The thing we’ve been discussing on my team is that since only your 3 best challenges get scored, is it worth it to ditch the power port challenge and remove the shooter in your bot so you can be lighter for the other challenges. Were still undecided and I havent seen anyone else mention it.
Definitely consider the fact that challenge scores are also standardized based upon how well everyone else in your group does. If there is very little variance, then it’s hard to differentiate yourself from the crowd. Thus, shaving a few tenths of a second off a task might be worth less that competing in another you feel people will do poorly in.
One item to look at is how heavy IS your shooter? If it’s half the weight of the bot, then sure removing it may be a valid approach. If it’s 10%, probably better to keep it and try for the shooting challenges.
Since it’s already built, you should give each challenge a rough go before making big changes.
That’s partially a reason I made the poll.
Looking at the current results, I’m not totally convinced that the benefits of a slightly more favorable calculated score in 2/5 challenges is worth losing precious tenths of a second in the mobility oriented tasks. I’m not 100% i set it up right, but a 50th percentile performance in a challenge with 24 participants and a 50th percentile performance in a challenge with all 30 only differs by about 3 points, not worth it imo.
Today at practice we did a few tele-op runs of Galactic Search Path A to see what the actual capabilities of our bot were and were able to whittle down out time to a little under 4 seconds, and our bot is in no way designed to be mobile. I genuinely think a bot with solely mobility in mind could get a sub 2 second time. If a team without a lot of resources wants to have a sizable chance of winning their judging group, they’re going to have to specialize their bot somehow. For us I just think that mobility is the way to go about it.
We actually discussed this and decided not to. We’re well under max weight anyway and the robot is designed so that it would be unbalanced without the low goal scoop mechanism on the front. We have a very nice shifting gearbox, so our top speed is about 20 fps in the current configuration, which means we need to hang on to all the stability we can. More speed is not something we have to worry about. Besides, we want the whole package on the robot for the judged awards.
Totally fair point. Just a heads up though, I cant find anything saying the robot shown in the skill challenges has to be the same as the robot used in the at home judging. I think this is because stop build date isn’t a thing anymore. Closest thing I could find was SC8 but thats specific to the skill challenges.
We’re currently planning to contend in all five. We want to develop the pathing and vision capabilities required to do well in the Galactic Search and AutoNav challenges, but we’re also refactoring the robot to lay down solid scores in the other three challenges so no one challenge becomes a must-do-great-or-else situation.
With brand new programmers and a robot originally set up to dump low, we’ve got plenty of opportunities to improve where we’d otherwise land. Whoever gets further along gets further along.
Sub 2 second seems is next to impossible. Doing some quick math, the minimum distance between the start and the 3 pick up points is 36 feet. Taking off 3 feet for the length of a robot, that means a minimum travel distance of 33 feet. Even if the robot was doing all 33 feet in one straight line sprint at a max speed of 20 fps and acceleration of 25 fps^2, it would still take ~2.2 seconds. That doesn’t account for a variety of other things that would slow a robot down.
This may however further prove your point that 10ths of a seconds count. I’m going to have to think myself to sleep on that one.
I find it interesting that the least popular option in this poll is also the only one with a definable max score.
100% agree. The only one we’re not taking seriously is the power port challenge. Sure we’ll give it a shot, but the rules around that one seem awfully loose and exploitable that we aren’t taking it seriously (yet).
You can’t roll balls to the bot cuz they don’t want an infinit loop
Nothing directly in the rules about the bot creating a loop by just catching rebounds and yeeting them back
Blue box says they don’t want the bot sitting in one place
No direction yet on how much movement is required
Personally I think they should have 3-4 zones and you can’t shoot from the same zone more than 3 times in a row (a full payload). You can still go back and forth from 2 zones of your choice, but it seems like a simple method to get rid of static loops.
I honestly think that because the Interstellar Accuracy Challenge is the one furthest from the general robot skillset. Most teams have a sweet spot, either a single spot or an “arc” of set distance. We certainly didn’t ever have a plan to shoot from green or red zone in 2020.
Having such a huge range and being able to do it effectively is something we didn’t see from a lot of teams. This is what I believe is driving teams to consider other challenges.
While I agree with the general premise of removing static loops, I don’t think the zones should be added. Of all the challenges, this most accurately represents the gameplay that was put forth by infinite recharge, and thus should not be restricted in ways that changes how it’s approached. Many teams didn’t design for huge variances in distance because it provided such minor benefit. It wasn’t fundamentally part of the game.
Cycle as many balls into the goal, any way you like it.
What about saying you have to cross the back line every 3 shots. Even if you do come back to floor collect.
Also, our robot only shoots from the protected zone at this moment, so multi-zone game means we’d need a new shooter.
Pretty much any significant movement would due. Defenders always have to more back 6ft after a pin. Seems like a good number here.
I think the main thing is that the GDC needs to define what “minimal movement” is. Otherwise we’re taking the rebound approach and having the robot do a jitter step at the 59 second mark.
Agreed. I think 6 feet is a good number. I don’t think it needs to be strictly followed, because I think it will create a clear movement between shooting attempts. I think having to move after 3 set shots is reasonable.
While well-defined, the line bumper rule probably is unfair to teams collecting bounce backs in the scoring zone because they would have to first move to a place which might not be close to the power cells they want to collect, so a movement rule would mean they could not have that extra motion to go collect power cells.
How about something like: Robots must be in the reintroduction zone for any scored ball to be re-entered into play.
Forces robots to move at least 6 feet, but not if they are trying to rescore a rebound.
A rebound being a ball that missed (unscored)? So if a team catches a scored ball, they have to back out to the re-introduction zone to “reintroduce” it, but if the ball is a miss (or if they count it as a miss) then they can try to rescore it without backing out.
I can think of at least one robot that naturally would be good at catching rebound to re-score them with an open hopper design. But most don’t have the hopper designed to catch bounce backs, but if allowed I can imagine teams wanting to put up high scores would redesign. Even moving every 3 shots wouldn’t be detrimental to their game because they would use the backup time just to reload the spindexer from the hopper. I’m not sure if they’d be faster to go to the reintroduction zone and back or recollect off ground if they did angled shots (so as not to catch them).
If I may channel my inner Carl …
I was thinking he meant rebounds of made shots. If the goal is a taped hexagon on the wall, with careful control of the ball’s speed and spin, a robot could set up a loop in which it’s retrieving its own made shots without much lateral movement.
Oh! I see what’s going on. This was a fundamental miscommunication.
Our team has a fully build 3d scoring area, and I didn’t even consider the idea of bouncing a ball off the the 2d goal back in the robot. I was using the term “rebound” a little too loosely.
What I was trying to suggest was something akin to 2014, where until a ball is official scored, it keeps just getting re-entered into play. Once it is scored, the ball has to go all the way back to the starting point on the opposite side of the field. Since that doesn’t exactly work for 2d, I’m suggesting this instead:
After scoring points, a ball must enter the reintroduction zone (either by human or robot) before being able to score points again.
I physically shuddered at the idea of someone channeling their inner Carl, and I am Carl… What… What does that say about me?
it says you should maybe go back to bed … er … cargoship … uh … whatever.