Do you think FIRST should extend the autonomous time

[EricWilliams]

I'll just play Devil's Advocate here:

Originally Posted by Tom Bottiglieri

You can score 75 in auto this year. Most qualification matches do not peak 100 points. This seems plenty important.

I know my game insight and strategy skills are legendary (I predicted a strong goalie bot would be a reasonable strategy this year), but could this not be little bit of confirmation bias? You're not comparing average autonomous to average match scores, and it's hard to say that the average telop scores are what the game was theoretically designed for.

Originally Posted by Andrew Schreiber

No. The fact is most autonomous periods are boring. Many robots don't even move. Having 30 seconds of 5 robots sitting still while 1 moves forward, does some simple task for 5 seconds and then sits still is BORING.

Couldn't you draw the conclusion from this (absolutely agreed upon) sentiment that teams aren't focusing on autonomous mode BECAUSE not enough (perceived) importance is being placed on strong autonomous performance?

[Joe Ross]

Originally Posted by EricWilliams

You're not comparing average autonomous to average match scores, and it's hard to say that the average telop scores are what the game was theoretically designed for.

The average robot scores 9.74 points in Autonomous and 8.22 points in the high goal (and fouls). It also scores 5.85 points from assists and 4.05 points from truss/catch. Alternately, the top 100 robots average 20.87 points from assists, 22.76 from Auto, 12.57 from Truss/Catch and 24.82 from Teleop/Fouls.

[kevin.li.rit]

I think they should have a button that you can flip for a bonus autonomous mode. Flip the switch and you just stay hands off your robot for the rest of the match.

[Andrew Schreiber]

Originally Posted by EricWilliams

Couldn't you draw the conclusion from this (absolutely agreed upon) sentiment that teams aren't focusing on autonomous mode BECAUSE not enough (perceived) importance is being placed on strong autonomous performance?

I'm drawing the conclusion that the average team inhales audibly and is therefore incapable of fielding a robot that moves reliably let alone one that moves autonomously for 30 seconds. It's not a function of focus; it's a function of lack of training, support, and reasonable expectations.

[IronicDeadBird]

Also just consider the time constraints, would you add the auton' time to the game or take it out of the teleop? What would have been nice is to see an auton redemption at the end of each match where robots attempt to position for another shot at auton. Other then that I feel like in order for the change to actually have a significant impact you would be adding far too much time.

[cbale2000]

I would say yes to this idea on two conditions:

1. The match length was increased in proportion to the auton time increase (Don't reduce driver time to accommodate an auton >15 seconds)
2. Allow robots to interact during auton. Some of the most exciting parts of matches from older games were during auton because other teams could attempt to disrupt opponents auton. Granted it would depend on the game, but I think its crucial to making any long auton interesting for spectators.

[Jon Stratis]

The length of autonomous really depends on the game. For a game like this one... there's a limited amount of stuff they could really have us do for autonomous, so a short auto makes sense. Go back to a game like Overdrive, and a longer auto starts to make more sense, as the limit for auto points was really based on how fast/far you could reliably drive in auto.

Given that, I would not say they should extend it for all games, just that it could reasonably be extended for some games, so long as there's plenty to do and it's easy enough to do it that everyone can get it done to one extent or another (like mobility points this year).

[Cory]

FIRST can't even make the field work properly for 10s of auto. Why would we think they could make it work right for 30s?

[DampRobot]

Originally Posted by Andrew Schreiber

I'm drawing the conclusion that the average team inhales audibly and is therefore incapable of fielding a robot that moves reliably let alone one that moves autonomously for 30 seconds. It's not a function of focus; it's a function of lack of training, support, and reasonable expectations.

I don't think it's just that teams are all really bad generally or are afraid of autonomous. It's just really hard to do autonomous. We're a veteran team with very experienced controls mentors and students, and we just spent the last three hours diagnosing why our robot wasn't driving straight in autonomous. If autonomous was "easier" for the average team, it might make sense to extend it. Because most teams have so much difficulty getting autonomous to work, longer auto=more boring matches until you get to very high levels of gameplay. Almost all robots will probably just sit on the field for a while and not do much.

[Andrew Schreiber]

Originally Posted by DampRobot

It's just really hard to do autonomous.

No, it's really not. Doing it reliably, well, accurately are all very hard. But merely doing it is really easy. 125's 2 ball autonomous was written in a hotel room in under 10 minutes. It went 11/12 balls last weekend. Yes, I am relatively experienced, but the core code blocks in that auton were written by a sophomore and junior in HS with very little help.

The core of my statement wasn't that teams don't prioritize auton. It was that they simple can't reliably move. I had a robot start to drive and have their intake fall apart, didn't hit anything, just started decided the field needed some debris. I also had robots sit and spin one way, then the next in a futile attempt to move. And they wouldn't accept help, they insisted everything was fine. For TWO events...

[Gregor]

Originally Posted by Andrew Schreiber

No, it's really not. Doing it reliably, well, accurately are all very hard. But merely doing it is really easy. 125's 2 ball autonomous was written in a hotel room in under 10 minutes. It went 11/12 balls last weekend. Yes, I am relatively experienced, but the core code blocks in that auton were written by a sophomore and junior in HS with very little help.

That's nice.

You've got 15 years experience on a team of 17 years. Yes maybe the code was mostly done by a sophomore and junior, but as they're from the type of team that 125 is, I'm sure they've been well trained and nurtured from the second they showed an interest. Your auto mode manages to drive straight all the way to the low goal. That feat itself is beyond most teams.

It isn't actually that easy.



Not accepting help is a different story.

[Chief Hedgehog]

I do see how this would benefit a game like Ultimate Ascent where the teams could have been allowed to climb for some bonus points - but in a game like Aerial Assist - no.

The programming team is already responsible for so much - it is nice to allow for the Drive team and HP (in games such as Aerial Assist) to develop a strategy. The tele-op portion is so crucial to gamesmanship and excitement, it would be counterproductive to the game itself.

As for the argument that 'Aerial Assist' already has too many problems - I differ. The elimination matches that I have watched (and my team was a part of) have been so exciting! Yes, I could offer up instances where it struggles to compete with games such as UA - but that game was the standard that FRC games will always be measure against.

I believe that Aerial Assist was designed for matches that take place in eliminations - without a doubt. When you get that many great robots and teams on the field and behind the controllers - the intensity rivals Ultimate Ascent.

[AlexD744]

Originally Posted by Andrew Schreiber

No, it's really not.

I think the hardest part about autonomous is getting a robot that is designed to do autonomous well. Many robots are simply not designed to do a two ball auto (lack of good pickup, lack of ball control). Other robots are inherently squirly and even driving straight can be challenging without some fairly sophisticated feedback from a gyro/encoders and decent time to tune. (whereas the person behind the sticks learns to correct fairly quickly)

Compound that with inexperienced coders making a mess out of the code and debug time can increase exponentially, and since the coders are inexperienced, things won't work the first time.

Again, these are things that come with experience. It's the little decisions early that make the difference between coding a two-ball in 10 minutes or spending 2 days debugging your mobility points.

[Laaba 80]

Originally Posted by AlexD744

I think the hardest part about autonomous is getting a robot that is designed to do autonomous well. Many robots are simply not designed to do a two ball auto (lack of good pickup, lack of ball control). Other robots are inherently squirly and even driving straight can be challenging without some fairly sophisticated feedback from a gyro/encoders and decent time to tune. (whereas the person behind the sticks learns to correct fairly quickly)

Compound that with inexperienced coders making a mess out of the code and debug time can increase exponentially, and since the coders are inexperienced, things won't work the first time.

Again, these are things that come with experience. It's the little decisions early that make the difference between coding a two-ball in 10 minutes or spending 2 days debugging your mobility points.

100% agree with this.

I used to wish the autonomous period would be extended to 20 seconds, but I am pretty happy with 10 or 15. It keeps the potential boring time to a minimum, and adds excitement when an elite team finishes with no time to spare. I wish they would allow for "buzzer beater" shots in auto though, instead of only scoring pieces that passed through the goal in auto. This would make it easier on refs, and to this day watching 254 hang their 2nd ubertube during the Galileo elims was one of the most exciting things I've seen in FRC.

[Andrew Schreiber]

Originally Posted by Gregor

That's nice.

You've got 15 years experience on a team of 17 years. Yes maybe the code was mostly done by a sophomore and junior, but as they're from the type of team that 125 is, I'm sure they've been well trained and nurtured from the second they showed an interest. Your auto mode manages to drive straight all the way to the low goal. That feat itself is beyond most teams.

It isn't actually that easy.



Not accepting help is a different story.

Then steal it. https://github.com/FRC125/NU14/blob/...Train.java#L64 Our code has, and as long as I'm around will always be, open sourced from the moment of kickoff. I can't promise it to be readable, good, or proper all the time (these are HS kids) but I can promise it'll always be available.

Or, as every team in New England is welcome to do, come in and ask for help (I'm only there on weekends). Or find us at an event and ask for help. Or PM me here.