Week 5 Impressions of the Rack. The Conclusion of the Games.

[hayakuneko]

concerning the end of each match, I would have to disagree. At my first regional (LA) lifting did become a match decider. There could be no blue ringers on the rack and 4 red on the rack, but blue still won with their 60pts.

However, in SD, I quickly realized that this lifting bonus had almost no effect on the game depending on the alliances. Before the pairings we lifted 330 in hopes of lifting every match during the elim rounds (like 1717) but it turns out our third alliance robot wasn't even able to get on our wing despite our 1" ground clearance.. haha. anyways, in the final round, we figured that because lifting for 60 pts would only tie us with 1717 and their lifters, we would have to outscore them on the rack

and we even took a strategy from 330, but if your team hasn't figured that strategy out yet, then watch 330 at Atlanta (haha i feel like i'm advertising for you guys)

[Lil' Lavery]

Weeks 4 and 5 say the beginnings of what I think Championship will look like (at particular regionals). Teams have figured out how to beat ramps, and beat them consistently. Many winning alliances still have ramps as an options, but no longer rely on them (and/or only use them for 30 a match). Spoilers are becoming bigger and bigger (when rows get full, spoilers are worth much more than ringers). Autonomous is now important (1114; 3/4 Vegas SF alliances scored keepers often; keepers determined several matches in 111s march to the WMR finals; etc.), especially in close and/or defensive matches. The rack is beginning to fill up (watch Waterloo, Boston, Toronto, Vegas, and Palmetto for proof). The winning alliances can either grind it out defensively, or pour on the points, depending on how their opponents play. Spoilers scare people, a lot (the "skunkworks spoiler" might just be my new favorite strategy ). Refs are cracking down on ramps and defense. The hybrids are shining (254, 330, and 1114 each won 2 regionals). The multi-threat alliances have come out firing. Welcome to Rack N' Roll, here's where the real fun starts.

[shamuwong]

I can appreciate the whole 2^n scoring thing this year, it makes fast and efficient scoring extremely important, as well as effective defense to allow you to establish your rows.

I'm also happy with the rack and the movement required. This was the year for creative drive trains, as opposed to last year, where all you needed was a simple drivetrain with brute strength and a turret. Crab drive was a huge advantage for our team, along with other high scorers, such as Wildstang and Beatty. Not required, mind you, great scoring was achieved by other "conventional" drive robots, but it was definitely a great advantage for us, and many others, based on how many I saw this year.

However, the randomizing algorithm - There are a lot of strong teams out there with high numbers, and weak teams with low numbers. It's unfair to judge a team based on their number. At least in previous years, when we had less than favorable alliances, we had nothing to blame but luck. There are a lot of other things I could say about the so-called "randomizing", but they'd all be euphemisms for "utter crap". So I'll stop there.

Automode was also given too little value. So it's a ringer you can't spoil. There are plenty of others that you can that will do the job just as well. Plus, we have a rotating and translating light and rack, along with legs that swing wildly. Despite it's rather minute value to the game, its difficulty made ringers scored in auto very satisfying.

The kit of parts also proved to be another design challenge. Nothing like having to deal with sub-standard parts and unneccesarily large/heavy components to stimulate the engineering mind.

I hope FIRST develops more games where more complicated drivetrains are advantageous, in more ways than just ground clearance. Wildstang's 2004 robot comes to mind. I see them as being just as important as manipulators, if not more so.

All in all, a well-done game, except for the dodgy alliance algorithm.

[rjmah]

The Autonomous ringer carried no bonus for a reason this year. Consider an alliance that can score 6 ringers on average. The Autonomous ringer doubles that score at the end of the round. You shouldn't think of the original value but it's value at the end.

We removed our camera as a quick fix to get to our weight. In retrospect, we should have kept it since we might have been scoring autonomous by our third match at GTR. The camera programming had been tested on our last year's robot.

[Tom Line]

#1 Autonomous was not worth enough
#2 Ramps are worth too much, or ringers aren't worth enough
#3 Defense rules the day if it is played

Even at Western Michigan, we saw that it is very easy for an "ok" robot to shut down an "excellent" scoring bot. This competition can very easily turn into a snooze fest while 3 ramp bots play defense on the other teams scorers, then ramp for the win.

Nationals will see some high scores, but it's also going to see insane defense played by some of the top end bots there with good drivetrains. It's not even necessary to push the other bot in the corner - just stay between them and the rack and 90% of the offensive bots will not be able to score.

[IndySam]

A defensive match isn't a snore fest if you are the defender.

To me defense is fun, even when I played little league football I always liked playing defense more than offence.

[Travis Hoffman]

Originally Posted by Lil' Lavery

Refs are cracking down on ramps and defense.

I didn't see that at my Week 5 regional in GTR. I didn't see the rack filling up in any of the elimination rounds 48, 379, and 1006 were in. We were a hinge and a blown tread away from the finals.

How does a ref crack down on a ramp, anyway?

It's easy to put up huge tube scores if no one plays defense on you (especially if you don't have a full alliance against you - many alliances could score 300+ points if they were only facing one robot... ). If you want the snoozefest of 2005 and want the top offenses alliances to waltz to a boring victory, go right ahead and leave them alone. ZZZZZZZZZZZzzzzzzzzzzzzzz......

This is a call out to all teams at the Championship - play defense. It works. Play it fairly and smartly and within the rules, and it works. Don't be afraid to face up against the giants, stare them in the face, and get down to business. No one is entitled to a cakewalk. Make them earn those tubes!

I LOVE Rack n' Roll the way I've seen it played. Go GDC!

[Steve W]

Defense is not ramming other robots. If you watch any of the matches that Team 48 at GTR you will see excellent defense. I hate to blow Team 48's horn (man do their heads swell) but they played the best defense next to Team 4 that I have seen this year. They play smart and clean not rough. A lot of the time they just kept between the Rack and their opponent. If anything the opponent was ramming them trying to get by or push them to the rack.



Disclaimer - this post was in no way meant to allow the swelling of Travis's head. Man we all know how big it is already.

[Lil' Lavery]

Originally Posted by Travis Hoffman

I didn't see that at my Week 5 regional in GTR. I didn't see the rack filling up in any of the elimination rounds 48, 379, and 1006 were in. We were a hinge and a blown tread away from the finals.

How does a ref crack down on a ramp, anyway?

A ref cracks down on a ramp by using strict interpretations of the support rules and carefully examines to see if they are in contact with any field elements. I'm seeing more and more situations where ramps aren't counting because of it. It only takes one real defender to turn a match into a defensive grinder (example-West Michigan), but the refs are being a lot stricter in recent weeks, and teams are figuring out how to score through defense with moderate success. Multi-threat combos are getting more coordinated, more common, and harder to completely shut down (look at Vegas).
Your alliance in GTR is a slight exception (your opponents still scored 36.5/match, not terribly low), but look at the teams. 1006, 48, and 379. Defence, Defense, and Defense. It was so defensive that 1006 was assigned to offence! (Where 1006 is, I apply the Canadian spelling ). Of course your matches are going to involve a more defensive approach by both alliances, and this year, defense and ramps go hand in hand. But the ultimate victors of GTR was a triple-threat offensive alliance, 1114, 2056, and 176.
Champs in Vegas, 1425 (scorer), 254 (scorer/ramp), 1983 (scorer, spoiler, defense).
Champs in Palmetto, 1319 (scorer/ramp/defense), 832 (scorer), 342 (scorer).
Champs in Long Island, 229 (scorer), 234 (scorer), 545 (ramp).
Champs in Waterloo, 1114 (scorer/ramp), 2056 (scorer), 771 (defense).
Champs in Boston, 126 (scorer), 69 (scorer/ramp), 125 (scorer).

Of the 9 teams that won 2 regionals, 7 can score tubes (and proficiently). 2 are exclusive ramp/lifts, and 4 are hybrids (although 254 only had their ramps for one of their two regional wins).

[Dave Flowerday]

Originally Posted by Lil' Lavery

(1504, 111, 302 losing WMR F-3 on outside bumper zone penalties)

Maybe this is what you meant to say anyway, but the 2 penalties our alliance had in F-3 at WMR were home-zone violations, not contact outside the bumper zone.

[Lil' Lavery]

Originally Posted by Dave Flowerday

Maybe this is what you meant to say anyway, but the 2 penalties our alliance had in F-3 at WMR were home-zone violations, not contact outside the bumper zone.

Woops

[dtengineering]

Having competed at both first weekend (Portland) and fifth weekend (Toronto) regionals, I think it is perhaps not unfair to suggest that there were two very different games being played at the two tournaments.

In Toronto, for instance, our alliance scored 276 points in our first match and would have broken 300 had we made it up our partner's ramp at the end of the game. That was more than the combined total of all our matches in Portland.

In Portland ramps were quite dominant. In Toronto, the rack was quite dominant. In Portland, it seemed relatively easy to defend against all but two or three teams... and keep them from scoring ringers for the whole match. In Toronto, defence just slowed the scoring down a bit -- not to say that there wasn't great defense being played... particular kudos to 48 as mentioned above for playing strong, clean D -- but there were so many teams at Toronto that could punish you on the rack that it was hard to stop them all.

So when discussing the game of Rack'n'Roll, your impressions will be quite different based on what game you saw being played. I wasn't that impressed by the game on the first weekend (well... okay... the finals at Portland were pretty good) but by the final weekend it was evident that the game had evolved to be the exciting, dynamic sport that the GDC must have been hoping for.

My one real concern surrounds the match scheduling algorithm. I have already posted my concerns regarding this to another thread, so I'll leave it at saying that I really don't like being able to walk into a match, identify the "middle third" of the teams by number, and saying, "well... we'll NEVER have them for an alliance partner."

Jason

[Travis Hoffman]

Originally Posted by dtengineering

In Portland ramps were quite dominant. In Toronto, the rack was quite dominant. In Portland, it seemed relatively easy to defend against all but two or three teams... and keep them from scoring ringers for the whole match. In Toronto, defence just slowed the scoring down a bit -- not to say that there wasn't great defense being played... particular kudos to 48 as mentioned above for playing strong, clean D -- but there were so many teams at Toronto that could punish you on the rack that it was hard to stop them all.


Jason

Jason:

I think the key thing here is to understand that an elimination alliance needs to be willing to employ TWO robots for defense in order to shut down a quality opposing alliance. One isn't enough to keep the scoring down. I won't go into details (do your homework!), but this strategy has led to a finalist, a championship, and a semifinalist finish at the three regionals we've
attended.

People may be all like "OMG 2 defensive robots? Like, that's totally bogus! Gag me with a spoon!", but it works. I feel we've got the formula for success down, and we're looking forward to trying it at least 2 more times this year, at the Championship and at IRI. For the mere sum of $10,000, you too can own the secret to this formula. Please send all checks to.........

[Travis Hoffman]

Originally Posted by Lil' Lavery

Your alliance in GTR is a slight exception (your opponents still scored 36.5/match, not terribly low),

Much of the contribution to that score was ramp-related. Which, as we all know, is kinda hard to defend - we're still working on ideas for that.

But the ultimate victors of GTR was a triple-threat offensive alliance, 1114, 2056, and 176.

That won the finals with a double ramp, ironically. The opposing alliance kicked up the defense a notch.

Sigh, I wish we had gotten the opportunity to put up defense alliance vs. offensive alliance in Mississauga and see where it would have gotten us. I have a feeling we still wouldn't have won, but it would have been interesting. Ah well, there's always Atlanta....and beyond.

[Lil' Lavery]

Originally Posted by Travis Hoffman

That won the finals with a double ramp, ironically.

They could have won with only 30 ramps points in the 2nd match, or none in the first.