Defense in Ultimate Ascent

[GeorgeM]

Originally Posted by Koko Ed

There was a fan rookie robot at GTR-West (4716 who qualified for the championships as a GTR-West finalist due to the Wildcard). Karthik refereed to them as a Peacock due to the tail fan the robot had. They can only play defense but against the 1114/2056 juggernaut they were not fast enough to keep up (1114 would elude them by darting under the pyramid which they were too tall to get under)and because they did not score any discs in autonomous it put their alliance at a tremendous disadvantage.

They were a great defensive robot! I'd go as far to say that they were the integral part to making it to the finals. They played lights-out D on Lo-ellen in Quarters and shut-down Theory6 in Semi's.

They would have been good enough to stop 2056/4814/1114, except they made a few mistakes that 1114 capitalized on. Instead of staying closer to the center of the field, they attempted to defend Sims in between the loading station and the pyramid. Since their "Peacock" could not get under the pyramid, 1114 took that opportunity to evade them. I think if they would have stayed closer to the center, their defense on 1114 would have been magnified.

...A great defensive robot!

[Nemo]

How tough is it to play defense on a climb and dump robot?

After poking around in the rules and the Q&A for a while, here's what I've gathered:

Letting the climber push you into their pyramid is bad; probably gives you a technical and gives them a free 30 point climb.

It is possible to legally play 2 on 1 defense against a single robot.

It is legal to block one of the feeder stations, but not both at the same time. It is not a penalty to touch the opposing feeder station as long as you're not also touching an opposing robot that is in the feeder station.

Those indicate to me that climb + dump robots should be defended on their way to the feeder station and back. But after they reach their own auto zone again, it gets really risky to try to defend them unless you think their drive is too weak to push you around.

[Chris is me]

Defensive strategies will be a factor but the days of picking a robot to just play defense and hang for 10 are long gone. There are too many ways to score points, too many robots that can do it, and too many alliances that will need to be taken down with sheer firepower.

I don't expect any "pure" FCSes on Einstein (robots that can't cycle / hang past 10 / do something else well without tipping).

[Ben Martin]

As a FCS team attending the championship, we are hopeful that we could be seen as a great late 2nd round pick for two other strong teams, forcing another robot to be taken out of play to sit in front of us. This could make the game effectively 2 vs 2 except for autonomous and the endgame. In this way, the robot could be a "defender" in that we would prevent another robot from taking action during the teleoperated period.

[MrJohnston]

Originally Posted by Chris is me

Defensive strategies will be a factor but the days of picking a robot to just play defense and hang for 10 are long gone. There are too many ways to score points, too many robots that can do it, and too many alliances that will need to be taken down with sheer firepower.

I don't expect any "pure" FCSes on Einstein (robots that can't cycle / hang past 10 / do something else well without tipping).

I don't think this is necessarily true... Even if you put up a perfect defense against a FCS, it is still going to score at least as many points as the defender. More likely, it will score at least a couple of Frisbees, meaning that its alliance partners (who are effectively playing a 2 on 2 game) will have an advantage. The key will be for the FCS to no mindlessly shoot into a pool noodle. Rather, it must move around a little and find open shots; the worst thing it can do it drain its allies disks by needlessly firing away - and taking them away from undefended cyclers.

I am not saying that there *will* be a pure FCS on Einstein, only that it is a real possibility.

[DjScribbles]

Originally Posted by MrJohnston

I don't think this is necessarily true... Even if you put up a perfect defense against a FCS, it is still going to score at least as many points as the defender. More likely, it will score at least a couple of Frisbees, meaning that its alliance partners (who are effectively playing a 2 on 2 game) will have an advantage. The key will be for the FCS to no mindlessly shoot into a pool noodle. Rather, it must move around a little and find open shots; the worst thing it can do it drain its allies disks by needlessly firing away - and taking them away from undefended cyclers.

I am not saying that there *will* be a pure FCS on Einstein, only that it is a real possibility.

After putting up 7 discs in autonomous, our team will gladly sit in front of your FCS, blocking your shots and your protected feeder station for the cyclers.

This strategy leaves the field relatively uncongested for the alliance defending the FCS, and pretty blocked up for the FCS alliance; even though it removes a bot from scoring, it is still inhibits the FCS alliance beyond just blocking full-court shots.

[MrJohnston]

Originally Posted by DjScribbles

After putting up 7 discs in autonomous, our team will gladly sit in front of your FCS, blocking your shots and your protected feeder station for the cyclers.

This strategy leaves the field relatively uncongested for the alliance defending the FCS, and pretty blocked up for the FCS alliance; even though it removes a bot from scoring, it is still inhibits the FCS alliance beyond just blocking full-court shots.

Remember: Both alliances will likely have robots with a 7 disk autonomous routines. Overall, in elimination, scores should be reasonably close after the 15 second autonomous period. It then becomes a contest to see which team of three can score more points. If you spend the entire 2 minute blocking us, you will score zero teleop points. We will likely score 3-5 disks (down from about 40). Therefore, in teleop, our team would outscore yours. Yes, our protected feeder would be plugged up if we just stood there. However, the unprotected one would not be. Unless you are going to remove yet another robot from offense, our partners could use the unprotected feeder unhindered, so the loss of the protected one would be inconsequential.

[pntbll1313]

Originally Posted by DjScribbles

After putting up 7 discs in autonomous, our team will gladly sit in front of your FCS, blocking your shots and your protected feeder station for the cyclers.

This strategy leaves the field relatively uncongested for the alliance defending the FCS, and pretty blocked up for the FCS alliance; even though it removes a bot from scoring, it is still inhibits the FCS alliance beyond just blocking full-court shots.

That sounds like a decent strategy and would result in a net gain for your alliance. The problem is that all of the teams with a 7 disk auto must fit under the 30” pyramid bar. I haven’t seen one that can then extend up to 84” and effectively block a tall full court shooter. If you are one of those teams and decide to go that route I would suggest you reconsider. I think your floor pickup (which you must have for the 7 disk) would be wasted sitting parked on the opposite end of the field.

[andres20perez]

It really all depends. I saw in the Boston regional that defense was played by few robots. It was mostly 3 v 3 shooting matches, and it wasn't a surprise to see a score near 150. In the NYC regional it was the traditional game of defense and offense, and there the scores rarely made it past 95. Whether or not a robot chooses to play offense of defense matters not as long as the robots out preform the opposing alliance. If its 2 reds offensive 1 red defensive vs 3 blues offensive, as long as the defensive robot does its job well, and outperforms the robot its blocking, the match becomes a 2v2 match all offensive. In that case it matters only which pair of robots performs the best. If the third red robot decided to play offense then it matters only which team performs best. Whether or not robots play offense or defense it all really depends on how well the robots can function by themselves, and with others.

[cmrnpizzo14]

Originally Posted by karomata

Defense has actually probably had one of the largest impacts in Ultimate Ascent. At kickoff, many people thought that because of the placing of the pyramids, it was insane to get anywhere near robots shooting near their pyramid. However, with the rise of full court shooters, defense quickly became a good idea. Taller robots with more material higher up were successfully able to block full court shooters, however alliances without them were knocked out consistantly because of the sheer scoring power of full court shooter. For example, 3173 had an average score of 80 at the Boston Regional. This was not an easy team to beat, regardless of who their alliance partners were, this is shown in 3173s continuous success throughout the elims in both the Finger Lakes and Boston Regional. The only way to have any chance at stopping full court shooters, like 3173 or 148, is to have a defensive tall robot, or as shown in silicon valley, have them fall over on the way to their feeder station.

+1 to that, we could have averaged more without some motor troubles on Saturday that we unfortunately brought onto ourselves. We also noticed that when we had our shooter up to speed our flywheel acted as a gyroscope and helped prevent us from being tipped.

I think that this year defense is whatever you make of it. If you choose to try and actively defend a cycle shooter, you can do it. Maybe not to the extent that a FCS could be blocked (nice job 3003 ) but you could do it. You could also though just try and purely outscore a team because unlike last year whatever you score doesn't become available to the other alliance. Heck, 48 even showed us that you can defend a 30 pt. climber that wants to start climbing immediately after auton.

There have been alliances that have tried to get up in auton and then hold the lead. There have been the more traditional style alliances (469, 3539, 217) that used 2 Offense and 1 Defense. There have been teams that will just outscore you (2056, 1114, 1325).

I think that this is the reason that this year's game is my favorite: There are so many different combinations that you could use to effectively compete.

[Navid Shafa]

Originally Posted by CrazyMohawk

Undefended any good FCS will win this happened at Portland and Seattle.

Not entirely true. Defense was certainly attempted in Seattle, but what really sealed the deal was a strong third robot who could play as a screen for the FCS. 492 kept blockers off our FCS (948) and left them free to pound at the goal.

In Central, our partners 1425 had plenty of defense played on them, but something they were good at was moving and shooting around blockers. They have a really skilled drive team and scored under pressure, unlike most FCS.

FCS's will have a tough time at champs, especially if that is all they do. It is really crucial that you can fill other roles if need be, especially in quals. Floor loading and cycle running will give both 2471 and 2130 a leg up at champs. They might not have the accuracy and point capability that 948 had in Seattle, but the versatility will be just as important, if not more so.

[Adamc4]

Originally Posted by CrazyMohawk

Undefended any good FCS will win this happened at Portland and Seattle.

I seem to remember 3145 blocking 2471 in finals matches but you guys still won despite our efforts

[PayneTrain]

One thing people are going to need to be very aware of in building an Einstein worthy alliance is the possibility that you will run out of game pieces before the match ends. We have seen matches where no discs were left behind the walls, but do not discount the possibility that two really good cycling robots and a floor pickup could score every single disc on their side, in addition to a few on the other side. At that point, you're going to either regret not having any big point climbers or the 8+ points you threw away by not giving discs all of their potential value (3 for white, 5 for red).

An interesting defensive strategy could actually be forcing robots to go to the 2 point goals or abandon pyramid goal disc scoring and actually make a team play itself out of contention for the match.

Scouting at championships is going to be quite the saga, because not only are there going to be a lot of great robots, the "middle class" of robots have had a better year and the Wild Card system makes the event more competitive as well. There are going to be 20 or so robots in every division that will look second-pick worthy on paper but finding the best one for each alliance is going to be interesting.

I think an ideal second pick search begins by establishing a "floor" based on the abilities of the other two partners. The third robot needs to be able to contribute to a perfect autonomous routine by never missing its 3 discs and maybe can pickup the center discs or sit on top of them to keep them away from the other alliance. It would have to "cover" cycles the other robots can't complete every match, and 10 point hang. After that, I can imagine a lot of evaluation on a team's ability to play defense while transitioning between cycles, how well the team could commit to load-zone/blocking defense, man-to-man defense, lane-to-center (what I call "contact zone") defense, and zone pyramid defense (those are the four types of defense to play that I can see, feel free to add or subtract). Driver communication with the coach and coach communication with alliance partners for evolving the strategy in qualifications is also a quality that will be looked at.

(If you would like to see me shamelessly and subtly plug the Team 422's machine qualities as a robot specifically designed to be an ideal second pick at Championships, feel free to PM or just watch my posts all week.)

[Bstep]

Originally Posted by Adamc4

I seem to remember 3145 blocking 2471 in finals matches but you guys still won despite our efforts

Our collector mechanism is very familiar with your..."efforts."

Your driver actually came and signed the bar that got bent by your climbing hooks. It was fun. On a side note, why didn't you use your 30 point climb after that practice round we had with you?

Has anyone noticed how 469 loads discs at the feeder station? They have four discs in there within 2 seconds! The human player shoves two nested frisbees through the slot at once. It's remarkably efficient and a huge contributing factor to their ability to squeeze in seven runs. It's like they only have to bump the feeder station and there are already four frisbees in there.

[rwkling1]

Team 4054 had an amazing defensive robot this year. We went against them one match and we could only make one trip back to the feeder station even though we were one of the fastest robots at our regional. They would shoot for 12 autonomous points and have a 10 point climb and then they would shut down any team on the opposing alliance that tried to score. They ended seeding 5th. Defense is important in UA.