Team Update #18

[JB987]

Originally Posted by Lil' Lavery

I didn't really watch many webcasting this weekend, but from the bit of Bayou I saw, the referees were still using their judgement to call the minibot races. There were many false positives, though I only recall one instance of a minibot not triggering at the top.

So...local refs under the direction of local head refs disregard the update and deal with the reality of imperfect triggers? Let's hear it for common sense!

[AdamHeard]

Originally Posted by JB987

So...local refs under the direction of local head refs disregard the update and deal with the reality of imperfect triggers? Let's hear it for common sense!

I love it!

[MrForbes]

Why bother having rules?

[Michael Corsetto]

Originally Posted by JB987

So...local refs under the direction of local head refs disregard the update and deal with the reality of imperfect triggers? Let's hear it for common sense!

Wish they followed this path at Sacramento, we clearly hit the top with 1868's minibot (1.2 second climb) in our last finals match and the tower didn't trigger. Didn't change the match results, but I'm trying to imagine the outrage when a tower doesn't trigger on Einstein...

[billbo911]

Originally Posted by Michael Corsetto

Wish they followed this path at Sacramento, we clearly hit the top with 1868's minibot (1.2 second climb) in our last finals match and the tower didn't trigger. Didn't change the match results, but I'm trying to imagine the outrage when a tower doesn't trigger on Einstein...

This is interesting as I never saw a single time at Sacramento where the towers missed a trigger. Now I must admit, I always focussed my attention on our minibot, so I might have missed yours. Ours is a bit of a tugboat, but it always triggers the tower, and even won several races.

By the way, you guys did a great job in Sacramento and deserved the win. Congratulations!

[Tristan Lall]

So, what's the objective of the minibot race, anyway? To trigger the tower first, or to reach the top first?*

Originally Posted by JB987

So...local refs under the direction of local head refs disregard the update and deal with the reality of imperfect triggers? Let's hear it for common sense!

I know what the rules say, and that isn't it.

If the object of the race was to get there first, then sure, the referees would be totally right to award the race to the first minibot to arrive, irrespective of what the sensor did. But by defining the race in terms of the sensor, and then making an update to emphasize that point, FIRST has explicitly included the mechanical and electrical characteristics of the tower in the problem of winning the race.

Judging it based on sight—but only when the tower fails to trigger on what appeared to be a sufficient impact—fails to distinguish true negatives from false negatives. (Are the referees making the calls fully aware of the possibility and likelihood of a true negative?)

Incidentally, one could argue that the likelihood of a true negative is remote enough that they can all be disregarded—but I haven't seen that earnestly proposed, and frankly, if FIRST wanted that to be the case, they shouldn't have gone to all the effort to define the race in terms of the sensor. By taking pains to define it as they have, I think their thought process was that by removing human judgment from the outcome, they would have a lower rate of false outcomes than otherwise—but to achieve that accuracy, they had to make the sensor response a factor in the definition.

As for the "common sense" argument, I think our preconceptions about how a race is decided colour our opinions. It may well have been a better idea to conform to those expectations instead of crafting a rule that is slightly more complicated—but the rules are supposed to be the same for everyone, and by trying to correct this on a sporadic and unsystematic basis, I'm concerned that we're letting one element of equity infringe on another.

*Hints: <G67> and the definitions of trigger and minibot race. Why would they write that, but intend something else?

[Adam Freeman]

We have a decently fast minibot (~1.5 sec climb rate). It was 14/15 triggering the target at West Michigan. It triggered the sensors on all (4) towers throughout the weekend.

In SF1-1 we launched the minibot on the Red left side tower, it "won" the race but did not trigger the sensor. We were told that the ref saw that it hit first, but since we had won the match any way they were going to "trust the field".

I was not concerned about getting the actual score corrected, but more so about the next matches and how things would be scored if it didn't register again. We were told that the refs have the power to over rule the sensors if needed...but we were not specifically told that they would manually score our minibot if it didn't register again.

Our minibot never went up that tower again, but 2054's registered every time on that side.

This tower had the light knocked off it at one point on Friday and had not registered a couple other minibots at other times on Saturday.

We have a non-random pattern that our minibot is sufficiently designed to trigger the sensors. The field has a non-random pattern that their sensors are functioning correctly. There is some interaction between the tower and the minibot that in certain instances the variation in their performances line up to produce a false negative.

Is it FIRST, the team, or the refs responsibility to handle the < 5% chance that these variable line up and the sensors don't trigger?

We may attempt to make our minibot robust to this issue, but I am fine with leaving the instances where the match outcome would be effected in the hands of the refs...as long as the benefit of the doubt goes to the team.

[Mike Copioli]

Originally Posted by Chris Hibner

An update from Detroit:

All of our minibot runs triggered the tower. Everyone that I saw triggered the tower except one:308's minibot failed to trigger once and it cost them the match. I'm not sure if it hit the bolts, but it hit the plate with more than adequate force. They have a very reliable minibot that triggered every other time they ran it.

I think I may have saw one false trigger, but I can't confirm.

Chris, I spoke with a team member of 308 after that match. He stated that the tower was disabled due to early deployment. Anyone from 308 can correct me on this if I am wrong.

Our towers triggered every time we successfully deployed at Detroit. Minibot ~2.5lbs or .025 Karthiks at ~ 1.2 - 1.6 seconds.

BTW it was great working with you Chris, I will save my additional comments for the detroit thread.

[jspatz1]

Originally Posted by Tristan Lall

But by defining the race in terms of the sensor, and then making an update to emphasize that point, FIRST has explicitly included the mechanical and electrical characteristics of the tower in the problem of winning the race.

Many feel the problem is they did not also PROVIDE the mechanical and electrical characteristics of the tower. As far as I am aware the length of signal required to register with the FMS is still a mystery.

[MrForbes]

Originally Posted by Adam Freeman

Is it FIRST, the team, or the refs responsibility to handle the < 5% chance that these variable line up and the sensors don't trigger?

We may attempt to make our minibot robust to this issue, but I am fine with leaving the instances where the match outcome would be effected in the hands of the refs...as long as the benefit of the doubt goes to the team.

Which team? the team that failed to TRIGGER the TARGET, or the teams on the opposing alliance?

Come on now....

[Adam Freeman]

Originally Posted by squirrel

Which team? the team that failed to TRIGGER the TARGET, or the teams on the opposing alliance?

Come on now....

I was referring to the team that failed to trigger the target.

I would not want to win a match, District Championship, State Championship, or World Championship because my opposition built a minibot that was faster than mine, beat it up the pole, hit the target with the specified FORCE, but the sensors did not TRIGGER the target.

Especially if their minibot has an approaching 100% success rate and the failed TRIGGER is the result of some interation of unknown variables.

Seems to me checking to see if minibots trigger the target could be part of inspections. If you get the ok at inspections, then it is assumed that the minibot is designed correctly.

The refs would have the ultimate power to either believe or over rule the sensors. If a sensor does not trigger, but the team has passed minibot inspection (including triggering), the refs can determine the winner and assign the points accordingly. If it is too close to call, then the match will be replayed due to a field fault.

Its already bad enough that we spent 6 weeks designing a HOSTBOT that can achieve all the tasks of this game at a very high level, only to have it marginalized by a minibot that every team can build in one day.

[MrForbes]

Originally Posted by Adam Freeman

I would not want to win a match, District Championship, State Championship, or World Championship because my opposition built a minibot that was faster than mine, beat it up the pole, hit the target with the specified FORCE, but the sensors did not TRIGGER the target.

FIELD – .....

FRAME PERIMETER – ....

I don't see an official definition of FORCE in the rules.

I would be fine winning a match because my opposition built a minibot that beat ours up the pole, but failed to TRIGGER the TARGET, which is what the rules require it to do. The rules do not make any exceptions.

I guess the thing to do is ask the head ref at our next regional whether they will be calling minibot races based on the rules, or something else.

[Matt Krass]

I'm torn on this one.

On one hand, I believe the intent of the minibot race was to get their first (it's a race after all, it's displayed as such on the scoring metric) and I'm glad to see that refs stepped in when the technology couldn't.

But I'm also frustrated at the refs changing the rules again, while I don't agree with this team update, and I think it's a bad idea, if that's the rule, and that's what every team started working towards, it should be enforced as such. This particular dilemma is precisely why I thought it was a bad idea to throw all their weight behind the system like this, I don't think it's ready yet. Of course, we have no way of knowing if it's ready since we're still working on vague definitions of the 'black magic' actually required to trip the sensor. For all we know it's actually working 100% perfectly, but I suspect that isn't the case.

Matt

[MrForbes]

Originally Posted by Matt Krass

Of course, we have no way of knowing if it's ready since we're still working on vague definitions of the 'black magic' actually required to trip the sensor. For all we know it's actually working 100% perfectly, but I suspect that isn't the case.

I talked to someone who should know at AZ, and I get the feeling that it is working 100%.

I'm also eagerly awaiting a Q&A response....

[Lil' Lavery]

Originally Posted by squirrel

I don't see an official definition of FORCE in the rules.

I would be fine winning a match because my opposition built a minibot that beat ours up the pole, but failed to TRIGGER the TARGET, which is what the rules require it to do. The rules do not make any exceptions.

I guess the thing to do is ask the head ref at our next regional whether they will be calling minibot races based on the rules, or something else.

Okay, let's just bash at the base of the towers until it triggers. No need to actually race up the pole. So long as we're hitting the pole with our minibot, the minibot is pushing on the tower, which causes the tower to trigger, so it should count.

Fine with me.

Originally Posted by squirrel

I talked to someone who should know at AZ, and I get the feeling that it is working 100%.

I'm also eagerly awaiting a Q&A response....

I can guarantee you from watching the Bayou webcast, it is not working 100%. Not even close.