Teamwork Strategy for Scoring

[Rick Wagner]

It seems to me that teamwork between robots might emerge as an important strategy this year. Two robots, working together, should be able to easily overpower an opposing alliance robot, push it into range of a payload specialist, and hold it there while the payload specialist fills up its trailer. What do you think? If all teams adopt this approach, it could get interesting, with teamwork counter-strategies for defense.

[EricH]

It seems to me that teamwork between robots might emerge as an important strategy this year. Two robots, working together, should be able to easily overpower an opposing alliance robot, push it into range of a payload specialist, and hold it there while the payload specialist fills up its trailer. What do you think? If all teams adopt this approach, it could get interesting, with teamwork counter-strategies for defense.
You do that. I know right where I'll be scoring with my robots.

[smurfgirl]

It seems to me that teamwork between robots might emerge as an important strategy this year. Two robots, working together, should be able to easily overpower an opposing alliance robot, push it into range of a payload specialist, and hold it there while the payload specialist fills up its trailer. What do you think? If all teams adopt this approach, it could get interesting, with teamwork counter-strategies for defense.

I can see a lot of robot pile-ups happening this way... it tends to happen like this every year. In those instances, the robots that can avoid the pile-ups tend to be the ones with the advantage.

[Rick Wagner]

You do that. I know right where I'll be scoring with my robots.

That, too, will be interesting. It's not clear to me yet how easy or hard it will be to defend against a scoring robot on that slick surface. That is, the third robot on the alliance might be able to hold off two robots attempting to score on the two robots holding the one for the payload specialist to shoot at.

Further, the two teamwork robots holding or pinning the robot being scored on will be close to their own alliance's payload specialist, so any opposing robot coming in to score on them will also be in range of the alliance's best shooter.

[MattLi]

It seems to me that teamwork between robots might emerge as an important strategy this year. Two robots, working together, should be able to easily overpower an opposing alliance robot, push it into range of a payload specialist, and hold it there while the payload specialist fills up its trailer. What do you think? If all teams adopt this approach, it could get interesting, with teamwork counter-strategies for defense.

Uhh I have a different opinion on this matter. This overpowering of which you speak is in other words pinning a robot. After doing some quick math I truly believe that pinning a robot this year is going to be harder than one thinks.

[nHouse]

I think two robots can probably hold a single bot down near the player stations even with the added carpet traction however those two robots would be sitting ducks for the other teams remaining two bots to go score happy on. It'll be interesting to see how well some teams work together this season, especially those teams that work together on a yearly basis.

[logicalyrandom]

This seems like a bad-ish idea to me, using two of your robots to make one of your opponents robots sitting ducks not only takes 66% of your team "out of the game" leaving only one alliance robot to pick up balls, ceding the field to the two unpinned bots, it also makes the two pinning robots sitting ducks for the two unpinned bots. Plus the math seems to say that pinning is going to be hard.

[Rick Wagner]

Here's another teamwork scenario that occurred to me:

Two robots (say, blue) work together to push one red alliance robot into a blue corner (best alliance shooter). Then, one blue robot could hold that red robot there. Even if the red robot has two wheels on carpet for more traction, as soon as it gets all four back on the regolith, it's only even in pushing force with the blue holding robot. With good driving, blue can probably hold the cornered robot for enough time for the shooter to fill up its trailer. If a red alliance robot comes to the rescue, it too is vulnerable to the human shooter.

[smurfgirl]

Here's another teamwork scenario that occurred to me:

Two robots (say, blue) work together to push one red alliance robot into a blue corner (best alliance shooter). Then, one blue robot could hold that red robot there. Even if the red robot has two wheels on carpet for more traction, as soon as it gets all four back on the regolith, it's only even in pushing force with the blue holding robot. With good driving, blue can probably hold the cornered robot for enough time for the shooter to fill up its trailer. If a red alliance robot comes to the rescue, it too is vulnerable to the human shooter.


I'm not under the impression that there is really that much difference in driving on carpet vs. regolith- the rover wheels are a much bigger factor than floor surface.

[thefro526]

I'm not under the impression that there is really that much difference in driving on carpet vs. regolith- the rover wheels are a much bigger factor than floor surface.

This is true. When people talk about rover wheels on carpet I just think of those nifty little hard plastic furniture sliders that slide super heavy furniture on low pile carpet.

Back to strategy though... I think more often than not you're going to want to keep your alliance's robots moving. This would pretty much make the strategy of having 2 robots pin one down pretty ineffective. What would be good though, would be if one robot pinned down an opposing robot and then their alliance member ran some kind of interference to protect the scoring robot from being score on by another robot.

Or another thing we may see is a robot pinning another robot and then scoring in it but then not moving until it's alliance partner gets to the robot to score on it. This would basically make it so that both robots don't have to track down their respective targets.

Something that will probably happen in a lot of matches is seeing two teams on the same alliance chase down the weakest robot on the other alliance. They could use the advantage of having two robots worth of traction to pin the opposing robot and then one robot leaves and continues moving but leave the other one there to score.

Regardless, good communication within an alliance will be key for victory. One of the discussions I've been having with prospective human players is that they'll be giving balls to other robots that aren't ours but on our alliance. Also the HP will have to make the decision of who to give the empty/super cells to and there 66% chance that it won't be their teams robot.

[smurfgirl]

Regardless, good communication within an alliance will be key for victory. One of the discussions I've been having with prospective human players is that they'll be giving balls to other robots that aren't ours but on our alliance. Also the HP will have to make the decision of who to give the empty/super cells to and there 66% chance that it won't be their teams robot.

Yes definitely. The payload specialists are going to have to be just as keen on strategy as the commanders, considering there is also a 67% chance they aren't going to be remotely near the commander in any given match. I think that payload specialists and commanders should probably do their "training" and their pre-match strategizing together so they'll always be on the same page, because we'll see a lot of payload specialists operating on their own a lot of the time.

[Rick Wagner]

... What would be good though, would be if one robot pinned down an opposing robot and then their alliance member ran some kind of interference to protect the scoring robot from being score on by another robot.

Or another thing we may see is a robot pinning another robot and then scoring in it but then not moving until it's alliance partner gets to the robot to score on it. This would basically make it so that both robots don't have to track down their respective targets.

Something that will probably happen in a lot of matches is seeing two teams on the same alliance chase down the weakest robot on the other alliance. They could use the advantage of having two robots worth of traction to pin the opposing robot and then one robot leaves and continues moving but leave the other one there to score.

Regardless, good communication within an alliance will be key for victory. One of the discussions I've been having with prospective human players is that they'll be giving balls to other robots that aren't ours but on our alliance. Also the HP will have to make the decision of who to give the empty/super cells to and there 66% chance that it won't be their teams robot.

Some good ideas there.

[Redskins1666]

I really do prefer the dumping method over the rest because you have multiple chances of scoring rather than just that one shot.


I really dislike this years game, because of it's human player abilities. Where is the talent in your robot when a human is shooting the ball? I would like to see the ratio for robot to human scoring and see how much more the human will score. This is more of a athletic game with a moving robot.

[ScottOliveira]

I feel that using two robots to hold another in place is justifiable, especially with a good systematic approach. After a certain number of balls, the trailer will start to fill up and it will become too costly to try to score again on that opponent. So take 2 robots (Bots B1, B2) from the same alliance and have them pin an opponent(Bots R1) next to a payload specialist(PL1). The remaining bot(B3) runs interference/ replenishes cell supply.

After they dump balls in and give the specialist a chance to load up the trailer, they move to the next opponent robot(R2). B3 and B1 engage R2 in front of a different payload specialist(PL2), whose ball supply will not be depleted. B2 runs interference to keep R1 and R3 away.

Repeat process for R3.

Ideally this would occur in front of corner PL specialists, to increase the effectiveness of the third robot out running interference.

[Rick Wagner]

I feel that using two robots to hold another in place is justifiable, especially with a good systematic approach. After a certain number of balls, the trailer will start to fill up and it will become too costly to try to score again on that opponent. So take 2 robots (Bots B1, B2) from the same alliance and have them pin an opponent(Bots R1) next to a payload specialist(PL1). The remaining bot(B3) runs interference/ replenishes cell supply.

After they dump balls in and give the specialist a chance to load up the trailer, they move to the next opponent robot(R2). B3 and B1 engage R2 in front of a different payload specialist(PL2), whose ball supply will not be depleted. B2 runs interference to keep R1 and R3 away.

Repeat process for R3.

Ideally this would occur in front of corner PL specialists, to increase the effectiveness of the third robot out running interference.

You got it. I think teamworking alliances will dominate this year.

[Taylor]

I'm not under the impression that there is really that much difference in driving on carpet vs. regolith- the rover wheels are a much bigger factor than floor surface.

I disagree - with the robot off, if we hand-push the robot on linoleum, the wheels don't turn the transmission gears. If we hand-push the robot on carpet, the sprockets turn. The wheels are grippier on carpet than they are on regolith.

[smurfgirl]

I disagree - with the robot off, if we hand-push the robot on linoleum, the wheels don't turn the transmission gears. If we hand-push the robot on carpet, the sprockets turn. The wheels are grippier on carpet than they are on regolith.

Well yes, I realize they are grippier on carpet, but from what I've heard about people testing their robots in both situations, its not as much of a difference as people are speculating. A situation I see described a lot is one of a robot pushing match where a weaker robot gets pushed by a stronger robot onto the carpet, where it regains so much traction that it overcomes the stronger one and gets back onto the regolith, then gets pushed back, etc. I don't think this is going to happen as much as people predict- there are other driving dynamics, such as the trailers, the robots slipping around each other sideways rather than a strictly linear pushing match, and the real difference in pushing power of the robots compared with how much traction the carpet will add. I'm not saying the dynamics of carpeting and regolith are identical, but rather that I don't think they're as astronomically different as some people thought (especially toward the beginning when not much testing had been done).

[Rick Wagner]

Having given it some more thought after a couple of days, it seems to me that super cell scoring in the last 20 seconds will decide a large number of matches, if not most of them. This would seem to nullify any objections to vulnerability to robot scoring while pinning an opposing robot while the PS scores.

That is, during the last 20 seconds, it's not likely that there will be enough time to feed super cells to robots to then score, so the real danger will be from the fueling station PS. The key tactic will be to recognize which PSs have super cells and then take offensive/defensive action as necessary. It's a great opportunity for two alliance partners to shove a robot/trailer into a corner while a sharp shooting PS easily scores two super cells. I think we will see this tactic performed again and again, and then we will see coordinated defenses evolve to defeat it.

[Rick]

You got it. I think teamworking alliances will dominate this year.

This is a very general statement that is true every year. Last year, we had alliance plans down to which ball each team would get, which lane to score in, and who would play defense. If you do not work together before a match, it really shows on the field.

[MattB703]

Here is another thought;

It will be much easier to stop a robot for a few seconds somewhere on the field than it will be to push a robot into a position to be loaded by your payload specialist. This is a large part of why I have been an advocate for the "mass dump" scoring strategy. My thought is; 1. fill a robot with a quantity of balls. 2. chase an opponent's trailer till he runs into something or your partner "sets a pick". 3. dump your whole load while the opponent bot struggles to extricate itself. 4. go get more balls and repeat.

I'm excited to see how the game really plays out.

Matt B

[jamie_1930]

The wheels will have more grip on the carpet, but as soon as you leave the carpet and re-enter the regolith the wheels will lose all traction, due to the speed they are spinning at and slip back into the carpet. So most likely it will work to pin with two bots, and in the worst case scenario the pinned bot will jerk around.

[Rick Wagner]

Here is another thought;

It will be much easier to stop a robot for a few seconds somewhere on the field than it will be to push a robot into a position to be loaded by your payload specialist. This is a large part of why I have been an advocate for the "mass dump" scoring strategy. My thought is; 1. fill a robot with a quantity of balls. 2. chase an opponent's trailer till he runs into something or your partner "sets a pick". 3. dump your whole load while the opponent bot struggles to extricate itself. 4. go get more balls and repeat.

I'm excited to see how the game really plays out.

Matt B

I agree that a lot of points will be scored by the chase, bump, and dump strategy. The pursuit algorithm with matched robots on a limited field will always result in the chaser overtaking. The minimal turn radius at 10 ft. / sec with 0.06g acceleration is 8 feet. The real question is who wins in the contest of bump and dumpers versus pinners and PS shooters.

[nahstobor]

The good old basketball "corner trap" might play out well.

[FRC1672]

The idea of pinning a teams robot is somewhat effective. I want to see how plausible this idea was with the simulation that was shown at kickoff but thats yet to be released. I believed this strategy will be tested and finnally decided on in the week 1 regionals. We will be at Trenton so we we be one of the the first teams possibly employing this technique.

[MGoelz]

I disagree - with the robot off, if we hand-push the robot on linoleum, the wheels don't turn the transmission gears. If we hand-push the robot on carpet, the sprockets turn. The wheels are grippier on carpet than they are on regolith.

Well yes, I realize they are grippier on carpet, but from what I've heard about people testing their robots in both situations, its not as much of a difference as people are speculating.

The wheels will have more grip on the carpet, but as soon as you leave the carpet and re-enter the regolith the wheels will lose all traction, due to the speed they are spinning at and slip back into the carpet. So most likely it will work to pin with two bots, and in the worst case scenario the pinned bot will jerk around.

First of all, I have to point out that the likelihood of your robot being entirely on carpet is rather slim. I think that was something that FIRST did to a) give the ability for some teams to just drive around the edges (but that would make your trailer an open target) and b) to help if you run into a wall to provide that tiny bit more traction to propel you back onto the regolith.

I think more often than not you're going to want to keep your alliance's robots moving. This would pretty much make the strategy of having 2 robots pin one down pretty ineffective.

I completely agree. Movement will be key. If you're not moving, your trailer is much more vulnerable. Imagine if all the robots are on one half of the field. A PS that would throw from the side where no robots are would have a small chance of scoring if you move, because by the time the ball gets there, your robot could be gone.

I really dislike this years game, because of it's human player abilities. Where is the talent in your robot when a human is shooting the ball? I would like to see the ratio for robot to human scoring and see how much more the human will score. This is more of a athletic game with a moving robot.

I would not say that this is an athletic game with just moving robots. It's all a matter of opinion, but I would like to point out that FIRST is about robots. They wouldn't create a game where the robots weren't important. I think the idea is to see that although the PS is important, robot scoring can be just as, if not, more effective. A PS has limited opportunities to score, and has the time factor of having to bend down, pick up a ball and throw it. Throwing more than one at once would be highly inaccurate (provided the trailer isn't next to the fueling station) and therefore they must throw one at a time. This goes along with:

Here is another thought;

It will be much easier to stop a robot for a few seconds somewhere on the field than it will be to push a robot into a position to be loaded by your payload specialist. This is a large part of why I have been an advocate for the "mass dump" scoring strategy. My thought is; 1. fill a robot with a quantity of balls. 2. chase an opponent's trailer till he runs into something or your partner "sets a pick". 3. dump your whole load while the opponent bot struggles to extricate itself. 4. go get more balls and repeat

Mass scoring robots will greatly sway the game. In my opinion, this is probably the most efficient way to score. When you have a trailer in front of you, and you have any number of balls in possession, you want to get rid of them as quickly as possible, before you lose that opportunity to score.

Teamwork is important in every game, but so is self-reliance. If your alliance partners get in a bind, you don't want to be helpless. You have to know that, should that happen, you can still be an effective player for your alliance and do something to help bring up the score. It's always to good to plan for the worst case scenario.

[lenny8]

I really do prefer the dumping method over the rest because you have multiple chances of scoring rather than just that one shot.


I really dislike this years game, because of it's human player abilities. Where is the talent in your robot when a human is shooting the ball? I would like to see the ratio for robot to human scoring and see how much more the human will score. This is more of a athletic game with a moving robot.

i think the human interaction wont be as great as you mite assume just because a. robots are moving b. im going to assume most of the balls will be on the field so the humans will not have many chances.

[Akash Rastogi]

Just pointing out what I see as the obvious, BUT, all I know is no matter what your strategy (and yes I've played out a CRAP load of strategies in my head) you will be compromising something, whether it is you getting scored on, or you having to not score on other robots.

Think about the best strategy you think will take place on the field, and then just think about the cons of it. Not the pros.

P.S. If you can come up with a strategy that you think has no compromises, PM or IM me and we'll talk about how wrong you are... ;)

[Rick Wagner]

Regarding the vulnerability of robots pinning another while the PS scores in its trailer, that will be true in the first 100 seconds of the match, but in the final 20 seconds, it's more likely for super cells to be scored by PSs than by robots. A teamwork pinning strategy could be decisive in the case of an alliance having super cells to score.

[fuzzy1718]

assumeing that there will be super cells to throw. if by some chance you shut your opponets out the previous match you lose what? 3 super cells I think? Supercells will not be a huge thing this season, they will be the spoilers of 2007, rarely played

[thefro526]

Just pointing out what I see as the obvious, BUT, all I know is no matter what your strategy (and yes I've played out a CRAP load of strategies in my head) you will be compromising something, whether it is you getting scored on, or you having to not score on other robots.

Think about the best strategy you think will take place on the field, and then just think about the cons of it. Not the pros.

P.S. If you can come up with a strategy that you think has no compromises, PM or IM me and we'll talk about how wrong you are... ;)

Stogi, yes we are just talking about theoretical strategies based on assumed circumstances and robot capabilities but speaking here is very useful. You're right, no strategy is ever completely without flaw but what we're trying to do here is get feedback on prospective strategies. Shoot me an IM When you get a chance tho.

To get back on topic, A friend and I came to a realization this week. For most robots, including our own, it'll be extremely difficult to get and score a single isolated ball (Super Cell) in 20 seconds. So we figured out the best way for our team to score them:

We'd make sure that both of our Human Players on either end have at least one Super Cell engaged by the last 30 seconds of the match and that our robot has 5 or more balls ready to be scored. Then in the final 30 seconds we'd wait for a robot to go anywhere close to one of our human players on either end and try to pin them in front of the airlock with the help of one of our teammates. The instant those final 20 seconds come around our human player would then have to lob a single Super Cell over the wall into your opponents trailer which is incredibly easy. At this time we'd score any balls we have in our possession and so would our team mate. If timed correctly the robots would only have to be stationary for 5-10 seconds at the most but we'd have the opportunity to score 20 or more points.

Of course that strategy would have it's flaws but it might be worth looking into.

[Rick Wagner]

... We'd make sure that both of our Human Players on either end have at least one Super Cell engaged by the last 30 seconds of the match and that our robot has 5 or more balls ready to be scored. Then in the final 30 seconds we'd wait for a robot to go anywhere close to one of our human players on either end and try to pin them in front of the airlock with the help of one of our teammates. The instant those final 20 seconds come around our human player would then have to lob a single Super Cell over the wall into your opponents trailer which is incredibly easy. At this time we'd score any balls we have in our possession and so would our team mate. If timed correctly the robots would only have to be stationary for 5-10 seconds at the most but we'd have the opportunity to score 20 or more points. ...

That's exactly what I've been thinking for a winning alliance strategy. Our scouting team will be looking for empty cell runners (it will take the full 100 seconds available for one robot to deliver the four empty cells) as well as attack 'bots (shooters or dumpers). We have a shooter bot that can pick up moon rocks from the regolith and believe that the winning alliance will have two attack 'bots and one empty cell runner.

[rjmah]

What do you think the perfect empty cell runner looks like? It could have bin that is exactly at the outpost height but probably better to have a efficient ball harvester that will always get the dropped cell. It should shoot or dump into the airlock. Pushing the balls through the airlock is not efficient and prone to having the empty cell stolen by the opposing alliance.

Blocking the outpost will also be a strategy, preventing access to the empty cells. The best defense against this is to have the outpost retain as many moon rocks as possible. In fact, the outpost should only take shots that are easy shots at the beginning. At the end game, most robots will try to be mid-field to avoid the super cells and should be easy targets.

[Rick Wagner]

What do you think the perfect empty cell runner looks like?

I'm not sure what it would look like, but it will be specialized to run empty cells quickly with no chance of losing one to the opposition. It should run four empty cells in less than 100 seconds. That will cover approximately 120 feet in that time (three and a half round trips) with time for picking up and dropping off empty cells. I don't think it should pick them up from the floor because of the chance of an opponent 'bot swooping in, bumping, and stealing. It must be a positive holding mechanism that operates quickly because the runner 'bot will be an easy target while sitting still.

[pontiacdude210]

Yeah, most teams are building a scoreing robot this year, so any cell-runners will be highly prized and grabbed up quickly in the final rounds.

[Redskins1666]

I realize that the human player will have a very hard chance of scoring, but the limits that the human has the robot has also. The aim for FIRST is to show kids about engineering and science. Yes, the pinning idea will score you major points, but where is the engineering in only making a chassis that has a lot of resistance?

[Vikesrock]

I realize that the human player will have a very hard chance of scoring, but the limits that the human has the robot has also. The aim for FIRST is to show kids about engineering and science. Yes, the pinning idea will score you major points, but where is the engineering in only making a chassis that has a lot of resistance?

There is a ton of engineering that would have to go into a robot designed to pin other robots. In order to be able to reliably pin other robots you will likely need either more maneuverability, more force or both. Getting either of those two aspects in a robot this year is quite difficult and will require significant engineering (optimized drive system, propeller/fan propulsion, or other unique robot components). Your team may feel that it is not inspirational to design a robot with such a "vanilla" function, but many teams see the elegance in selecting a strategic niche of the game and designing a robot that can fulfill that role.

[Rick Wagner]

There is a ton of engineering that would have to go into a robot designed to pin other robots. In order to be able to reliably pin other robots you will likely need either more maneuverability, more force or both. Getting either of those two aspects in a robot this year is quite difficult and will require significant engineering (optimized drive system, propeller/fan propulsion, or other unique robot components). Your team may feel that it is not inspirational to design a robot with such a "vanilla" function, but many teams see the elegance in selecting a strategic niche of the game and designing a robot that can fulfill that role.

Well said. Crab drives are very difficult engineering feats, with non-trivial control system problems to solve. A crab drive 'bot may be the best to apprehend and push another robot this year. And nobody suggests you wouldn't want to fill the victim's trailer while you're pushing him, either.

Coordinating a teamwork strategy within an alliance presents its own challenges that will task the leadership, social, and organizational skills of your team. FIRST is about math, science, and engineering, but it's also about a lot of other things, including management, leadership, cooperation, and teamwork.