50 Pt. Climbs

[cmrnpizzo14]

Originally Posted by Brandon Holley

At Orlando we chose 4451 as our first pick from the 2nd seed (declining the 1st seed, and passing up several other shooting robots) because based on our scouting data, they had the highest probability of scoring 50+ points a match. The reason for this was their near flawless execution of just getting to the top every match. They did not disappoint us as they climbed to the top in all 9 of our elimination matches.

Why weren't there more teams like this? No offense to 4451, they are a great team. However, if a rookie can build a robot that can deliver consistently enough to do this, why didn't more teams? Someone please correct me if there were but I don't remember any other bots off the top of my head that did that.

[Andrew Schreiber]

Originally Posted by cmrnpizzo14

Why weren't there more teams like this? No offense to 4451, they are a great team. However, if a rookie can build a robot that can deliver consistently enough to do this, why didn't more teams? Someone please correct me if there were but I don't remember any other bots off the top of my head that did that.

How many teams at Orlando had already attended an event? Not to take away from 4451 but they had already attended a regional and had a chance to shake down their bot.

[EricDrost]

Originally Posted by Andrew Schreiber

How many teams at Orlando had already attended an event? Not to take away from 4451 but they had already attended a regional and had a chance to shake down their bot.

4451 was also nearly flawless at Palmetto, a week 1 event. These guys are the real deal, rookie or not. Without seeing their future robots, I would put money on them being a top 20% team every year from here on out.

[AdamHeard]

I'll likely catch some flak for this, but a lot of teams don't know what they don't know.

A 30 point climber that was even halfway sort of kind of fast was not a system you could just BS the design on. You had to have some idea of what you're doing. Double this if you wanted it to be easily aligned, and work every single time.

Without an ME (or a tradesmen with some real experience) completing such a system was going to be real difficult. Some teams probably did without this, as it is possible to guess right on things sometimes.

Our scouting data had the 24th team in our division averaging ~50 points. This was their average though. A dedicated 50 point climber could not possibly average their maximum (unless they were literally flawless). So at the champs level it wasn't viable.

At regionals if a team averaged 30-35 points of a 50 point climb, they wouldn't miss elims at any event (most likely) and would likely be a 1st round pick/seed.

[BrendanB]

Originally Posted by cmrnpizzo14

Why weren't there more teams like this? No offense to 4451, they are a great team. However, if a rookie can build a robot that can deliver consistently enough to do this, why didn't more teams? Someone please correct me if there were but I don't remember any other bots off the top of my head that did that.

I think these teams got lost in the giant group of teams who attempted to 30 point climb, had the mechanisms but didn't use them due to weight, complexity with a system as a whole, or abandoned them at the last second.

Our team was close to falling into this catagory when our original shooter/climber combo didn't integrate by 1/2in. at our first regional. We went with the climber for the rest of the event because our shooter game wasn't that great and improved it for Pine Tree (along with cutting our climb time drastically by 65%) to become an auto, 3-4 cycles, climbing robot. I know of a lot of other teams we encoutered this season who either had a 30 point climber on their robot or the space was clearly allocated but never used. Other planned on making or integrating their climbers later on in the season but the effort wasn't worth it if you could pull of 1-2 cycles in the final 20-30 seconds of a match.

[Chris is me]

Originally Posted by Thunder910

Let's think from a picking perspective. At Connecticut, there was a very good, consistent Climb/Dump, Team 2170.

2170 is a very good example of a climber dumper. In my opinion, their fatal flaw wasn't their design or strategy so much as their choice of regional. Connecticut is an incredibly deep regional, with several three cycler alliances forming and 100+ point scores being a requirement to break the quarter finals. I'd argue that if a team like them was available at the Championship, they had a respectable shot at being selected for the elimination rounds in the weaker divisions. They were even faster than 4451, had a drivetrain suitable for playing defense, and with the right driver and strategy could have robbed two cycles from an alliance while also scoring 50 points a match.

We talk about 4451 as being one of the most successful climb/dumps, but they never won an event(you can srgue that they should have, but the fact remains that they didn't), and they weren't picked at Champs because of this.

If we're defining success as "winning an event", I guess 67 didn't have a very successful year in 2013...

[Kevin Leonard]

Originally Posted by Chris is me

If we're defining success as "winning an event", I guess 67 didn't have a very successful year in 2013...

I think for HOT, anything less than a world championship is a less-than successful year.

[Jon Stratis]

We found, generally speaking, that climbing and dumping was great for qualifications, but sucked for eliminations.

Simply put, in qualification matches there was so much randomness in the match-ups that, most often, climbing and dumping successfully would guarantee a win.

However, once we got to eliminations, the alliances were tougher, and often that 50 pts wasn't enough help to the alliance to guarantee a win. When you consider that, at the regional level, most alliances were set up as two scorers and 1 defender, that 50pts is essentially half of the alliances ability to score. When scores go over 100 points, you start to have a problem with not pulling your weight at just 50 points.

All that said, climbing and dumping was a great way to be picked. Either high-ranking teams figured they could handle enough frisbees that they wanted someone complimentary climbing, or lower ranked teams took a gamble on the potential big impact a climb and dump could have.

As a result, we did great at our first regional, seeding first and being picked first. We did worse at our second regional, seeding low and yet still being picked in the first round. In both cases we lost in the elims because our alliance just wasn't capable of putting up enough points to win.

[D.Allred]

Originally Posted by cmrnpizzo14

My question, mostly to teams that attempted this but also to those who considered it, is why did you find this strategy successful or unsuccessful?

We found the strategy very successful for multiple reasons.

First of all, we owe a great debt to 1114’s Effective FIRST Strategies guide. Our goal before the season was to find a simple and effective approach to scoring following 1114’s golden rules. Apparently the Game Design Committee had a different plan.

We knew a disc shooting robot would have a higher scoring potential and therefore be a “better” strategy. But we didn’t know if we were capable of building that machine our first year. Climbing and dumping would still give us a very high score without the strain of prototyping disc shooters and queuing systems. We dedicated all our efforts to climbing hoping the dumper would be easy. (As it turned out, dumping was a bit trickier than I first imaged.)

We knew the scoring potential of our strategy put us in almost certain position to make regional elimination rounds. We also felt climbing and dumping would be a great complementary robot to a shooter making it harder for the opposing alliance to pick someone to defend. Making in the top 8 twice was a pleasant but slightly unexpected bonus. Getting picked by 125 and playing with 233 at Orlando was a blast!

Ultimately this strategy, or our performance, wasn’t strong enough to make the elimination rounds on Newton. We added some defense to our Newton game plan, but I agree that our scoring cap, low autonomous output, and all-or-nothing scoring was the key limiting factor. I was very happy to see 190 make eliminations with a similar robot. They played great and had a better autonomous scoring potential than our robot. Well done!

Here’s the real success story to the strategy. We also knew a climber / dumper would differentiate us and bring more visibility to our program in the FRC community (hopefully not the spectacular failing kind.) This success also gave us greater exposure in the school and community. As a result, we’ve had more students apply for the team this season than we can handle.

There’s more to this game than robots.

David

[efoote868]

I think another thing to look at is the timeline of the rules, specifically.

Early in the season, rules pertaining to climbing were different. Given a strict interpretation of the rules before January 15th, climbing the pyramid was incredibly difficult and was likely written off by many teams.

Two weeks into the season, the rules relaxed and rule 23-1 was introduced.

I'd venture to guess that many teams can't toss out a third of their build season to react to a rule change of that magnitude. That is to say if they weren't planning to climb before that point, they likely weren't going to redo their designs to include it.

Edit - That also includes the assumptions they made about climbing. If it wasn't worth a team's resources to climb, their outlook on the game was likely different.

[LeelandS]

As a team that had originally planned for the 50pt climb + dump, here is my take on things.

We decided to go for the 50 points because we were basically going to dedicate the entire 6 weeks to the effort. We were confident that we would be able to develop a robot to make it work in 6 weeks. The 50pts would have been a huge swing, especially at our Week 1 regional at Finger Lakes, especially where we only saw two robots climb for 30 all weekend, one who didn't do so until elimination rounds, and one who stopped half way through qualifications after a fall. Of these two, only one could dump.

In a lot of ways, I was the unraveling of this plan. I have a deep issue with one trick robots, largely along the lines of "Don't put all your eggs in one basket." Especially when the dropping of this basket breaks all the eggs, the basket, and anything beneath it. Oh, and the basket costs $3500. I pushed to develop a shooter (for the autonomous points, and as a fall-back plan). The extra weight and complexity drew away from the resources for climbing, and we ended up scrapping the climber Week 5, and rapidly developing a 10pt climb in the last week, and purely cycling. Would we have successfully developed the climber if we hadn't gone to shooting? I don't know.

Truth be told, I'm glad we did this. We ended up seeding 5th at the event, captaining the #4 alliance, and winning the event. As a cycler, we were mediocre at best, averaging about 2 cycles in a good match. But the 18pt autonomous (in which we were very consistent), the 10pt climb, and the 24 disc points we earned were more than enough, and we were more consistent then I imagined we would have been climbing for 30, then dropping those 4 discs into the goal. Plus, we weren't put into that precarious position that could have ended our season.

The plus side is, improving our ability to score (i.e. better cycling) raised our point ceiling. After swapping gearboxes at Worlds, we were consistently getting 3 cycles, sometimes 4. Were it not for unrelated consistency issues, we would have been much more effective. Meanwhile, dedicated 50pt climb and dumpers were still scoring 50pts. It doesn't matter how quickly or efficiently the climb and dump is accomplished, as long as it is done.

In the long run, I think it comes down to consistency vs points, factoring in a team's resources. 1114 and 67 were the top climb-and-dumpers this season. Not many teams have their resources. We definitely didn't. If a team had the resources to effectively manufacture that kind of climbing device, more power to them. Many teams don't have those resources, but many teams tried. So we saw many teams of varying consistencies go for those points, with many failed attempts.

[Jared]

Our team built a 30 point climber, and it was the most successful robot we've had in a long time. We seeded fourth at our regional, almost exclusively due to our climber, and 13th in Archimedes (we weren't picked though).

Our strategy from the beginning was to shoot three in autonomous, full court shoot about 10 discs, and climb/dump. However, we had some problems. In both our events, our competition shooter performed significantly different from our practice one. At home, we could fire 10 out of 12 discs in the three point goal from full court in 40 seconds, including line up time, but we never made any full court shots at competition, only autonomous ones. We spent a LOT of time on our climber, and our climb time was 30 seconds to the top, which we bumped up to 18 seconds, then we broke our last spare worm wheel, and then went back to 30 seconds. Besides for this issue (we used 3 CIMS on a transmission designed for only 2, which is why it failed), our climber was mechanically perfect, and made it to the top every time it started.

If you're interested, here's a video of us climbing with the slow deployment mechanism.
http://www.youtube.com/watch?v=wL7eT4H2ARM

However, the climber had a fair share of disadvantages. The whole subsystem weighed about 30 pounds, and took a long time to perfect. We must have made 20 different iterations of the hooks before settling on the final one. It was expensive, as it used two five start (on the faster version 8 start) lead screws, which weren't cheap. Also, it was difficult to line up quickly, and our deployment mechanism at our first event was extremely slow. We used our shooter to dump the discs at the top, which was a poor strategy, as the depth of the goal in Archimedes varied by over 1/2", something we didn't think to test at home.

We also fell from the top of the tower, and really destroyed the robot. The frame of the robot above the wheels was destroyed and had multiple broken welds, so we ended up removing our shooter, hopper, and all the motors up there and replacing the structure of the robot with a spare one. We also built a new hopper. The scariest damage was the climber deployment lead screw, which had bent at a 90 degree angle, but through some miracle, we managed to bend it back and get it working again. Then, we found that our power distribution board was dead, so we had to rewire about 20 really inconveniently placed connections. Overall, this was NOT worth it!

With that said, climbing was a lot of fun, and if we had a little more time (and not lost a week to the snow storm!!!) we could have had a really awesome robot. Overall, the biggest problem with most climbers was reliability.

Our robot scored an average of about 60 points per match (with a potential of scoring 68 max with auto points). For fun, we trying cycling in a practice match (no driver practice with cycling!), and got 5 cycles with missing only 3 discs (but with no defense/climb).