Climbing and dumping not winning?

[JohnFogarty]

I saw 1746 this past weekend that shot 3 in autonomous and then climbed for 30.

Back to the purpose of the thread....the elimination average score is no where near 70. The average score is more like 90-100+.

A team like 4451 is probably going to find a lot more success at world where they can be paired with two very accurate shooters and possibly ones with floor load for autonomous like teams 11, 1114, 2056, 624, 118, 148, 2415..etc.
If they added a shooter solely for auto like 1746 did then 4451 would be a team I wouldn't want to play against.

[Alpha Beta]

Originally Posted by Lil' Lavery

You know what confuses me about this thread? The timing. Didn't we just have our first climb+dump winner during week 3? 1806 was the #1 selection at winner at GKC (ahead of a team with a 7-disc autonomous).

It's already been stated by others, but let me add some perspective since I work with the scouts to create the final list.

1. We did take all climbers who didn't also have a shooting mechanism and solid drive base off of our pick-list. We needed the flexibility to move our 3rd robot into a defensive roll, and some climbers blocked the sweet shooting spot in the center of the pyramid for a significant portion of their climb.

2. 1806 was the only consistent full court shooter in the field, and when left unmolested could pour an incredible number of discs into the high goal. Because they could hit a wing auto to compliment our 7-disc center auto, and could climb and dump for 40 as a back-up if blocked, they had enough flexibility to be very attractive. There were safer picks, but no one with as much explosive potential to score. (They also took the defensive attention off of us.)

3. No one had an extra disc auto to compliment our 7-disc center. Hopefully others will develop a wing to wing 5 disc which includes the 2 (at minimum) center line discs. We are working on that as well so we can be the most flexible should we pair with other 7-disc specialists at champs. (Side-note: Any team who does not have floor pick-up and extra disk auto capability should spend more time in the tournament perfecting a wing autonomous. Getting 3 from the center is not going to cut it IMO.)

Originally Posted by Lil' Lavery

There was a seven disc autonomous available, even if both can't be used simultaneously. Denying the opposition the chance to score those additional four discs is every bit as good as scoring them yourselves.

That is a good point, and one the scouts discussed. The other 7 disc auto wasn't very consistent yet, and deemed as a lesser threat. Not picking them was a decision we knew we might regret later.

[Jaxom]

Originally Posted by Lil' Lavery

There was a seven disc autonomous available, even if both can't be used simultaneously. Denying the opposition the chance to score those additional four discs is every bit as good as scoring them yourselves.

True, and we were aware of this. I guess I didn't say it very well....what I'm trying to point out is that GKC isn't an example of climb-and-dump winning a tournament. It might have been an example of a full-court shooter winning a tournament, but the GKC teams were smart enough not to let us get away with that.

[sodizzle]

It looks to me like a perfect 3 disk autonomous in the top goal plus a climb and dump would be a valuable asset to an alliance. Especially if they can play defense first. So a team that scores 68 points (18 + 30 +20) that can play defense, so potentially blocking a cycle of top goal shots (blocking 12 points), could be worth it. Or a climber and dumper that can pick up a cycle of disks and then climb and dump could score 80 points (18 + 12 + 30 + 20). Thats when I believe it would be effective enough to win. But consistency is key. You may be able to score 80 points but if you cant climb and dump in a match you would only score 30 points. And that is out scorable by the majority of teams.

[theun4gven]

Originally Posted by Koko Ed

What shocks me is no one played D on them. They just let them go get their colored Frisbees and go to the pyramid and do their climb and no one bothered to even try to get in their way.

Swampdude already touched on it a bit, but anyway: We played against 125 and 4451 in our final qualifying match. One of our partners (4562) had no frisbee manipulation capabilities and was built solely for defense. Our other partner could only score 1 point goals, and slowly. Both of these robots did their absolute best playing D against 125 and 4451 with us doing our best to score. 4562 played great D for nearly the entire match but eventually 4451 got back to their pyramid. End result: 91 to 61 loss. D against 4451 doesn't matter unless it is effective nearly the entire match. Once they hit the pyramid they have their points.

With this knowledge our finals alliance decided very quickly to ignore 4451 entirely. Playing D against a good shooting bot slows them down and lowers their scoring. Playing D against 4451 takes perfect D and removes one of your bots from scoring. You mess up once and they get their points. For a great example of this watch Pink during the finals. Spectacular D every match but we still got through once or twice to put up a few points. You let 4451 through once and you've effectively taken yourself out of the game and didn't hurt them at all.

[EricH]

Originally Posted by Zebra_Fact_Man

How many more teams do you think would have attempted the 30pt climb (and/or perfected it) if the "climb in succession" rule had not existed?
[In other words, how many more teams would have climbed if they would have been allowed to grab on to something higher and pulled up.]

I personally feel like this is the single rule that makes the whole climb process so hard to complete.

Everybody and their little brother would have probably thought seriously about it. I'd put it at half the teams would attempt it; of those, half would actually succeed at the 30-point level. The rest, and a fair amount of other teams, would end up at 20. The 10-point hang would probably be almost worthless.

Think about it: 2000, 2004, and 2010 have all had hanging, with no contact with the support structure. Teams have those years to reference. But when you include the climbing in sequence, all three of those years are now out the window--you can still get ideas, but they're a lot harder to pull off (particularly with the "detach and winch" variety of hangers from 2004).

[Racer26]

IMO, its really all in the consistency, and 4451 showed that. No, they didn't win any events, but they performed quite respectably, and if the cookie crumbled only slightly differently, they might have won.

If you can't be consistent about it, its not worth dedicating your design to it.

I haven't seen ANY other consistently strong Climb+Dumpers. A Climb+Dumper with no auto, and 100% consistency should have an OPR of exactly 50. That would be a top contender at nearly every regional.

A fairly trivial 4 to 6 point auto could be added with minimal effort, and with 100% consistency, that would bring OPR in line with the likes of 610. Not as strong as 1114, 2056, 987, or 1986, but certainly strong enough to win weaker regionals, or be a top pick at a regional with only 1 powerhouse.

[thefro526]

Going back a few weeks to the end of build, I was one of those people that believed climbing and dumping was a viable and advantageous strategy. I still do. Machines that climb and dump for 50pts may not be winning events, but they still have their place in this game and can still be highly competitive.

If one were to take a look at the match scores from the last three weeks, they would see that the Average Qualification Match Score is ~50pts and the Average Elimination Match Score is ~80pts. This would imply that 3 robots, on average should be capable of scoring 50pts or so in a match and not necessarily be of Eliminations Caliber...

So, if a team were to field a 50pt climb and dump robot, that team would - in theory - out score the average qualification match with a disc or two scored in autonomous, or a partner that puts up a point or three, and that same climb and dump robot, if paired with another reasonably capable machine, should be able to out score the Elimination Match Average without too much Hassle.

I think the reason that we're not seeing a lot of climb and dump robots winning events is more complicated than it seems, but the primary reason is that there just aren't that many robots that went for the climb and dump strategy, and many of those that did, aren't climbing and scoring reliably enough to really make a dent.

All of that being said, I think that data shows that building a climb and dump machine wasn't a bad decision, especially if it's effective. Regardless of the time the machine takes to do it's climb and dump, 50pts are 50pts and there are a heck of a lot of machines that couldn't touch that number on an open field.

[eedoga]

So I may be a bit biased on this one. My kids decided to build a climbing dumper. I'm still not sure they are going to pull it off because this was one heck of a challange, but with luck and hard work I think they will make it.

The thing is...I wonder if a climber/dumper takes away from the scoring capability of the other two bots on its aliance. So for example if you have a full court shooter, and a floor loading cleanup bot would having a 50 point climber/dumper be an acceptable trade off to a third shooter/defence bot? So would it be better to have gone with a defence bot than a climber/shooter? In which case it depends on what you are up against and what you are alied with.

Unfortunately I don't think a 50 point robot is a guarantee of anything. In fact it may prove to be too much of a liability to be desirable.

One thing that might change this is the fact that the scoring system might change at the finals. Not sure which way that would go, but if relatively few climbers make it to the finals then maybe it would break in favor of climber/dumpers...

Edoga

[Chris is me]

Originally Posted by eedoga

The thing is...I wonder if a climber/dumper takes away from the scoring capability of the other two bots on its aliance. So for example if you have a full court shooter, and a floor loading cleanup bot would having a 50 point climber/dumper be an acceptable trade off to a third shooter/defence bot? So would it be better to have gone with a defence bot than a climber/shooter? In which case it depends on what you are up against and what you are alied with.

If it takes the whole match to climb / can't drive, and especially if it only makes it to 30 points, it's not worth it. The faster you get the climb, and the more defense you can play before climbing, the better.

There's a reason 3467 was the first pure climber selected at BAE, despite only going to 30. They played great defense for the first minute of the match, then climbed to 30 pretty consistently starting Friday afternoon. In my mind that is a lot more valuable than a 50 point dumper that does nothing else.

I think the best case scenario for teams that built 50 point dumpers would be to go for that strategy in most of their qual matches, but be prepared to switch to defense + quick hang in elims if they have to, or if they don't have the 20 point portion of their game consistent enough.

[apples000]

To me, it seems like robots who just climb and dump aren't enough to win. They also need a 3 disc auto and they need to be able to do some pyramid runs or full court shots. From what I can see, a robot that shot 3 in auto, drove to the feeder station, made some full court shots for 1 minute (if they were defended, they could do pyramid runs), then climbed/dumped for 1 minute would dominate any competition that I've seen so far. Also, it doesn't need a floor pickup. The problem is that a maneuverable, strong, good shooting, and good climbing robot that can win pushing matches is an extremely difficult thing to build.

[MisterG]

2481 Roboteers came really close this weekend at WI. They were allied with 2826 Wave from the #4 spot and made it to the finals.

They had a good auto, were very mobile in the early part of teleop (could defend or do at least one cycle) and could climb in around 30 seconds.

I didn't watch the finals closely so I am not sure what was the eventual downfall.

I think that the climb/dump or FCS roads are not totally un-doable but I think primary shooting is a safer bet.

In the words of Karthik "FIRST games are a trick question." Even knowing that I have not learned which aspects are the trick.

-alg

[Xavbro]

The Bayou Regional this past weekend as a great example of the main point of this thread. 2992 and 1421 were the best 50 point climb and dumpers at the regional. Both made elims with 1421 coming one match away from winning it all in the finals. What made them so successful wasn't the fact that they could score 50 but it was because they forced their opponents to focus on them to try to stop them, thus causing them to slow their scoring down.

2992 was picked by the #5 seed captain and they took the #4 alliance to 3 matches. They forced them to focus on their ability to get 50 points which took away from #4 being able to just focus on scoring. They ended up losing due to 233 awesome defense against them in the 3rd match.

1421 was picked by the #2 captain and they undefeated to the finals due to 155's shooting and 1421 being able to climb and dump 50 with little resistance. In the finals, 1421 had some resistance but was able to still climb and dump and beat the powerhouse scoring #1 seed of 118 and 3753. In the second match, #1 launched an all our defense assault on 1421 slowing them down from their process. With 30 secs left, they were able to complete, in my opinion, the most impressive climb and dump this season (still looking for video of that match) and lost only by 2 points. In the third and final match, #1 decided to let 1421 go through their process, didn't play any defense on them at all and 118 and 3753 just decided to out shoot the climb and dump. 1421 got the climb but missed the dumb but it didn't matter because 118 and 3753 put up at least 127 disc points.

So after watching the elims, it proved that climbing and dumping might get you to the finals, but if the opposing alliance is a very good shooting alliance, you're going to need to scout and have a really good strategy in order to win. 1421 was an awesome bot and had a great climb and dump mechanism but they needed more scoring either from 155 or with 155 and/or someone to slow 118 down to have a better chance of winning.

[RyanShoff]

2481 Roboteers lost because they fell over very early in their last match. Had that alliance got the 50 points from that robot, they would have won. It was their first regional and I'm sure they won't be that tippy at Crossroads.

4143 is a climb and dump with a 4 or 6 point auto dump. We were 8-2 at Wisconsin quals and lost one match by 6 points. 6 more points and a climber/dumper is #1 seed.

Climb and dump can be a winning strategy. Most of the good climbers at Wisconsin where outside corner climbers. We make one trip down the field, get the colored discs. Then we can line up to the corner in the opposing alliances protected loading loading. They bump us while we are lined up touching the pyramid and its a technical foul.

I think the best alliances will be 2 fast shooters and an outside corner climber/dumper.

[RyanShoff]

Also I don't buy the whole 2 on 3 thing.

4143 often a had a bot playing man-to-man on us trying to stop us from getting the colored discs.

One match we pushed it all the way to the feeder station and back to the pyramid.