I would clarify this one since I hadn't worded it the best, and it might be causing some confusion. Here is an example of the situation I was trying to describe:
3 red robots are in the neutral zone, all are in the process of crossing a defense. To my knowledge, with the current "zone" system of refereeing, 2 referees would likely be watching them, and each would have a panel to input the crossings. With the "man-to-robot" system, there will still be 3 robots to watch and 2 panels to input crossings, but there will be 3 referees instead of 2 watching.

[EricH]

Originally Posted by Caleb Sykes

I thought I would clarify this one since I hadn't worded it the best, and it might be causing some confusion. Here is an example of the situation I was trying to describe:
3 red robots are in the neutral zone, all are in the process of crossing a defense. To my knowledge, with the current "zone" system of refereeing, 2 referees would likely be watching them, and each would have a panel to input the crossings. With the "man-to-robot" system, there will still be 3 robots to watch and 2 panels to input crossings, but there will be 3 referees instead of 2 watching.

Just got around to fully reading this. It's actually 1 ref entering the crossings. But in that situation, there'll probably be 4 refs using at least peripheral vision if not focusing. Maybe 3 if the other side of the field has the same situation.