Week 1 Impressions of Rack 'N Roll

[Priyadarshy]

yeah those three matches showed how important the rack was and the ramp 612 removed their ramp for the tournament to improve visibility and it paid off but the rack is crucial the ramp WAS the outcome in qualification matches but in the tournament it was a whole new game

[Eldarion]

Originally Posted by ldeffenb

But then, we also changed the lens on our camera to a 90 degree wider angle.

That's the difference--we used the stock lenses with about a 60° field of view. With the 60° field of view, there is one rack orientation in which neither light is visible, even when you are pointed directly at the center of the rack. Darn, that was frustrating! (Of course the lights were always in our field of view after the arm broke! )

Originally Posted by ldeffenb

I disagree with 8 lights making it easier. That would have just made more cases with multiple lights in view which (at least in our implementation) is a much more difficult autonomous situation to solve.

Oops, I forgot for a second that everyone else is using the CMUCam. We have a vision system that is capable of seeing up to 6 objects simultaneously with ease (www.falconir.com -- I know it's just a teaser, that will be 'fixed' in the coming months.) If you do a search for 'Team 81 progammers teaser explained' here I think you can find my other threads describing the first hardware rendition of this system.

[dangerousdave]

Originally Posted by wilsonmw04

There was a count down during each match. It was posted in the middle team's control area for each alliance. It counted down the two different periods of each match. Now if you are only talking about the big screen, then you are correct.

Yes, I was talking about the big screen being viewed by the vast majority of people. We were unable to see "exactly" how much time was left in a match. We were left to guess by looking at a green line.

[platypus]

Originally Posted by Noah Kleinberg

To the people who think autonomous mode isn't very important this year, don't think of it as two points, scoring is exponential. especially if more than one robot on your alliance scores during autonomous, you're at a pretty big advantage for the human control part of the match.

I have to respectfully disagree here. You've made the general statement that I imagine most justifies autonomous, but at the same time, you've missed the point that I've tried to make.

Yes, autonomous is a good thing to have -- after all, it does score points, and when is scoring points a bad thing. The point that I have tried to make is that it doesn't score enough points to be worth the effort that most teams would have to put into it.

It's a difficult autonomous to code for, and most autonomous modes are not 100% successful, with this year being no different. The reality is that autonomous integration has a serious opportunity cost for all but the most efficient teams, and an unrealiable chance to score an extra ringer (since spoilers aren't used, that's what it is) was hard for most teams to justify when they were being delayed by typical mechanical setbacks and even atypical mechanical setbacks (think Banebots). Even the software teams have better things to do, like program a particularly complex control system.

As a good yardstick for the value of autonomous, compare this year to last year. Last year, the team that won autonomous won the game at least 75% of the time, and that's a conservative estimate. There isn't much data to make similar estimates this year because of the lack of autonomous scoring, but I can say from observation that it really didn't make much difference. Maybe in the national eliminations, alliances will score 3 in a row in automonous and it will matter, but realistically, the GSR eliminations sported an alliance that once scored two keepers, only to lose to the steamroller of the top alliance. Last year, if two teams on an opposing alliance had unloaded their clip of nerf balls for 40+ points + 10 more points for winning autonomous, it would have taken an act of god to beat them (or at least a DQ from one of them for placing a robot part into the side goal).

Autonomous is always nice, but it isn't worth what it needs to be worth this year.

[Nawaid Ladak]

That last QF match at VCU was probally the best match i saw so far this weekend.
I think of this as football.

There are three parts to the game

Offense
Defense
Special Teams

there will be alliances such as

The Indianapolis Colts: The best defense is a god offense, never back down and keep on filling the rack (sort of like the winning alliance at NJ)

the Baltimore Ravens: Defense is going to be key, you stop your opponent from doing anything on the rack, with minimal offense, you are sure to win the match

The Chicago Bears: Special Teams, in this game the special teams are pretty much the the autonomous and ramp part of the game, couple that with the spoilers and you fill the missing peace to a great alliance partner. someone who can play some of both offense and defense

so what do we learn from this

To succeed in this years game you must have, A KISS B have alliance partners who specialize in your weaknesses, for example if your a defensive robot, then be sure to have a team who can ramp and or atonomus, plus a team who can score,

that i think will determine the championship this year

Another thing i noticed, and i think this is for every year. defense is disregarded at regionals, we all want to see offense, like a famous saying goes "offense brings the crowds, but defense wins games" or in this case, regionals.

Teams in the top 8, (most likely offensive matchines) will need to do A LOT of scouting and choosing, because one wrong move, could cost your team a trip to Atlanta in April.

[GaryVoshol]

Originally Posted by platypus

Yes, autonomous is a good thing to have -- after all, it does score points, and when is scoring points a bad thing. The point that I have tried to make is that it doesn't score enough points to be worth the effort that most teams would have to put into it.

True. I haven't seen any of the games yet, but right from the beginning I thought autonomous was the least important aspect of this year's game. Yes, you can have that non-spoilable keeper, and if you arrange your ringers properly so that the keeper is in the middle of a row, it gives you a boost of points. But those same points could also be scored in teleoperated mode. I know a ringer could be spoiled, I know you would "waste" 15 seconds of play time without autonomous. But a good fast scorer, especially if it is a spoiler-remover bot, can overcome the lack of autonomous. Teams without a spare robot at home to practice with, and code for, probably made the best investment of time in perfecting their ramps and teleoperated arm operation.

As a good yardstick for the value of autonomous, compare this year to last year. Last year, the team that won autonomous won the game at least 75% of the time, and that's a conservative estimate.

I've heard a variety of statistics and guestimates as to the value of last year's autonomous. Part of the reason the autonomous winner won the game so often is that the way to score points won in both portions of the match was exactly the same. If you could put 10 balls into the air in autonomous, you could also score a lot during teleoperated, assuming you had an effective way of reloading. I also think that teams that use the light to aim during teleoperated had a bigger advantage last year than this, because last year the target stood still. This year you can use the lights to guide you during teleoperated, especially while on the other side of the ramp, but most teams will still have driver control over placing the ringers, not pure programming control.

The value of this year's autonomous is more similar to that in Triple Play. There, we had a truly hard task which gained you very few actual points. If you capped the center goal with the vision tetra, you got points for scoring that tetra plus the two that were given to you on your corner goals. You didn't automatically get any row points - just like this year the keeper by itself is only worth 2 points. It's what you do to augment the autonomous scoring that makes the biggest difference in your final score.

[Andrew Blair]

The trouble is, the only reason a team would spoil your row is if it is seven or longer in length- and therefore extends into their viewable portion of rack. A keeper only protects that row if it is placed where the opposite alliance will want to spoil- or in other words, the HARDEST POSSIBLE SPOT TO PLACE IT! The opposite side of the rack.

The keeper's special advantage is lost on this year's game. You score a keeper, and you're just one ringer up effectively. Nice, but not super special.

[Ahnxlazyman]

I think only 2 spoilers were ever "used" in VCU, the first one was on Friday, In which a Human Player Successfully Threw it onto the rack, the only problem, It landed on an empty spider leg making it pretty much useless, the other, was a robot trying to get a SPoiler on their but they failed (out of time, or fell out of their grasp)

I saw A lot of Ringers accidentally being tossed onto robots...They would be tossed by Human players Towards the robots but would be caught by a protruding Objects like Flag holders or tall arms, It made the robot Useless (offensive-wise)
I didn't enjoy that (even when it wasn't our own team)

Ramp Bot's are nice (especially when they can raise 2 up 12") but they take the fun out or the match, all they do is drive and line up in their zone, drop their ramps, and sit for the last 1:50 seconds, It's rather boring.

And I also have a question, In autonomous, do alliances that scored the most points in autonomous get an extra points for winning that period?

[EricH]

Originally Posted by Ahnxlazyman

And I also have a question, In autonomous, do alliances that scored the most points in autonomous get an extra points for winning that period?

No. I wish they did.

[Azure]

Originally Posted by Ahnxlazyman

I think only 2 spoilers were ever "used" in VCU, the first one was on Friday, In which a Human Player Successfully Threw it onto the rack, the only problem, It landed on an empty spider leg making it pretty much useless, the other, was a robot trying to get a SPoiler on their but they failed (out of time, or fell out of their grasp)

Team 345 tried to use a spoiler in our quarter-finals match and they nearly were successful. Also, when I was doing some scouting on Thursday, I saw a couple spoilers being used in the practice matches.

lol talking to peeps from my team on Chief Dephi is kinda weird, but whatev.

[Frinkahedron]

Originally Posted by Azure

Team 345 tried to use a spoiler in our quarter-finals match and they nearly were successful. Also, when I was doing some scouting on Thursday, I saw a couple spoilers being used in the practice matches.

lol talking to peeps from my team on Chief Dephi is kinda weird, but whatev.

That was the only serious spoiler attempt I remember from VCU, would have destroyed our 100 pt string if they got it on in time.

[teenmisfit]

he ramps will win championships. Having a ramp/defensive bot will keep the other team trying to score tangled up but at the end those ramps mean 60 point. team 1824 was solely a ramp/defense bot and he won us sooo many round just because of the ramps. everyone gets sooo good at scoring tubes in the finals it comes down to bonus points to win rounds. You need a well rounded alliance to win. not a solid offensive won. I will stick by my word that the two teams i went to finals with at BAE (175 Buzz, and 1824) was the best alliance and teamwork i have seen in four years of first. Nothing equates to how we all did our job flawlessly everytime and us mere "kids" showed professionalism and sportsmanship both to the opponents and our alliance.
We may not do that good at nationals as we did at the BAE but Buzz and 1824 will. They are valuable, cooperating teammates that just work everytime.

we were the only team i saw at BAE that effectively could put on a spoiler and take it off. we did so with ease (but only in the practice rounds on thurs.)

[+()c|D]

I feel that the idea of getting all 8 around is not that hard.

Also, rampbots are a HUGE deal in this game... you could not score any ringers and still win with a good rampbot on your team!!

[wolfj]

I think that too many points are given for robot elevation. It seems as though an alliance that gets any bonus is a winning alliance. This seems to be a common opinion. It takes six in a row to cancel out a good rampbot's lead. But, I hate to criticize, since the FIRST game designers are doing an amazing job, and probably have creative capabilities we could only dream of. (But then again, we do build robots.