A Recycle Rush Reflection

[dcarr]

Originally Posted by Squillo

Unfortunately, I have not been able to solve this problem. One idea would be to give out points in some manner throughout the season, and then the teams with the most points get to fill, say, 2/3 of the Champ slots (with the rest for award winners; I do think it's good to let, say, Rookie All-Stars, EI and CA winners, etc. go to Champs). I'm not sure how all the points should work, we could debate that all day long, but I would think that being a "2nd pick" for a winning alliance would garner (way) fewer points than seeding high, being alliance captain of a semi-final competitor, etc. Even winning an engineering-based award could count for some points. There would also have to be a way to give single-competition teams as much of a shot as those who attend several competitions. This is really a difficult problem to solve... but that doesn't mean we/they shouldn't try.

District Points System

[Sunshine]

Originally Posted by Squillo

I think @evanperryg has hit the nail on the head. As the best teams get SO much better than the VAST majority, it gets harder and harder for the GDC to meet what I see as one of its biggest challenges - designing a game/engineering problem that provides some challenge for those "top" teams, while still allowing the rest of the teams to meaningfully participate. While "playing defense" is one way to solve that problem, I don't think it's the best. It would be better, IMHO, if the less advanced teams could contribute to offense, but at a lower level. (I happen to think this year was a good year in that regard, I know rational minds do differ in this regard.)

I am not sure what the answer is. Part of me keeps thinking that maybe it's time to split into two competitions, FRC and "super-FRC," while still keeping some connection so that the "ordinary" teams can be inspired by and learn from the "powerhouses." I don't really like the idea of a split, but I'm not sure how to solve the problem otherwise. The "wild card" slot helps a little (at least when a single high-powered team would get multiple 'invites' to Champs, another team or two gets to go), but then there is the whole issue of the "first alliance second pick" that usually gets a MUCH less capable (i.e., fourth-or fifth-string) robot/team into Champs, while the "second stringers" are left behind (sometimes again and again... consider a team that regularly seeds 3rd or 4th - maybe in top 10-15% of all the teams by OPR or whatever else you could use to accurately measure performance - but never is on the winning alliance (because 1 picks 2 and then handily defeats all others). If FIRST wants the BEST robots/teams at Champs (often given as the rationale for allowing #1 to pick #2), then why let a third robot "tag along" merely by the luck of being drawn? (Or worse, because they happen to be "friends" with the #1 alliance? Like the old "popularity contests" in for grade-school student government.))

Unfortunately, I have not been able to solve this problem. One idea would be to give out points in some manner throughout the season, and then the teams with the most points get to fill, say, 2/3 of the Champ slots (with the rest for award winners; I do think it's good to let, say, Rookie All-Stars, EI and CA winners, etc. go to Champs). I'm not sure how all the points should work, we could debate that all day long, but I would think that being a "2nd pick" for a winning alliance would garner (way) fewer points than seeding high, being alliance captain of a semi-final competitor, etc. Even winning an engineering-based award could count for some points. There would also have to be a way to give single-competition teams as much of a shot as those who attend several competitions. This is really a difficult problem to solve... but that doesn't mean we/they shouldn't try.

Although interesting, you are getting off topic. I hope this does not cause us to get off track. Purhaps a new thread with this info would be appropriate.

[Squillo]

Originally Posted by dcarr

District Points System

Wow, they should do something like this for Regionals, too! I love it.

Sorry for going off topic. We now return you to your regularly-scheduled endless discussion of whether Recycle Rush is boring or not.

[Green Potato]

I saw this challenge as a throw back to the old days. We saw that coming.

Honestly, whether or not one misses the "old days" of FIRST may indicate whether they like this game. Sure, this game isn't a copy of the old, but people who generally liked the older games were also the kind of people who liked this one. I saw this as I was walking around and talking with people both at Palmetto and Virginia regionals. I personally don't like this game so much due to it being more of an engineering match than a real competition. ( I actually suggested that this game be called a "challenge" instead), but different people have different opinions.

[Ginger Power]

Originally Posted by Siri

I agree completely, but I don't think the assertion is really "RR is inherently boring". I think it's that it's statistically more likely to suffer from the boredom you've identified (e.g. Wave----->everybody else at the event). The point spreads are nuts this year. When they're not, it's not so bad (as long as you're good with watching uniformly-ish good or bad box stacking).

Yes

[MrForbes]

We kind of accidentally made a "can specialist" robot, for your entertainment.

Every match is different for us.

[Caleb Sykes]

Originally Posted by Squillo

If FIRST wants the BEST robots/teams at Champs (often given as the rationale for allowing #1 to pick #2), then why let a third robot "tag along" merely by the luck of being drawn? (Or worse, because they happen to be "friends" with the #1 alliance? Like the old "popularity contests" in for grade-school student government.))

Emphasis mine.

I disagree with a couple of things in your post, but this one flat-out confuses me. If the best two teams at the event are not picking the team that they think gives them the greatest chance at winning the event, then I have absolutely no clue how they managed to build the two best robots at the event.

I have never seen a case of the highest seed picking a team just to let them "tag along" for the ride to champs. My guess is that you have seen a case where the one seed picks a team they are familiar with because they trust the team to be useful contributors to their alliance. For example, there are generally a couple of teams at each event that I don't trust, primarily based on past experiences with them. Even if one of those teams had a slightly better robot than a team I am familiar with, I would want to partner with the team I am familiar with because I know that we will work well with them.

[evanperryg]

Originally Posted by Squillo

If FIRST wants the BEST robots/teams at Champs (often given as the rationale for allowing #1 to pick #2), then why let a third robot "tag along" merely by the luck of being drawn? (Or worse, because they happen to be "friends" with the #1 alliance? Like the old "popularity contests" in for grade-school student government.))

Yes, I believe that an event where one robot is so far beyond any of the others is upsetting to the other teams. I've seen plenty of that at certain regionals. However, I think you miss the importance of a second pick. I won't speak for any other team, but as my team's lead strategist and the primary guy behind the picklists, I'm willing to lay out the importance of some of my team's second picks over the last few years:

2014 Crossroads: 2338, 4085, 2358: We were in a position at crossroads where we knew we were not going to win the event, but I'm not one for giving up. 2358 was vital to this alliance by providing a robot that could inbound, bounce assist, or truss shot quickly, while their maneuverable drive and ~50" height made them good for defending against 1024.

2014 Midwest: 2338, 1732, 3936: Almost everyone at Midwest knew that 1625 and 2451 would win it effortlessly. I was certain they would run 1-assist+truss cycles and use their last bot as a defender. Realizing this, I knew the only way to have a chance against them was to get the 2-assist+truss cycle down fast. So, 3936 was perfect for us because their robot was just a collector on wheels, and a pretty quick collector at that, and their drivetrain was strong enough to get in 1625's way while they tried to get a truss shot.

2015 Wisconsin: 2338, 2077, 167: This one's pretty simple. We knew we needed to put up 5 6-stacks capped if we wanted to beat 2826 and their 28 point auto. since 2077 and 2338 were both landfill bots that needed upright cans, we needed a bot that could both tip up our center-pull cans, and could have a ramp for 2077 to use. We had worked with 167 in a qualifier, and knew they could tip up cans quickly, and they were a joy to work with.

2015 Midwest: 2338, 1756, 171: 1756 tips over 2 (usually 3) cans in auto, and our center can auto usually leaves one can tipped over. This time, too, 2338 and 1756 were both landfill bots that preferred ramps. So, we needed a robot that could upright cans, and be light enough to have 2 tethered ramps. Through all of quals, no team had proven themselves to be a really good can tipper, so we searched for some teams that could be light enough for the ramps, and suggested those teams work on can tipping. 171 proved themselves to do it the fastest, and they became a valuable alliance member. Even when 1756's collector broke and we weren't doing much better, 171 was able to score a few points that helped keep our average high enough to advance to finals.

A second pick doesn't have to put up massive scores to be valuable. A really good second pick can be the deciding factor in an alliance's success. The fact is, high-scoring teams put a significant amount of time and effort into deciding what their second pick is. Even if it sometimes looks like a power alliance's second pick is a lottery winner on a free ride to champs, that team was chosen for a good reason, and they are pulling their weight just as much as (and sometimes more than) their captain.

[Aroki]

Lets not forget that at most regional events, the last pick in the draft is above average. A game where the tops teams are able to question whether an above average robot is more of a liability than an asset is completely unbalanced.

[dcarr]

Originally Posted by Aroki

Lets not forget that at most regional events, the last pick in the draft is above average.

This may be true at larger events with a deeper field, but I'm not sure I'd qualify that as "most."

[Aroki]

Lets not forget that at most regional events, the last pick in the draft is ranked above average. A game where the tops teams are able to question whether an above average ranked robot is more of a liability than an asset is completely unbalanced.

That said, I think the true issue with this year's game lies with how it plays with low tier teams. Other people have talked at length about how the points with RC's favors top teams and the lack of defense removes the major role of weaker teams so I will instead talk about the ranking method in this year's game.

In past years, if six weak teams were playing each other, everyone can get excited for a 3:2 finish. The winning alliance feels good about clutching out the victory and maybe a team scored a last second goal or something and is feels like the contributed to the alliance effort. Objectively speaking however, all six of these robots are weak and none will advance to eliminations but they are still able to be happy about getting a win. This year however, with average match score deciding ranking, there won't be a single one of those six teams walking off the field feeling good about how their robot did.

Originally Posted by dcarr

This may be true at larger events with a deeper field, but I'm not sure I'd qualify that as "most."

Edit: On average, regional events have 51 teams this year.

[MichaelBick]

Originally Posted by Aroki

Lets not forget that at most regional events, the last pick in the draft is above average. A game where the tops teams are able to question whether an above average robot is more of a liability than an asset is completely unbalanced.

I think you mean above median, not above average

[Rachel Lim]

Originally Posted by Aroki

Lets not forget that at most regional events, the last pick in the draft is above average. A game where the tops teams are able to question whether an above average robot is more of a liability than an asset is completely unbalanced.

The 24th best team at a regional isn't usually that different from the 25th best one, or the 30th best one, and so on. Generally speaking, it's quite obvious who the top 2-4 teams are, and then the next 4-8. But beyond that, many teams are quite similar. Being "above average" may only really apply to the top 8-12 teams to the event, and "below average" to the absolute bottom few. "Average" means very little; between each of those groups is a huge drop.

What truly matters is how much those teams can contribute to their alliance. And being in the 99th percentile (i.e. ranking first) does not mean scoring twice as many points as being in the 50th percentile (i.e. being average). In many cases it is 5, or even 10 times as much.

That said, I do agree that top teams this year are even farther above the average than top teams last year. However, in both years, the top teams were very, very far above the average--or 24th best--team. The reason 2nd picks last year were more crucial was because they were needed for assists and for defense.

[Siri]

Originally Posted by Caleb Sykes

I have never seen a case of the highest seed picking a team just to let them "tag along" for the ride to champs. My guess is that you have seen a case where the one seed picks a team they are familiar with because they trust the team to be useful contributors to their alliance. For example, there are generally a couple of teams at each event that I don't trust, primarily based on past experiences with them. Even if one of those teams had a slightly better robot than a team I am familiar with, I would want to partner with the team I am familiar with because I know that we will work well with them.

I'll contribute to this as well. I don't know that FIRST wants the "best" robots at Worlds. They seem to want a good cross-section that can also make for exciting and meaningful world championship matches. Does it take luck to make participate in that as a 2nd picks? Some, and 2nd picks are likely less consistent with Saturday afternoon appearances, particularly when including Worlds. But 1640's current high school seniors have won back-to-back MAR Championships in 2012/3 and back-to-back Worlds divisions in 2013/4 as second pick.* I don't know anyone who argues that you can go along for the ride in district or division championships, but if that was just luck of the draw, it was about 4.7 million to one. I'd be interested to know how common similar repeats are.

As for doing that in this game, I'm sure 2nd picks are still valuable. I think the root of most draft complaints this year isn't that there aren't 2nd pick niches that teams can make themselves valuable for, it's that the niches themselves are upsetting to people. The most divisive example being a "boat anchor" to teams' ramps, but other cheesecake is also controversial this year.

*We didn't personally know our alliance teams beforehand (we knew the MAR guys of course, but so did everyone else there). It was our first time picked by all of them except 341 in 2012, who had picked us second to win Horsham a few weeks prior. EDIT: Actually, we won Philly 2011 (and then Newton 2013) with 303, but in 2011 we were the AC and they were the second pick. Also, this makes me feel old in that all the strategists on both teams were too young to remember it.

[dodar]

More often than not, 3rd robots at regionals are niche teams.