Good strategy?

[tkwetzel]

I came up with what I think is a very good option for strategy and game play, however the rest of my team decided it was not the best idea. It involves making a net or basket about 8 feet wide that you can unfold and position under the ball dump in the autonomous mode. If you had a reliable partner, they could knock off the 10 point ball, hopefully not too soon, and you would catch most of the balls if not all of the balls. You could then have a way for them to be put back out in the corner near your ball corral, so they would not all be spread across the field and you would be able to easily push them into the corral as fast as your team could remove them probably. If you could consistently do that you would most likely be chosen in the finals. You may not automatically go to finals, because you may have one or more partners who cannot make it to the 10 point ball. However, a team who could get to the ball easily and is in the position to pick alliance partners would love to be able to control the balls. During STUbots (students act out game) we found that to be one of the best moves. However, our team decided that we don't want to have to rely on the other robot to get to that ball.

[mightywombat]

I think it is a fine strategy. Many, if not all, of the 'bots SHOULD be able to knock off the ball. I would suggest that your 'bot also be able to knock the Bonus Ball off and not catch the falling balls. Just in case.

[Aignam]

I think it is a fine strategy. Many, if not all, of the 'bots SHOULD be able to knock off the ball. I would suggest that your 'bot also be able to knock the Bonus Ball off and not catch the falling balls. Just in case.
That seems like the best option. The strategy itself seems pretty good.

[Doug Traster]

It involves making a net or basket about 8 feet wide that you can unfold and position under the ball dump in the autonomous mode.

Personally, I think this is a wonderful idea!! Interestingly, my team had a similar reaction to your team's when I suggested it. I can't feature wanting to chase balls all over the field when you can just catch them and then funnel them into the ball chute all at once. It might be interesting to consider a telescoping arm or tethered minibot to go out and trigger the ball dump while your robot waited patiently under the dumper. This would seem to overcome their primary objection. In the simple case, however, it makes autonomous almost fool proof -- move 3 feet, unfurl net. I'm still trying to convince my team of its merits. Good luck with yours.

[Marygrace]

Our team talked about something along those lines. Maybe not catching them and containing them all but having a way to direct them towards the chutes instead of them bouncing away. We decided against it, we thought other aspects of the game were more important....well, not more important, just other things would give us a larger advantage in the game.

[KenWittlief]

actaully I think you have nailed the 'ideal' ball gathering concept - get them while they are most concentrated in time and location - that is, all at once

I noticed you mentioned a net - that is an implemtation detail - there are many ways to catch them as they fall.

as for the release ball - your side will have 6 starter balls behind the wall at the start of the game - even if the release ball is not moved, the 18 balls on your side will fall 30 seconds after auton ends

that gives your side 5 seconds / ball to shoot the starter balls, then the 18 drop and you have 75 seconds left - still 4 seconds per ball if you can deliver them quickly to the corral

so even in your worst case, you end up with control of 24 balls with a potential score of 120 points - just for your bot alone - not bad at all! in fact, that is the MOST points that one robot can 'score' if it can only do one thing!

PS: you spelled Strateejury wrong :c)

[tkwetzel]

Our team also thought about the arm/tethered bot, however, the arm would be difficult over that distance, and the tethered bot would still have to be able to knock the ball off in 15 sec., so there were arguments that shot that down.

[edomus]

a funnel would give only one side of the alliance the balls, thus hurting the time it takes to get them into the goals. Also, what if your alliance has a bad autonomous?

[tkwetzel]

If your alliance partner has a bad autonomous mode, you either wait till 45 seconds, as mentioned by KenWittlief or you go after other objectives. Whis will not kill you if you can still get under the drop consistently and efficiently. You may throw a few matches, but you will most likely be chosen by a team for finals if you can control the balls and they can get knock the 10 pointer.

[Celestial Rain]

Hey guys! I have just been reading this and I came up with an idea, which im not sure if its illegal. U know how u can cover your side with a net or pole? Well have u thought about doing the same exact thing on the other side of the field (with a telescopic pole). I mean the rules state that you cannot goal tend. Or block a ball on its fall. But wat about doing it to the ball slots on the other side. =P Its just a thought

[KenWittlief]

think of it in other terms, and the beauty of it comes out.

say for example, I want to buy 5 pounds of M&Ms from the bulk section of the supermarket

would i dump them all over the floor (even if it was clean) then try to scoop them into the bag

or would I take the scoop and dump them directly into the bag?

[tkwetzel]

That has been debated here:
http://www.chiefdelphi.com/forums/showthread.php?threadid=23651

[tkwetzel]

That last post of mine was referring to blocking the ball chutes.

[Celestial Rain]

ok well i read the post and understood it. My next question is it good to grab the 10 pt ball in the autonomous mode or hit it

Is it illegal to smack the pvc bars underneath the ball and try to get it inside the arena?

Also i noticed that the upper deck is a foot high. With torques involved, is there a good way to get up there?

[abeD]

Its a foot high off the ground, so if you are on the first platform(which is 6" off the floor) then you only have to go up another 6" to get to the top platform

[stuy694]

I like the idea, I also mentioned it in a post http://www.chiefdelphi.com/forums/showthread.php?t=23680.
But I'd also like to ask, how often as robots gonna hit/knock down both trigger balls in autonomous mode as there is like a foot of space between the two. Cuz if its likely that your opponent will trigger both ball releases then this design has a much better chance of winning even if your partner can't trigger the ball release.

[animater31405]

Ok. Now I have to make you look at the flip side. The cost, weight, power, and size of the robot are HUGE factors if you want to make a device like this especially the height issue. Not to mention how easy it is to break that. And, you may have a problem if you want to get the 50 points there for the taking. It's a good startegy if your robot is not able to do anything else for about a minute in the match.

[tkwetzel]

It is easy to design a catch system with enough power, strength, and size to accomplish this. And it should not be that exspensive either. And it still a good strategy, even if you still want to do other stuff.

[KenWittlief]

Ok. Now I have to make you look at the flip side. The cost, weight, power, and size of the robot are HUGE factors if you want to make a device like this especially the height issue. Not to mention how easy it is to break that. And, you may have a problem if you want to get the 50 points there for the taking. It's a good startegy if your robot is not able to do anything else for about a minute in the match.

now you are jumping from the big WHAT (should our robot do) to the big HOW (will the robot do that)

funny thing about engineering - the biggest mistakes are often not going after the best WHAT - taking on the best concept and finding a way.

I promise you - there will be bots at the events that find a way to catch every ball before it touches the floor - because they will focus on that, they will keep at it till an elegant solution pops into their mind at 3AM - and when you see it your jaw will hit the floor and you will spontainously say "Why didnt we think of that!"

Why - cause you gave up on it too soon - the solution didnt come to you in 30 minutes or less, so you thought it was too: hard, complicated, expensive...

Once you have the best answer to WHAT your robot should do, its time to start kissing frogs - keep thinking up possible solutions - most of them will be frogs - but sooner or later you will kiss one and it will turn into a beautiful thing.

Anything worthwhile takes a little time!

[Thulium]

a funnel would give only one side of the alliance the balls, thus hurting the time it takes to get them into the goals. Also, what if your alliance has a bad autonomous?

Considering it is stated in rules that balls are allowed to be passed to alliance members, I think 1:45 is plenty of time for 2 people to throw 18 balls (assuming good efficiency).

[KenWittlief]

what if your alliance has a bad autonomous? help them improve it before your first match together :^)

[Lil' Lavery]

I didnt have that many major problems with it, with the exception of the possible waiting for 45 seconds. But if we use a runner, we could knock the ball off ourselves and get under the ball dump in the autonomous mode. Plus with a fabric, unfolding chute, it wouldn't be too heavy, and we'de gain control of the balls quickly. Hmmmm, another possibility for the runner.

[Jedi Padawan]

We thought about that too in conjunction with a few other doo dads we could attatch and then realized how expensive in time for us it would be to produce that so we canned it. (Despite the fact I still think it's a relatively good strategy)

[KenWittlief]

what do you mean by a 'runner'?

a tethered bot? there is an entanglement rule this year - having a detachable baby bot on a wire probabally wont be allowed.

[John JediMaster]

I don't know if i'm wrong or not. But in the past when it came to strategy we didn't want to have a primary strategy that involved relying on our aliance partner. Now I agree that this is a good strategy, but not for the qualifiing rounds, but it could work for the elimination rounds when you know your alliance bot's strategy. But I see the point your teammates made. Try mentioning that it could work during the elimination rounds.

All the best.

[KenWittlief]

our team talked about this idea, catching all the balls as they fall, and we decided not to use it. Reasons being:

1. its all or nothing - if it fails to work you get none of them - if you are tying to get the balls in auton mode, and something goes wrong, the balls will be all over the field, and you will have no way to collect them.

2. its complex to build, would be vunerable to damage, and might be difficult to fix between matches

3. as the game is being played you need to keep the HP supplied with a steady stream of balls - this approach would dump all the balls on one human player all at once, and your bot would probabally have to stay there for a while - 18 balls wont fit in the corral all at one time, so your bot is not free to go do a different task. we decided it would be more time effecient to gather a few balls, deliver them, then gather a few more.... or go after the 2X ball in between gathering.

[kevinw]

our team talked about this idea, catching all the balls as they fall, and we decided not to use it. Reasons being:

1. its all or nothing - if it fails to work you get none of them - if you are tying to get the balls in auton mode, and something goes wrong, the balls will be all over the field, and you will have no way to collect them.


I recall seeing some amazing ball collecting bots in 2000 and in 2002, many of which had a large basket for storing balls on top. I don't see why you couldn't try to get the balls that fall, and still be able to collect the ones you miss.

[ngreen]

My dream robot would be one that could catch all the balls...teamwork can be great...while still be able to load them from the bottom, it would then grab the mobile goal uncap it and hold it..go infront of the ball coral and release the balls while the mobile goal was on the back of it. The HPs would be trained to not overshoot because if the shot is short it would return to the basket of balls and return to the HP. Your own ball rebounder/returner. Once the mobile goal is full the robot would cap it and try to retrieve another 2X. When accomplished it would wait in front of the stationary goal. And rebound shots while returning them by aiming its conveyor return and shoot the balls into their corral. At a certain time or when the goal is full the robot would cap it, climb to the bottom platform via the stairs and hook to the bar and pull itself up. But I'm still dreaming. I hope I get woke up at 3 in the morning with how to do this all effectively.

[tkwetzel]

"1. its all or nothing - if it fails to work you get none of them - if you are tying to get the balls in auton mode, and something goes wrong, the balls will be all over the field, and you will have no way to collect them."

This is all part of the strategy. This is true, but it is not a big enough reason to eliminate the use of this strategy. If your alliance member (or opponent) fails to hit one of your bonus balls in the auton mode, you can wait the extra 30 seconds and then give all the balls to your player, you can try to get small balls from your opponents side, or you can attempt to control a 2x ball or the goals. Another thing about this is that you don't have to be one of the top teams to be in the finals. If your robot can reliably collect the balls and deliver them to your player station effectively, then you will most likely be picked by one of the top teams who did well being able to control the goals and 2x balls.


"3. as the game is being played you need to keep the HP supplied with a steady stream of balls - this approach would dump all the balls on one human player all at once, and your bot would probabally have to stay there for a while - 18 balls wont fit in the corral all at one time, so your bot is not free to go do a different task. we decided it would be more time effecient to gather a few balls, deliver them, then gather a few more.... or go after the 2X ball in between gathering."

This is false logic, because if the robot supplies all the balls to the player station in the first 45 seconds of teh match (including auton), the human player can then have the rest of the match to be throwing all the balls and the robot can focus on one other task, instead of having to interrupt the 2x ball manipulation or bar hanging to get more small balls to the human player. And if you are efficient about collecting the balls in the beginning, it will ultimately give your human player more balls in a shorter period of time then collecting them across the field. Your human player does not need a steady stream of balls, if he can get them all in the beginning. When the robot delivers all the balls, the human player and coach and possibly one of the drivers can help remove all the balls from the corral as they quickly come in. Then there are plenty of balls for your alliance players throughout the match. It will also free up your alliances robot to concentrate on other tasks, such as goal manipulation, 2x ball control, and/or bar hanging, so they will be able to more effectively perform whatever tasks possible.

I believe I have worked through all the negatives of this strategy and found that it will be more effective than any other strategy, as long as the robot is reliable. The reliability all depends on the team, the design, and the auton coding, which is all dependent on the team. I welcome any other suggestions for or against this strategy as I try to look at every aspect of it.

[KenWittlief]

I recall seeing some amazing ball collecting bots in 2000 and in 2002, many of which had a large basket for storing balls on top. I don't see why you couldn't try to get the balls that fall, and still be able to collect the ones you miss.

you could, but that means you need two separate systems on your bot to collect the balls, one to catch them from the 8' wide dropper, and a second to collect them from the floor.

We decided if you have the floor-collector you dont need the falling ball collector.

[KenWittlief]

My dream robot would be one that could .......

yeah but can it make popcorn for your team while its in the pits?

[Aignam]

you could, but that means you need two separate systems on your bot to collect the balls, one to catch them from the 8' wide dropper, and a second to collect them from the floor.

We decided if you have the floor-collector you dont need the falling ball collector.
Can't the mechanism that catches the balls from overhead serve a twofold pupose in also being the hopper for the balls collected from the floor? Then, when you need to empty out the hopper into the corral, simply 'reverse' the ball collecting mechanism...

[KenWittlief]

This is false logic, because if the robot supplies all the balls to the player station in the first 45 seconds of teh match (including auton), the human player can then have the rest of the match to be throwing all the balls and the robot can focus on one other task

look at the drawings of the ball corral - its 4 feet wide and the balls are slightly over 1 foot wide - ie you can only get 3 balls through the opening at a time

and then where are you going to put those 18 balls, in additon to the 3 you aleady have there? the bot will be forced to sit there while the human player shoots most of them, maybe 2/3rds of them, before it can feed the last ball through the corral and go off and do something else.

Its gonna take a good shooter 4 or 5 seconds per ball to pick them up and shoot them - thats a good minute at least that you bot has to sit there feeding the balls through.

besides that, the best time to go after a 2X ball is in auton mode - thats when there will be the least amount of contention for them - and there is on reason why a floor collector cant sweep the floor for balls during auton mode too - having several to supply to the HP when auton ends.

dont get me wrong on this - catching all the balls as they drop IS a good strategy - but it DOES have drawbacks and potential problems (like all strategys will).

Based on our teams analysis of the game, our resources and funds, and the tradeoffs of each approach we looked at, this one got eliminated. That doesnt mean another team cant use it and pull it off.

[Jake177]

and then where are you going to put those 18 balls, in additon to the 3 you aleady have there? the bot will be forced to sit there while the human player shoots most of them, maybe 2/3rds of them, before it can feed the last ball through the corral and go off and do something else.
The rules clearly state that the human player does not need to be the one who grabs the balls. Since your robot isn't moving anyways, you could have one of your drivers grab the balls and just put them on the floor. I would imagine that, even with 18 balls in your hopper, you could deliver all of you balls this way in less than 30 seconds.

[Trinora]

A robot with a 8 foot bin is completely possible, as is making one that could contain all the balls. It's all about style and skill, and when you have a great idea, it's best to run with it before someone realizes what you are up to.

Our team also dismissed this idea, because it seemed too top-heavy and we had some other really awesome ideas. A robot like this would be a specialty robot, and, most likely, a first pick for finals. Simple attachments could make it the ideal Ball Manipulator robot.

Don't dis the idea when it's sound the whole way 'round. :.)

[KenWittlief]

Can't the mechanism that catches the balls from overhead serve a twofold pupose in also being the hopper for the balls collected from the floor? Then, when you need to empty out the hopper into the corral, simply 'reverse' the ball collecting mechanism...

having something that serves two purposes usually corrupts both of them. Its best to design a subsystem that has one function and optimise it for that one thing

when you try to morph two functions you end up with a kludge. Our team is gathering balls from the floor, and we have no reason to toss them up into an 8 foot wide dumpster - the less you have to move the balls the better.

another thing to consider for this - the passage way to the other side of the field is 8 feet wide. Build a bot wide enough to catch all the balls and you wont be able to drive it to the other side of the field, unless you collasp it back down first.

[JVN]

having something that serves two purposes usually corrupts both of them. Its best to design a subsystem that has one function and optimise it for that one thing

when you try to morph two functions you end up with a kludge. Our team is gathering balls from the floor, and we have no reason to toss them up into an 8 foot wide dumpster - the less you have to move the balls the better.

another thing to consider for this - the passage way to the other side of the field is 8 feet wide. Build a bot wide enough to catch all the balls and you wont be able to drive it to the other side of the field, unless you collasp it back down first.The question should be raised...

Why bother picking up small balls when you can catch 18 right off the bat?

I mean... let's say you are able to pick up 25 small balls off the floor, and deliver them to the HP during a match (in several loads, or all at once, or whatever). Wouldn't it be more beneficial to get 18 IMMEDIATELY, and then spend the rest of the match (~1:30) playing defense? Say by playing defense you keep your opponent from scoring 10 balls -- by slowing them down, and getting in their way. Now you've effectively swung the score 28 (or more) points in your favor.

Now... if you have a machine that catches all 18, and then can also pick up more small balls... you are going to be a force to be reckoned with. Every ball you grab is one your opponent can't.

Robot that gets 18 instantly. vs. Robot that gets 30 over time.

I know who I'd want for a partner in the finals.



Just something to think about.
Yay for stratejury.

John

Also -- I've seen plenty of mechanisms that accomplish more than one task, and do it better than several individual mechanisms. Prime Example -- 177's 2001 Robot. Look it up. It had an arm that could do EVERYTHING.

[ColleenShaver]

I have to agree with John on this one. And there are plenty of relatively simple ways teams could knock their bonus balls AND catch the 18 balls. Take some hints from some of the bots with tethers in 2002. There were bots that could spread across the entire field in 2-3 seconds using spring loaded scissor mechanisms and such things.

With a little modification, knocking off the bonus ball can be a light simple mechanisms based off some people's 2k2 devices (or your new ideas)... just in case your partner can't do it.

[KenWittlief]

with the entanglement rule I really doubt the judges will allow any tethers mouse-bots to be running around during auton mode or during most of the match

the tethered mice were only allowed in 2002 because they were released in the last few seconds of the game, and there was little risk of entanglement - if you entangle another bot I believe you are disqualified for that match. (not sure on this)

I posted all the reasons why our team choose not to try to design a dump truck to catch the 18 balls - I dont want to tie up the forum repeating myself.

You can only put so many subsytems on your bot - if you are going to catch the falling balls AND get them off the floor AND release the trigger ball AND get the 2X ball....

if you can do all that I will be impressed :c)

(but then again, engineers ARE easily amused! )

[tkwetzel]

Doing everything really isn't as hard as everyone thinks...or is it? We'll have to wait and see what happens. If I had my way with the design, the bot would easily collect the balls as they fall when my bot knocks the bonus ball off the tee, after 15 seconds it would then efficiently deliver them to my player station, after 45 seconds the human player would have all 18 balls and the robot would go after a 2x ball and both mobile goals, after 1:30 or less the bot would deliver the goals infront of our alliance player station then head towards the platform, dropping the 2x ball on the stationary goal with at least 18 balls in it and then head for the bar. The robot would easily climb the 6 inch steps and reach the bar to hang before the end of the 2:00. I believe that all that is possible, without it being to complicated. However, my team decided that we did not want it, so I am disappointed. Oh well, I hope I see a robot in competition like the one I envisioned.

[KenWittlief]

there is an easy way to design a robot that can do everything, like you want

Build a lightweight base platform with two motors, the RC and battery

and have a low, flat top

onto which you screw a pair of sneakers.

then find out which student on your team weighs the least

wrap that student up in alum foil, leaving openings for sight and breathing

and then stand in the sneakers, acting like a robot as much as possible

the hardest part about this will be getting shipped to the regionals in the crate - the trucks are not heated, and they will have to spend a few weeks in the drayage warehouse - but after that they really WILL be the lightest student on the team

:c)

[Aignam]

Doing everything really isn't as hard as everyone thinks...or is it? We'll have to wait and see what happens. If I had my way with the design, the bot would easily collect the balls as they fall when my bot knocks the bonus ball off the tee, after 15 seconds it would then efficiently deliver them to my player station, after 45 seconds the human player would have all 18 balls and the robot would go after a 2x ball and both mobile goals, after 1:30 or less the bot would deliver the goals infront of our alliance player station then head towards the platform, dropping the 2x ball on the stationary goal with at least 18 balls in it and then head for the bar. The robot would easily climb the 6 inch steps and reach the bar to hang before the end of the 2:00. I believe that all that is possible, without it being to complicated. However, my team decided that we did not want it, so I am disappointed. Oh well, I hope I see a robot in competition like the one I envisioned.
Robots like these have big red bullseyes painted on them. If your robot can do everything, make it very durable.

[Jake177]

The secret to building a good "everything" robot is going through as many designs as possible. That's how we arrived at our 2001 bot. We started desgining the claw as just a way of grabbing the big balls and designed a separate goal grabber. We then realized that, with just a few minor adjustments, we could use it to grab the goals as well.

[Levaman]

The secret to building a good "everything" robot is going through as many designs as possible. That's how we arrived at our 2001 bot. We started desgining the claw as just a way of grabbing the big balls and designed a separate goal grabber. We then realized that, with just a few minor adjustments, we could use it to grab the goals as well.

We're using a similar process this year. We designed a system to do one thing. Adding in the other systems, we're designing a functional robot designed around a few key systems. :]

[Tyler Olds]

I came up with what I think is a very good option for strategy and game play, however the rest of my team decided it was not the best idea. It involves making a net or basket about 8 feet wide that you can unfold and position under the ball dump in the autonomous mode. If you had a reliable partner, they could knock off the 10 point ball, hopefully not too soon, and you would catch most of the balls if not all of the balls. You could then have a way for them to be put back out in the corner near your ball corral, so they would not all be spread across the field and you would be able to easily push them into the corral as fast as your team could remove them probably. If you could consistently do that you would most likely be chosen in the finals. You may not automatically go to finals, because you may have one or more partners who cannot make it to the 10 point ball. However, a team who could get to the ball easily and is in the position to pick alliance partners would love to be able to control the balls. During STUbots (students act out game) we found that to be one of the best moves. However, our team decided that we don't want to have to rely on the other robot to get to that ball.


Hmmmmmm, sounds like a good strategy to me!!! I wonder how many teams will do that?

[PsiMatt]

We will see how it fares at the regionals