Yes, “inhaling the wall” and putting it in your isolation zone is a very viable hoarding strategy. However, it may not be the best idea. The center wall contains 9 scoring objects, out of 22 total. First of all, a team that can intake the entire wall at the same time, and focus on dumping it onto the other side of the isolation zone would probably be not very good at scoring (Go ahead and prove me wrong on this one). This leaves the other team to outscore 2 robots at the same time, while those robots have more goals to work with. The team inside the isolation zone is limited to 3 goals, 4 if they can reach over the fence. You could fit at most 15 game objects in 3 goals. The other alliance has 13 objects in 6 or 7 goals, and assuming that the hoarderbot isn’t very good at scoring, that’s very roughly 18 - 19 for the team with the hoarder. Of course, there is a fair bit of play with those numbers, but it really ends up that such a strategy would lead to playing the doubler and negation barrels. Not to mention what would happen if you had 2 robots designed only for the interaction zone.
Our idea for the game was to ditch the isolation zone as soon as possible to have 2 robots working the interaction zone before the other team can do the same. It’s almost impossible to outscore 2 robots, especially at Worlds when everybody is roughly at the same level.
It almost seems as if the college challenge is easier than the high school challenge this year. Especially since all goals are in the Isolation zone, so you would build a robot good at swallowing the wall, and then just dump it all in the isolation zone for you scoring robot to work with over time.
I agree. In fact, I thought of doing this for High School, then saw the College Challenge game and knew that it was really the best way to go. (I can’t exactly think of anything that could be better?)
The college game and the highschool game have a different distribution of goals.
The highschool game has a random selection of alliance partners.
Totally Generic Name just posted (in the highschool game) that 2 robots on 1 on 3/4 of the field have a good chance beat a hoarder, along with numbers to show it. Which of TGN’s points do you think is invalid?
As in Corewar, such a simple and rigid “fully embraced” strategy is easily defeated by an exactly appropriate simple counter-strategy.
Here is one that works, I’ll call it SOAH for “score once and hold”.
SOAH: is each move to the first goal, and put on a stack of tubes, and sit there, planning to unclamp at End-1 second.
SOAH beats FOLLOW, 26:0.
GreenEggs beats SOAH, 70:26 (not bad for SOAH considering finals score against GreenEggs was often Zero…)
ScoreAll beats SOH, 70:26
Actually, for Round up, there was a counter strategy for being followed. We were handing diagrams and explanations out in our pits. The “Reverse Shadow” is used in a situation such as above where you’re being followed by another team that is descoring all of your stuff, i.e. shadowing you.
What you do is to ignore the next few goals, leaving them empty for the team behind you to score on. When you skip those goals, you get behind the team following your alliance partner. Your partner will do the same thing, and now the teams that were following you are now being followed.
You might think that there’s the chance that both teams are doing the same thing and they end up chasing each other around in a circle, but that would almost never happen because you could probably count the number of teams that used complex strategies and maneuvering during a game on one hand.
What we’re trying to do now is get a head start on developing such strategies, because last year, we only had them for our last regional in March and Worlds. They worked extremely well, and we’re trying to do it again this year. Strategies tend to take the element of luck out of the game when it comes to 2 high scoring alliances against each other, and give a very big advantage to the teams using them.
Personally, I don’t like Gateway as much as Round up because it feels like they’re forcing the strategy into it, instead of being naturally there but hidden. There were a ton of things you could do with the movable bases and the ladder, even without considering scoring. Gateway has fixed goals that can only be used as goals, and the gates are only there because tape lines on the ground aren’t as definitive. Descoring was also a huge thing, and taking that out could be very limiting with only 22 game objects a side. What happens when teams can score all 22 game objects every time? Then it almost becomes a game of chance by depending on the doubler and negation barrels in the last 30s.
How do you expect hoarders to be able to keep their hoards in the last 30 seconds of the HighSchool game?
Which is better offensively, dehoarding, or double/negating?
Which is better defensively, guarding the goal with empty spots in it, or guarding the hoard?
I expect to see a lot of aggressive robot action, given statements about the “benefit of the doubt goes to aggressive team”.
Any board gamers out here?
It seems like a board game version would be allow strategy experiments.
Unless you had a robot that carried around all of the game objects like the bucketbots of Round Up, it would be very hard to keep your hoard. It would also be very hard to keep a hoard, if your isolation zone gate was up. On the other hand, we’re considering letting them HAVE our hoard…
On the subject of offensive and defensive strategies, it is really interesting as to deciding which to do in the middle of the match. We don’t know how that will work until the year progresses and robot designs reach a point where strategies can matter. Then we will be able to see how the flow of the game works, and plan around that. Of course, strategies only work as long as teams play within the bounds of standard gameplay (scoring first, hoarding/scoring more, and doubler/negations). If something happens to change that, like the base-in-the-ladder robots, then we have an issue.
Has anybody tried to simulate a game inside their head? Its very hard, but if you get all of the factors right, it can be very insightful.
You seem to be glossing over the last part of the strategy; and glossing over the adaptable nature of a human driver.
If an opponent executes the SOAH, while a COREWAR program is simple and rigid, I think we can trust a non-rigid human have fun scoring at will on other goals while the SOAH bot defends their one goal. SOAH loses.
The overarching purpose of the Follow/Shadow strategy was doing something simpler than many opponents would attempt and no more complex than almost all opponents. A strategy that involves simple actions can usually be carried out faster than one that involves complex actions, or will free up resources consumed by complex actions.
If the simple actions can repeatedly create relatively large (large enough) swings in the final score, then they make the complex actions a waste of resources (time or physical items).
Once the complex actions aren’t valuable, the game shifts to who can carry out the simple actions better.
Green Eggs… Get between them and the ladder (simple) so that they can’t put a goal in there (complex) without pinning you. Also build a robot that anticipates needing to remove and place tubes from/on a goalpost that is anywhere on the field (including under the ladder) (not 100% simple but relatively easy to do).
ScoreAll - Not sure what that is.
Wrap Up statement: RoundUp was definitely a good game in that there were several Rock-Paper-Scissors sorts of strategies and counter strategies. Consequently, being a one trick pony is a bad idea in a game like RoundUp. I think a Follower implementation, or any robot, needed/needs to be able to switch between a couple of effective strategies to do well in a typical VRC/FTC/FRC STEM robotics competition.
All good points Blake. My team is just bad at knowing when to be flexible.
I remember a QuadQuandry game we lost:
We had control of the goal, in the lead.
opponent big hook dragger robot was out in the middle of the field doing a little dance, not paying any attention to us…
So we left the goal to go grab more rings (greedy)
==> giving opponent an opportunity to hook the goal and drag it away.
Oops.
Speaking of GreenEggs vs the little under-ladder robot, (254?)
one blocking strategy would be to sit under the ladder at the goal dropping spot, and find out if the referee will call “deliberate damage” if the goal drops on you.
Most of our practice sessions are just generic robot operating skills, not game simulations where strategy might be important. One competition is typically 6-8 games, mostly against mismatched opponents, plus up to 6 more if you make it to finals. ie not that much opportunity to practice strategy.
A board game version might help, even more if the strategies enumerated and tested against every other strategy, like Corewar.
Multiple people suggested that we do that; we (1337, 1103, 254A) emphatically rejected that strategy. Moreover, I told 44’s drive team not to drop goals on us, because winning by DQ isn’t really winning at all.
Then again, deliberate destruction might not have been called at all. I’d confidently drop several goals on 254A - there was a polycarb shield over most of the robot.
Sure, tempting DQ isn’t really reliable anyway.
If they don’t drop on you, then is it easy to block them, or did their mechanum wheels make it easy for GE to sidestep you?
Not sure whether it bears any significance, but at the Mid-Atlantic competition, this very question came up when a team did drop a goal on another robot. Basically, the ruling was that if done intentionally, sure, DQ might be in order, but otherwise…
If you have a robot that went underneath the ladder, and dropping goals was happening quite a bit, by placing yourself there, you are putting yourself in danger… Whether a robot hurts you from the sides, or from above or below, all of the same rules apply…
Back to Original topic, predictions of Roundup that came true or not.
I couldn’t find the posts about “stop worrying about carrying the goals around and just play the game as it was intended”. Was that just a voice in my head, or does someone want to own up to it?
There was also a lot of strategic thrashing about whether hanging was ‘worth it’ or not. I remember making some prediction that there would be at least one hang-capable robot on each finals alliance; That may not have happened…
I didn’t stay for all the finals, just the Titan/GreenEggs battle.
Were there any hangs during either HS or College finals?
Any concensus on whether it was “worth it” to hang, or still debatable.
44 wasn’t using their mecanum wheels on their competition robot…the 8 motor drive gave us bigger problems when we were trying to change colors of the goals they had already deposited.
I guess now is a good time to post a bit of what Tech. Alliance 2 was thinking:
Match 1: 1337 & 254A vs. 44 & 1492Z
The plan was for 1337 to dump as many goals into the ladder as fast as possible, even at the expense of scoring on them. Once the goals started filling the ladder (both 1337’s and 44’s), 254A’s job was to turn them blue from the outside. 1337 would then cover wall, since they could score wall whereas Green Eggs could not. We anticipated some amount of defense primarily from 44 (who would then be mostly bored ;)), which we planned to simply run away from (pretty sure 254A was the fastest robot on the field). In the actual match, it seemed like both 44 and 1492Z started to direct their attention to us, because it took significantly longer to change each post than I had anticipated. One critical aspect we missed was 44’s fork - we knew it was there, but didn’t think that it could descore inside the ladder.
As I mentioned somewhere earlier though, if everything went the same and 254A’s autonomous curved left the way it should, 1492Z would have missed their second goal in autonomous and Blue would’ve won by 15.
Match 2: 1103 & 254A vs. 44 & 1114M
This was supposed to be much more of a guaranteed win, with 1103 on the field (Alliance 2’s backbone through all of Technology eliminations) and Science’s #3 subbing in for #2. Even so, we modified our strategy to look more like jgraber’s suggestion - passively preventing Green Eggs from depositing goals by simply running in the way.
Case 1: 44 doesn’t dump the goal and tries dumping it elsewhere. 254A follows, and I don’t think 44 would’ve kept it up for long.
Case 2: 44 leaves the first goal outside the ladder and later tries to catch 254A napping. I was planning to be on the lookout for 44 munching a goal down the stretch.
Case 3: 44 simply dumps it.
*** Case 3a: Referees call it intentional and rule a DQ after the match ends. I would’ve felt terrible.
*** Case 3b: Referees call it unintentional, but there’s no way of knowing the ruling immediately anyway. 254A lifts its arm, which had plenty of power to unseat the goal from wherever it landed on the robot. From inside the ladder this time, 254A changes colors of goals.
In the general case where goals do not end up inside the ladder, 254A and 1103 would’ve played like a normal match. Our combined strictly offensive capability was (IMHO) much better than 44’s and 1114M’s, so I doubt 44 would have gone down this path. 1103 would try to hang well before the end of the match, and 254A would clean up where it could by criss-crossing the ladder as usual.
In the general case where goals eventually end up inside the ladder, it would depend on how quickly 44 put them in. 254A would have changed colors as fast as it could from inside (we realized after Match 1 that the key was to descore the bottom ring, the one that 44’s fork couldn’t reach), but 44 might have been able to barricade 254A in. So now its time for more cases:
Case A: 44 dumps goals inside quickly enough to force a choice between staying inside and running back out. If 254A had changed over less than 3 goals by that point, I would’ve called stay in; otherwise I would have preferred our robot out in the open.
*** Case A1: 1114M starts scoring on goals inside the ladder (don’t know if they could, but it looked likely). 254A stays on the ground and clears each post.
*** Case A2: The goals stay blue or blank. 254A high hangs for the first time in the competition with 30 seconds left.
Case B: 254A tries to score/descore the remaining Red goals in the ladder from the outside and once again becomes susceptible to defense. Hopefully with GreenSimEggs obsessing over Poofs A, 1103 would find an opening to hang.
During autonomous, this strategy started to take place. For the first time all tournament, 254A lined up to run under the ladder - we were aiming at the spot where 44 makes their first dump. However, with 1103’s…minor setback…the best shot we would have had at a close score would have been to run under the ladder, hope that 44 dumped every single goal underneath after we were in place, hope that 1114M didn’t try to score on the empty posts, then clean them up and high hang (lots of hope involved).
The point totals would be:
RED:
3 wall goals * 11 points (this excludes the blank one that 1103 was blocking; I’m assuming 3 rings per post)
= 33 points
BLUE:
1 floor goal * 7 points (we would have grabbed one blue ring before running in)
20 point High Hang
10 point autonomous (negated because our robot evidently shifted before the match and the preload touched the post)
= 37 points = miraculous win
Unfortunately, 44 was busy doing its thing while we were trying to right 1103. By the time we all gave up on Currahee, I couldn’t see a route for 254A to get inside the ladder because it was full.
The deliberate damage clause is preceded by the phrase “sole purpose”. The primary purpose of putting a goal under a ladder is winning the match, not damaging a robot that intentionally puts itself in harms way.
At least that is how it was called at the Mid-Atlantic last season.