How many balls will you shoot during autonomous?

[Poseidon5817]

After seeing discussion on the low bar poll thread about how many teams will be able to do a two-ball auto, I wanted to see what teams are thinking without driving that thread farther off topic. How many balls is your team planning to score during autonomous per match at your first event? And don't worry, I made the poll results private, so you can tell the truth without fear of sharing secrets. :D

[Lil' Lavery]

I would be beyond thrilled with 18. ;)

[RoboChair]

I would be beyond thrilled with 18. ;)

How are you planning to teleport balls? Or your robot? I was pretty sure that the battery couldn't supply the required power draw.

[cait.schroeder]

Im doubting that there will be a lot of shooting during autonomous especially high goal.

[Lil' Lavery]

How are you planning to teleport balls? Or your robot? I was pretty sure that the battery couldn't supply the required power draw.

OP asked how many balls we would score in autonomous at our first event. I'll leave the rest as an exercise for the reader. ;)

[jwfoss]

OP asked how many balls we would score in autonomous at our first event. I'll leave the rest as an exercise for the reader. ;)

Someone wants to make it to the finals, but not sweep the elimination rounds?

[MrJohnston]

Im doubting that there will be a lot of shooting during autonomous especially high goal.

I think there will be lots of shooting.... Maybe not much scoring, but definitely lots of shooting....

We'll definitely be shooting one at the high goal...Once several routines are dialed in (different positions, different defenses) and our vision-tracking is consistent, we'll be exploring more interesting options that might involve multiple boulders.... Of course, we need to have more robot built than we have now to even think about it.

[GaryVoshol]

Sean's in the District system. He wants 12 quals, 2 quarters, 2 semis and 2 final matches. Shooting once in each.

Not sure how many balls everyone will shoot. Maybe we should count BOULDERS? ::rtm::

[Caleb Sykes]

"Shoot" and "score" are two very different things. I believe that many teams will shoot 1 ball in auto. Not many of those teams will be scoring though.

[pandamonium]

I really think that the 2 ball auto is not as difficult as everyone thinks.

[Poseidon5817]

"Shoot" and "score" are two very different things. I believe that many teams will shoot 1 ball in auto. Not many of those teams will be scoring though.

The poll was meant to measure how many balls a team plans on scoring. I'm sure a team won't be planning to shoot 2 balls in auto but only score 1 going into their first event, they'll be planning to make both, even though that might not actually happen.

[GeeTwo]

The way we see it, if you can't score on a high goal in autonomous, you can't score on a high goal. There are just too many obstructions to aim from your own castle. It may be possible to do this using a teleop camera, but by the time you get enough resolution to take aim, your frame rate over the network is terrible, at least for a regular video stream.

[Kevin Sevcik]

I really think that the 2 ball auto is not as difficult as everyone thinks.How would you compare the difficulty of a penalty free 2 ball auto this year to a 7-disc auto in 2013? Keeping in mind that in 2013, there weren't any obstacles to traverse. It was just shoot 3, drive forward and intake, shoot 4.

[Tottanka]

According to G13 you may not contact the Midline's volume. If that's where the boulders are located, how exactly are you planning on collecting more balls during auto, without any penalties?

[MrJohnston]

According to G13 you may not contact the Midline's volume. If that's where the boulders are located, how exactly are you planning on collecting more balls during auto, without any penalties?

You have to have a ball acquisition manipulator that doesn't need to grab but about 1/3 of the ball to pick it up.... Your autonomous routine needs to have precise control of the robot's location.... You have to be able to do it all quickly.... Picking up the extra boulder, crossing a defense with it and scoring it is not the hard part.... Doing so without committing a mid-line foul and all under 15 seconds is the hard part....

We had one of the better three-tote autonomous routines last year and, as we have analyzed this, we do think it will be trickier to pull off than last year's routine.... It can be done, but will the time it takes to develop it be justified by an extra 10 points a match?

[Dibit1010]

1 in the high goal....hopefully

[pandamonium]

How would you compare the difficulty of a penalty free 2 ball auto this year to a 7-disc auto in 2013? Keeping in mind that in 2013, there weren't any obstacles to traverse. It was just shoot 3, drive forward and intake, shoot 4.

I think that it really depends on your strategy. The way we are doing it seems about even in complexity to 2013. If you plan on getting the ball from the center line after crossing back into the neutral zone the task becomes more difficult.

[Jared Russell]

Look at how many teams consistently did a "shoot - get discs from midline - shoot" auto in 2013. Maybe 20 in the world?

Look at how many teams consistently did a 2 zone auto (including bump crossing and collecting/kicking balls in the second zone) in 2010. Maybe 10 in the world?

A 2 ball auto in Stronghold basically requires doing both of these tasks sequentially...all in 15 seconds...with a much narrower margin for penalties...into a much smaller goal.

Working 2 ball autonomous modes will exist but they will be more rare than multi-game piece autos in past games. You will never see one robot score three balls in a match in autonomous mode in Stronghold. Jar Nack has spoken.

(Don't agree? Get out there and prove me wrong)

[notmattlythgoe]

Look at how many teams consistently did a "shoot - get discs from midline - shoot" auto in 2013. Maybe 20 in the world?

Look at how many teams consistently did a 2 zone auto (including bump crossing and collecting/kicking balls in the second zone) in 2010. Maybe 10 in the world?

A 2 ball auto in Stronghold basically requires doing both of these tasks sequentially...all in 15 seconds...with a much narrower margin for penalties...into a much smaller goal.

Working 2 ball autonomous modes will exist but they will be more rare than multi-game piece autos in past games. You will never see one robot score three balls in a match in autonomous mode in Stronghold. Jar Nack has spoken.

(Don't agree? Get out there and prove me wrong)

Agree, and even looking at 2014 where all you have to do was shoot straight forward. I don't expect to see very many balls scored in auto let alone more than one from a single robot.

[Caleb Sykes]

Look at how many teams consistently did a "shoot - get discs from midline - shoot" auto in 2013. Maybe 20 in the world?

Look at how many teams consistently did a 2 zone auto (including bump crossing and collecting/kicking balls in the second zone) in 2010. Maybe 10 in the world?

A 2 ball auto in Stronghold basically requires doing both of these tasks sequentially...all in 15 seconds...with a much narrower margin for penalties...into a much smaller goal.

Working 2 ball autonomous modes will exist but they will be more rare than multi-game piece autos in past games. You will never see one robot score three balls in a match in autonomous mode in Stronghold. Jar Nack has spoken.

(Don't agree? Get out there and prove me wrong)

Jar Nack has been correct before (http://www.chiefdelphi.com/forums/showpost.php?p=1429184&postcount=4) in regards to auto. Wise teams will listen to his wisdom.

[philso]

Look at how many teams consistently did a "shoot - get discs from midline - shoot" auto in 2013. Maybe 20 in the world?

Look at how many teams consistently did a 2 zone auto (including bump crossing and collecting/kicking balls in the second zone) in 2010. Maybe 10 in the world?

A 2 ball auto in Stronghold basically requires doing both of these tasks sequentially...all in 15 seconds...with a much narrower margin for penalties...into a much smaller goal.

Working 2 ball autonomous modes will exist but they will be more rare than multi-game piece autos in past games. You will never see one robot score three balls in a match in autonomous mode in Stronghold. Jar Nack has spoken.

(Don't agree? Get out there and prove me wrong)

Agreed. Robots in a temporary (or permanent) Tortuga will probably not be scoring boulders.

[Chris is me]

The way we see it, if you can't score on a high goal in autonomous, you can't score on a high goal. There are just too many obstructions to aim from your own castle. It may be possible to do this using a teleop camera, but by the time you get enough resolution to take aim, your frame rate over the network is terrible, at least for a regular video stream.

Completely disagree here. It is an order of magnitude harder to drive to a "sweet spot" in autonomous than teleop. Most shooters won't be able to shoot from just anywhere that their camera has a clear shot on the goal either, but rather have the most accuracy from specific spots. Getting to those specific spots without losing position accuracy while crossing the various defenses is an immense challenge that teams seem to be consistently underestimating. Spy bot is a different story of course.

You also shouldn't need more than a single frame to figure out how much you need to turn to hit the goal, right? At least if you're using an auto-aim feature. This reduces network latency.

[Justin Montois]

lol @ more than 2.

[Taylor]

It can be done, but will the time it takes to develop it be justified by an extra 10 points a match?

If the number of teams who are working on a scaling mechanism is any indicator, 10 points a match are being taken pretty seriously.

[Knufire]

10 auto points is a pretty big deal considering that auto is the first ranking tiebreaker.

[PayneTrain]

I think just being able to possess a second ball after scoring one in autonomous consistently will be a commendable action.

[MrJohnston]

If the number of teams who are working on a scaling mechanism is any indicator, 10 points a match are being taken pretty seriously.

Yes, 10 points are 10 points.... The time investment for those 10 points is going to vary from team to team... Once the robot is fully functional, the programmers are going to need time to program any auto routines and the 2-boulder auto is going to be tricky enough that they will need quite a bit of time with the robot - and might even request that some things be altered mechanically to do it. All the hours invested in doing this will come at a cost of drive team practice.

I'm not saying that it's not worth the time investment - Heck, we are working on it. What I am saying is that there is a cost in time and teams will have to evaluate whether that cost in time is something they can afford when looking at the other needs their team might have.

Regarding the climber: We've put the strong auto routine ahead of the climber for several reasons:

1) We plan to have a very good shooter. If we can manage to put two additional boulders in the tower in the time it would take us to climb (assuming we still challenge), we have overcome the need to climb. We figure that if our partners are both climbing, there should be boulders available. If our partners are still shooting, boulders will still be fed to us... In other words, we see climbing as "nice" but certainly not necessary.

2) Auto routines are done during a totally separate time period. We look at everything we can do during an auto period as "bonus." There is no downside to scoring more autonomous points.

3) Scoring in auto will not require an additional manipulator of any sort - this saves us on cost, construction time, robot space, and the number of things that might need repair later. The code can be tested on our practice bot after bag-and-tag and therefore continually developed.

I've never said that a two-boulder autonomous can't or shouldn't be done - only that it will be difficult and require a significant time investment - at the cost of developing other parts of a team's game. For some teams it will be a good investment. For others (especially for those who don't get it figured out), it may be a poor investment in terms of robot performance. (Of course, that does not account of the potential learning that trying to figure it out will involve...)

[notmattlythgoe]

I think just being able to possess a second ball after scoring one in autonomous consistently will be a commendable action.

I think actually scoring a ball in auto will be a commendable action.

[Kevin Sevcik]

1) We plan to have a very good shooter. If we can manage to put two additional boulders in the tower in the time it would take us to climb (assuming we still challenge), we have overcome the need to climb. We figure that if our partners are both climbing, there should be boulders available. If our partners are still shooting, boulders will still be fed to us... In other words, we see climbing as "nice" but certainly not necessary.I think you mean if you can manage to put two additional boulders in the tower and CHALLENGE in the time it would take you to scale. Since you'd be foolish to miss the capture bonus and CHALLENGING, while trivial, does take a finite amount of time to execute.

[MrJohnston]

. You will never see one robot score three balls in a match in autonomous mode in Stronghold. Jar Nack has spoken.

(Don't agree? Get out there and prove me wrong)


Now, this is a challenge worth taking up.... **If** our two-boulder autonomous is up and running cleanly before our second district, I'll have to challenge to programmers to pull off a three-boulder routine... I can visualize our route and see it happening... We we can overcome all the challenges of a two-boulder routine, the only thing we'll have to do is go faster in order to pull off three... If there were 20 seconds, I bet we could do it... 15 will be difficult.

[MrJohnston]

I think you mean if you can manage to put two additional boulders in the tower and CHALLENGE in the time it would take you to scale. Since you'd be foolish to miss the capture bonus and CHALLENGING, while trivial, does take a finite amount of time to execute.

I think I said that in the message you quoted.

[JesseK]

I think actually scoring a ball in auto will be a commendable action.

Single-ball seems straightfoward for a consistent shooter. The audience and opponents will inevitably help out:
Category A: Portcullis may be slow, but easy to get consistent. Given the width of the defense and the width of typical robots, each bot on an alliance would be on its own (unlike category C). Cheval De Frise will be faster, but it's easy for the robot to come off at an angle and become misaligned.
Category B: Single ball shooters will laugh when the moat is put in front of them and cry when the ramparts are put in front of them.
Category C: They're slow to get through no matter what, but patience with autonomous mechanisms will inevitably show the way. Also don't rule out a team bringing one down for the partner in autonomous, then following that partner through. It will be fun to watch this happen.
Category D: Rock Wall has a bit of indeterminate behavior depending on high-siding of the drive train, but should be consistent if the drivers are precise in robot orientation when they set the bot down. Rough terrain is easy once a good path is figured out, but that also requires proper setup of the robot.

[Kevin Sevcik]

Getting to those specific spots without losing position accuracy while crossing the various defenses is an immense challenge that teams seem to be consistently underestimating.Quoted for truth. The most improbable part of a 2-ball auto is the part where you cross a defense, then cross back and find a gray ball on a gray carpet in a specific location. Without crossing the midline, if you're grabbing a midline boulder.

[Turing'sEgo]

Is anyone considering scoring their teammate's boulder in autonomous (given that they aren't planning on scoring it)? It seems that this is a safer alternative than getting a boulder from the center and risking G13, and it will accrue the same amount of points.

[philso]

Quoted for truth. The most improbable part of a 2-ball auto is the part where you cross a defense, then cross back and find a gray ball on a gray carpet in a specific location. Without crossing the midline, if you're grabbing a midline boulder.

And then cross a DEFENSE a third time. That's three opportunities to Tortuga...


Is anyone considering scoring their teammate's boulder in autonomous (given that they aren't planning on scoring it)? It seems that this is a safer alternative than getting a boulder from the center and risking G13, and it will accrue the same amount of points.

This would probably be a safer option up till Division Eliminations at Championship.

[Sunshine]

A one boulder score from spy bot will be common by week 1

A one boulder score from low bar bot will be common by week 1

A two boulder score from teams that dialogue with each other will occur by week 3. Especially from those with rear ball intakes working with those who have front ball intakes. But it will be a feat claimed by advnced programming and sensors.

[PayneTrain]

I think actually scoring a ball in auto will be a commendable action.

I think an alliance surpassing a 30 point auto at any district in our region is literally not going to have a chance in hell. When I meant commendable, I meant top 1% of teams.

[JesseK]

I think an alliance surpassing a 30 point auto at any district in our region is literally not going to have a chance in hell. When I meant commendable, I meant top 1% of teams.

Such a pessimist :rolleyes:

[Paul Copioli]

The 3 Boulder Auton is the Lochness Monster of 2016. Many people will claim it, but no one will actually achieve it.

Now, I have been wrong before....

[Turing'sEgo]

The 3 Boulder Auton is the Lochness Monster of 2016. Many people will claim it, but no one will actually achieve it.

Now, I have been wrong before....

I could forsee this happening: Robot A goes under the low bar and scores in the high goal. Robot B, discharges it's bot gently on the floor and moves away. Robot A goes back, picks it up, scores it. Robot C is the spy bot and also gently puts its ball on the ground in a convenient place. Robot A picks that ball up (as it's right next to it), and score it too.

Even that would be a time crunch.

[Kevin Sevcik]

I could forsee this happening: Robot A goes under the low bar and scores in the high goal. Robot B, discharges it's bot gently on the floor and moves away. Robot A goes back, picks it up, scores it. Robot C is the spy bot and also gently puts its ball on the ground in a convenient place. Robot A picks that ball up (as it's right next to it), and score it too.

Even that would be a time crunch.We managed that kind of autonomous, just, with the Wired Cats (2415) in 2012. They had a good shooter and 270 degree intake. All that 57 and 3997 had to do was feed them balls from the back and side at the right times to avoid controlling >3. And in our case, drive forward and tip the bridge to steal those balls. We JUST managed to get this programmed and tested over lunch. And 15 seconds was just about enough time to get it done. Grainy video (https://www.youtube.com/watch?v=2rPnc_xrgHg) evidence here. (https://www.youtube.com/watch?v=Pr4_za7viVc) So a successful 3 robot rube-goldberg auton in which all of the robots started perfectly positioned and none of the robots moved before the ball hand-off. And yes, it was 6 shots instead of 3, which would be slightly more time consuming. But still. 3 robots perfectly positioned, not moving.

But your thing? Yeah, I can see whipping that autonomous ballet out over lunch. I can't wait for the grainy video evidence (https://www.youtube.com/watch?v=lOxuRIfFs0w) for that one.

[Poseidon5817]

I could forsee this happening: Robot A goes under the low bar and scores in the high goal. Robot B, discharges it's bot gently on the floor and moves away. Robot A goes back, picks it up, scores it. Robot C is the spy bot and also gently puts its ball on the ground in a convenient place. Robot A picks that ball up (as it's right next to it), and score it too.

Even that would be a time crunch.

I don't think that three-ball strategy would be possible for almost all robots, as it would require multiple precise turns to intake the different balls and score (unless you have a two-sided intake (http://media.team254.com/2014/02/951707c4-2014-02-18-247x300.jpg) :D), which would be nearly impossible without going under the low bar, and even then it would be difficult. I almost think that two midline balls would be easier to intake and score for most teams than that. The sheer amount of variability and the amount of work to program that routine would not be worth the extra 10 points compared to a two-ball. In some matches, your alliance partners may even be able to score one ball, easily making up the difference.

[alexpell00]

this is a little off topic, but wouldn't you get the same number of point by crossing and extra defense as scoring a second ball?

[notmattlythgoe]

this is a little off topic, but wouldn't you get the same number of point by crossing and extra defense as scoring a second ball?

You can only get points for crossing a single defense in auto. Any additional defenses that you cross grant no points and does not it reduce it's strength.

[Dezion]

Personally, I believe that scoring two balls in autonomous will be unlikely for the majority of the competition season. It will more than likely not occur for the first few weeks and may become rare later.

One, you would somehow have to have a robot to retrieve the ball from the Midline without incurring a G13 infraction. This is possible to do, however extremely unlikely. After you score the first boulder, you will have to traverse a defense for the second time, which increases your chances of not lining up for the ball properly. (This will depend on your sensors for getting the ball.)

Two, consider other robots in autonomous. Your robot would have to avoid colliding with other robots, which would ruin both your autos. How do you plan on preventing his? (At least in Quals, I see preventing it very difficult; in Elims it could be possible do to better strategy and general better robots.)

Third, for a team considering picking up a boulder that an alliance member dropped, how do you plan on coordinating this? The alliance partner would have drop it roughly in the same spot, and it's possible that the boulder will roll away from the intended area. As well, it's possible that they may collide. (Possible in Elims, however I see it as very unlikely.)

tl;dr Shooting two (or even more) boulders will be very unlikely, and difficult, however possible and the best teams will be able to surpass the challenge.

[anishde]

Correct me if I'm wrong, but I thought that a robot is only allowed to control one boulder at a time, inclusive of autonomous?

[vann2648]

Correct me if I'm wrong, but I thought that a robot is only allowed to control one boulder at a time, inclusive of autonomous?

This is correct. 2 ball autonomous would involve acquiring and shooting the second boulder all after shooting the first

[anishde]

This is correct. 2 ball autonomous would involve acquiring and shooting the second boulder all after shooting the first

Then shooting more than two balls (or even two balls) during autonomous is an ambitious task...since you can't possess more than one ball at a time, and you can't shoot boulders over the outer works, this would entail:

1. Passing through the outer works
2. Shooting (and perhaps scoring)
3. Going back through the outer works
4. Collecting another boulder
5. Crossing the outer works for a third time
6. Shooting (and perhaps scoring)

I don't mean to be a pessimist, but I have strong doubts anyone's going to try to do that in 15 seconds.

[washedout]

The 3 Boulder Auton is the Lochness Monster of 2016. Many people will claim it, but no one will actually achieve it.

Now, I have been wrong before....

I agree Paul. And it saves me from having to print any stickers this season.

[MrJohnston]

Then shooting more than two balls (or even two balls) during autonomous is an ambitious task...since you can't possess more than one ball at a time, and you can't shoot boulders over the outer works, this would entail:

1. Passing through the outer works
2. Shooting (and perhaps scoring)
3. Going back through the outer works
4. Collecting another boulder
5. Crossing the outer works for a third time
6. Shooting (and perhaps scoring)

I don't mean to be a pessimist, but I have strong doubts anyone's going to try to do that in 15 seconds.

Many teams will *try* to have a two-boulder autonomous. Some will succeed. Of this, I am absolutely certain. No, it will not be easy.

[MrJohnston]

I agree Paul. And it saves me from having to print any stickers this season.

If anybody has a two-boulder autonomous that shoots its second boulder with more than a couple of seconds left, they are almost certain to take a serious look at doing three.... I'm not convinced anybody will pull it off, but every year I have thought that there would be something that "nobody" would be able to do - and I have been wrong every year.... If I had to guess, I would say that about five teams will make a legitimate attempt at pulling it off. One will succeed - but not be able to do it consistently.

[anishde]

Many teams will *try* to have a two-boulder autonomous. Some will succeed. Of this, I am absolutely certain. No, it will not be easy.

The reason why I'm relatively certain that few to no teams will actually succeed in scoring two boulders is because the last time there was a similar game with an obstruction between the starting position and the goal was Breakaway in 2010, in which robots could also typically only control one ball at a time. In autonomous that year, little to no scoring was done, with teams instead focusing on moving the balls around the field.