I have started this thread here to put into discussion arguably what I think the best climbing strategy of 2013 was. This is also to put into question how teams solved this engineering problem by ingeniously designing their mechanisms to climb. Feel free to post your thoughts of this and concepts of how you think teams climbed the corner of the pyramid successfully.
Look at all three Einstein winning teams. Pure cyclers (and 2 with ground pickups). 30 point climbing is nice, but it was not the most reliable (see 67, 1114, 254) strategy this year.
Correct, but then again the 1114 118 alliance in Galileo had the highest score out of any alliance at champs. If 1114 would have had the time to fix their climber before Galileo finals so it didn’t stick, they most likely would have won champs. 1114 and 118 were cyclers as well they just had problems keeping them from moving onto Einstein.
It depends how you define “best.”
Most reliable? Hardest to defend? Highest scoring? Greatest crowd pleaser?
And are we talking about corner climbing as the best “overall” strategy or in comparison to other climbing methods?
I wouldn’t say any team was most likely to win champs. If you remember 469/33 was consider by most a shoe-in to win it all before the finals started.
On topic now…
For climbing, I do not believe there was a “best” strategy. It was more about the room you had on your robot. 254/1986 robots were very fast at climbing along the faces. 67/1114 were fast at climbing the corners.
The one difference I see is that 254/1986 did not give up the ability to pick up off the floor with the face climbing, while 1114/67 did with corner climbing.
I think it is the best because it is the most risky way to climb. If you slip off a rung you probably will flip over and on to the ground. It is also the least stable because you only can hold on to one rung at a time. so the body of you robot can twist.
Agreed. Our team’s climber also went wonky in the quarterfinals on Archimedes and couldn’t climb higher than the second level. Our autonomous climb ran smoothly all season and failed at the worst time. Due to the complexity of climbing this year I’m not surprised by the wear and tear among the 30 point climb club. Looking back at the season, we would have still stuck with our climber it provides unique flexibility in a match.
I also think the reason 30 point climbers were so rare this season especially in the elimination rounds came from too many rules, low point value for the level 3 climb*, and Robot In 3 Days.
The reason I bring up RI3D is because I feel as though a lot of teams who would normally have focused on one aspect of the game, specifically the endgame, were inspired by the RI3D robot and strategy. Looking back at the season there were tons of teams who used a similar shooter and/or feed system. Looking at the RI3D robot, the real limitation in the point score was how quickly you could travel to the feeder station which for many teams was 3-4 trips, plus autonomous, plus a quick 10 point hang. This greatly outweighed a 30 point climb which for many teams it was a choice between one and the other.
That being said I love RI3D! Those guys worked hard and they really brought such a unique aspect to the season and served as an inspiration to many teams. I really look forward to seeing what the 2014 season brings for them.
35 would have been a better score for level 3 IMHO
Back on topic.
To be honest, I really loved the strategy 469 used this season. They have a very reliable autonomous with both center line and under the pyramid, quick cycler, quick floor pickup, extremely accurate full court (from both sides of the field) shooter, and a fast 10 point climb. Such a versatile machine that can cause a lot of damage in a match from many angles!
Similarly, 254 & 1986 (and us) gave up dumping the colored discs while 1114 & 67 did.
This year was really about trade offs and was a strategists dream with so many different aspects of the game to play. Its really hard to say what the best strategy was.
I would say this is how not to climb the corner. Actually we drove up the corner. For most of the season we were a cycler and our caterpillar drive did not work. It was only on Saturday at worlds that our system really started to work. It was an intense obsessive effort for the entire season. This is a video of it in action. It was taken by a Gopro camera mounted under the robot.
http://www.youtube.com/watch?feature=player_embedded&v=O4GK4HfHiBM
Notice that the 2nd leg does not release when it should. The season is over but we are still trying to dial it in.
This is a 3D model of the drive
http://wiki.team1640.com/index.php?title=File:Caterpillar2.jpg
We poured an unbelievable amount of resources into it but it works - kind of.
This is a link to our website detailing the project.
http://wiki.team1640.com/index.php?title=DEWBOT_IX_Caterpillar_Drive
Note the state machine diagram at the bottom of the page. It stressed our programmers too.
I don’t consider it economic success. Given the resources pour into vs. the contribution to points in the competitions. It’s a looser. How ever if we look at the skill sets and knowledge that our students gain from this design build project it’s huge. Really huge. It forced our students to really step it up. I’m proud of their accomplishment.
After surviving the caterpillar drive project, I can’t help feeling the GDC is evil. They created the pyramid climb thing knowing full well that teams like us could not resist the challenge and would jump at it. They knew the pain suffering climbing the pyramid would bring to teams. Well done GDC, well done, You got us good.
It might be hard to define best, but I think that given the constraints, I think the jury is in: 30 pt. climbers weren’t worth it. When 1114 can’t make it work reliably it’s fair to say that most teams would be better off with a different strategy.
If you had unlimited time and resources to build the ultimate robot to play ultimate ascent, I think you would end up with a corner climber. Specifically, you would get one that did a passive, after the buzzer 30-pt climb.
I felt floor-up was highly more valuable then the colored disks… There were also face climbers who could dump.
I looked at it more of method of climbing, than to climb or not to climb…
I could almost write a book on this subject. We had a fairly fast (under 30 seconds including alignment) corner climber with the ability to dump colored discs. It wasn’t the best climber out there, but we got it to work fairly well. It took over our lives most of the season and it dominated our strategy and fate. Was corner climbing the best strategy of 2013? I would say no. Was it the best strategy for us? I still don’t know, but I am glad we did it. It was what I call a “white knuckle ride” all season.
Our robot did a lot of things OK, but wasn’t outstanding at any of them. We had a floor pickup arm that let us run a 5 shot autonomous, but generally it was faster for us to run cycles than collect from the floor. We could run 4 cycles in a match if we didn’t plan on a climb & dump. In addition to the corner climber, we could do a “quick 10” hang from the low bar.
We only won once this year. That was Week 1 at Traverse City when we only had a 3-shot auton, never did more than 3 cycles, and never climbed - just hung for 10. Our OPR was around 40 - our lowest event all season, but we still won. Climbing was not a factor. General lesson: If you show up on Week 1 with a functional robot and a drive team that has had some practice, you can do well.
Our second event was West Michigan on Week 4. By then, our climber was ready to try in “battle”. We had also improved on some other “lessons learned” from Week 1. The climber set back our overall performance. Since we hadn’t used it in a match before, we would quit cycling and allow ourselves at least a full minute at the pyramid for alignment and climbing. If we wanted to carry discs, we needed to quit cycling even earlier in order to get them. Our best match was 5 shots in auton, 2 cycles, and a climb & dump (104 points). That was the exception, and we spent a lot of time & energy trying to repeat it. What usually happened was that our opponents would see us collecting colored discs, then swarm all over us until we didn’t have enough time to climb. We managed to make it to the finals, but it wasn’t because of our climber.
Our third event was MSC. By then we were getting desparate to make the climb & dump a strategic and visible success. We got off to a lousy start, going 1-3 in our first four matches, despite improvements that cut our alignment time and sped up our climb. We were almost ready to go straight from autonomous to loading colored discs and climbing in order to demonstrate a consistently successful climb & dump, when we realized that the dump had become our worst enemy. It was an all-or-nothing option, and in order to achieve the “all” we were giving up more than we stood to gain. We quit trying to dump, and didn’t even worry so much about reaching the 30 point climb. We went back to trying four cycles and then climbed as high as we could in the time we had left. Most of the defense went away. We were productive for the whole match. Our max scoring potential for a match was 108 (30 auton, 48 teleop, 30 climb). We never quite achieved that, but our average scoring was in the mid 80’s for most of Friday and Saturday. We won our last 9 qual matches. We had a good run through eliminations, but lost the semi’s in 3 matches. Partly because of a shooter motor failure, but mostly because we just got beat by the superior alliance of 2054, 67, and 2337. By that time, the climber was showing signs of wear and tear.
At Championship, we “struggled” (stunk) in quals due to shooter and drive train issues. The climber was working, and was one of the reasons we got picked by 4814 & 67 for elims. Our job was to shoot 5 in auton, create a shooting lane for 67’s FCS, then climb. The strategy worked well enough to get us three matches into the Curie finals.
In our case, the climber was the difference between a decent season as an average cycler in a crowd of better cyclers, and a great season (in our opinion) where we had the ability to do great things when it all worked. It was stressful, but exciting. We have lots of vivid memories, but they aren’t all about the high points. The design of our climber had a lot more mentor content than any other system on the robot (or any other robot we have built). I suspect that it true of many teams that climbed this year, and I am not sure that’s a great thing. Using it put a lot of stress on the drivers - especially the one that operated the climber. There has never been a game where a simple miscue could so easily result in a season-ending catastrophy. We wanted to have a student field coach this year, but I held that role for myself until we we were having good climbing success. If our season was going to end in a crash, I wanted it to happen on MY watch, after I told the drivers to go for it, when I gave them the go-ahead to proceed from one level to the next. I wasn’t confident early on, and I wasn’t going to put that responsibility on the students. We never did have a student field coach. That wasn’t such a great thing, either.
From what I observed, it sounds like almost every climbing team experienced failures at critical moments. In my opinion, the points for climbing to level 2 & 3 weren’t proportional to the work and risk it took to accomplish it. I don’t think a single robot on Einstein got there mainly because of their ability to climb. By that standard, climbing (corner or wherever) was definitely not the best strategy of 2013.
PS: We took two famous falls this year. Fortunately, neither ended our season.
(1) MSC Q101 where we won 247-70. We were paired with 33 and 67, and could have set the season-high score had we completed our climb and dump instead of falling from the 20 point level. We only contributed 30 of the 247.
(2) Curie Final match #1 where we won 186-0. We were at the 30 point level with a bad grip. When 67 reached the top, they hit the crown of the pyramid hard enough to knock us loose. The referees blamed 862 for knocking us off, and their alliance got red-carded.
In both cases, we were able to climb in our next match, but our poor robot is suffering.
It may not have been an economic design success, but in addition to what your students learned from it, with the excellent documentation you’ve collectively produced, I’ll be using this as an example of what’s possible to inspire my team for next season.
Awesome use of the GoPro! With that positioning, you had a way to see what had gone wrong on climbs so you knew what to fix on a part you’d never be able to see properly from driving position or the stands.
In my opinion, strategically the best climber is the fastest climber. A climber that takes half a match isn’t worth it.
A climber that takes half a match but also dumps is pushing it.
A fast climber is worth it.
A fast climber that dumps is gold. (<10-15 seconds maybe?)
See 1114 (pre-cmp elims) & 254.
Both valuable and strategic climbers, but different methods.
I’m going to throw a shout out to team 1421’s climbing concept. Their scissor lift inside climb with the dump in around 25-30 seconds was arguably the most reliable climb at champs. The best part about their climb was that they could extend their lift during the match to block quite a few fcs. I believe a corner climber had an advantage of being able to cycle better due to it usually being able to climb at a faster rate than the side climbers. My strongest point on corner climbing being the best way to climb for 30 is the teams that decided on corner climbing, 67, 71, 217, and 1114 + many other great teams. If those four elite level inspirational teams all decided to climb the corner over the side, then how can I argue that it wasn’t the premier climbing option. Yes you can argue that three of these teams didn’t make it to Einstein due to mechanical problems from their climbers( 217, 1114 both had problems and 67’s partner 1918 had a few issues from their memorable "pyramid dive"sorry for bringing it back up) but as far as strategy goes I think it’s a solid choice. It also says something that the only 30 pt. climb on Einstein was by team 1640, who had a corner climb.
217 never climbed past 10 (in a match), so I don’t know why they keep being mentioned as a elite 30 point climber.
They had the mechanism, but were never able to get it working. I haven’t even seen them pull off a 10 point hang.
I was just point out the strategy not the execution per say. I believe they intended to climb for 30 but they had a rough season this year as far as it goes mechanically. I was just analyzing the Thunderchickens choice. Instead of looking at them as an elite 30 point climber. I was looking at them as an elite team who attempted to 30 pt. climb from the corner. They just were unable to pull it off (which is extremely rare due to 217 being such a great team). Sorry for the confusion.:o
Although it may not have propelled us to the Einstein field, we, team 1448, were content with our 34 second corner climber that dumped four colored discs. I don’t think there was another robot at championships that used our rack and pinion system, which in my opinion was extremly reliable, as our robot never failed a climb given the time.
I wouldn’t say that there was any one specific “best” strategy for this season, for either overall game play and/or climbing.
Outside of the Einstein winning strategy, there were a number of very successful strategies that could have been just as successful given slightly different circumstances. Cycling, FCS, Floor pickup, Corner climbing, Side climbing, etc… were all in play leading into and up to Einstein.
Specifically to corner climbing for us…we decided to persue this type of climbing to allow for room for an additional climber. It was also the first method of climbing that we were able to meet (barely) the original 54" cylinder rule. We not a fan of swinging while climbing, and didn’t really have time/resources to proto-type any other methods of climbing. We didn’t even prototype our corner climber method…we were all or nothing on getting it to work or fail trying.
By going after this method of climbing and the design we chose, it significantly impacted our ability to floor load. We tried to pursue a flip down floor loader for a while, but were unable to meet all the packaging requirements to do both.
Climbing the pyramid for 30pts was our #1 design priority. We significantly under-estimated the # of discs that could be scored, either through FCS or cycling. The ease at which good teams scored discs really de-valued the importance of climbing. Combine that with the importance of autonomous as a tie-breaker, and I would say that our design priorities were really out of whack this year.
But, even with that being said there was a place for climbing in this game. 254 and 1986 showed it was possible to do while still pulling off multi-disc autos and 1114 showed it was possible to overcome the auto disadvantage by cycling 3/4 times and climbing+dumping quickly.
IMO - Being able to score discs + Climbing and/or dumping = Inspiration.
In the end, we may have had more success going a different route…but, I think getting the climber working and seeing it corner climbing+dumping at the end of the division finals was one of the coolest ways to lose an FRC event.