Reality of low-bar shooters
 

(Basing any conclusions on data from this post)

Throughout build season, there were many predictions that low-bar robots which cycled quickly to score balls, followed by a capture, would be a common and high-scoring design. However, it looks like the highest success rate was 9% during a given competition.

Does this mean a lot of low-bar bots failed to achieve what they attempted? And if so, do you think this will be changed as the season progresses?

Or, was CD wrong, with a large portion of First teams going for a breaching bot instead of the low-bar bot?

(Note: It could just be that this small sampling didn't have many of those bots, and the percentages will vary greatly with each weekend)

Re: Reality of low-bar shooters
 

We've only seen one week of competition, and when you evaluate ease of scoring vs. point value, scoring a breach consistently is easier to do than high goal scoring, and just as valuable. As the weeks go on, and teams get better at weakening defenses, that strategy has a maximum value.

Essentially low goals and breaches are easier challenges than accurate shooting, with similar outcomes in terms of points, so they'll be common during Week 1. As teams get better, I expect more high goal-shooting alliances to be successful.

Re: Reality of low-bar shooters
 

We plan to add scaling and LG scoring to breaching and see how that goes. Hopefully by day 2 we'll have auto HG aim going too in case of elimination rounds.

In SD it was all about Breaching then Scaling then LG...last items for this week are HG and cross+HG auto

Re: Reality of low-bar shooters
 

I don't think teams necessarily "failed" in what they attempted. I don't think we failed even though we never captured once in our 16 matches. For us to capture we required help. We could not guarantee a capture on our own since we needed a couple goals by our alliance partners, and to have everyone parked on the batter. We could however be pretty sure we could breach since there were more partners that could help us, so that's what we focused on.

We scored 6 high goals in teleop in our last finals match (the short blue bot). Just because the data counts us as 0% for capturing doesn't mean we're not capable. I have a feeling capture percentage will rise as the weeks go on :) By champs it should be pretty common.

Re: Reality of low-bar shooters
 

We need a little help too. Hopefully we can finally team up at 10k :)

Re: Reality of low-bar shooters
 

From my own observation most teams weren't even trying to shoot balls they focused more on breaching which is were the majority of the points have been coming from so far. I know for sure that my team is planning on being a shooter and a low bar effective.

Also from my own observations I have noticed that because of the lack of defense being played there really shouldn't too much trouble if you have an effective shooter.

Re: Reality of low-bar shooters
 

This is largely correct. In addition our team gages what we think we need at any given competition. For our week one we decided the priority was batter shooting and defense crossing. We were very happy with our performance. We learned a ton and are already making major bot revisions. We already know for week five scoring from the defensive positions and climbing will be necessary. That is our focus now.

This is the same every season. Trying to draw conclusions from performance after a week one event is a waste of time. Going into week one there are many variables that are totally unknown including field response and reffing.

Re: Reality of low-bar shooters
 

Breaching didnt occur as a preference vs. high goal scoring, in terms of points only.
Breaching was a priority because of the extra ranking point.
Ranking points is the name of the game in qualifications..

Re: Reality of low-bar shooters
 

This was week one. As always the game will be very different by week 6.

Re: Reality of low-bar shooters
 

As a comparison, think of how many 6-stacks you saw in week 1 last year.

Re: Reality of low-bar shooters
 

^ This... its much easier to earn a BREACHING RP than a CAPTURE RP since capture needs three bots and more time most likely. One bot can do a BREACH alliances need three bots to do a CAPTURE. Good luck with that until eliminations in most competitions.

Re: Reality of low-bar shooters
 

It is obvious that breaching is easier than capturing. Because of this, at low levels of play teams choose to forget about boulder and capturing points and focus on breaching because it is a sure ranking point. During our 17 matches at San Diego we captured 4 times (3 times during elims) and breached 16. At the San Diego event there was a total of 6 captures. By week 3 I think most regional's will have three times that many.
I don't think teams failed. I think they focused on getting easy points first and harder points second. This is why we will see a tremendous increase in captures as teams have more time to figure out vision tracking and dial in their shooters.

Re: Reality of low-bar shooters
 

I would like to make a comment in regards to low goal shooting.
It is much harder than you think it is. When looking down the field it is almost impossible to see the dividers on the batter.

Practice fields, if they have them at all, have wooden ones which can be seen pretty easily. The polycarb ones are almost invisible from the driver station. You know they are there but it is all too easy to run into one.

There are also reflections off of the diamond plate which can make things confusing too when you are trying to get close to it to line up.

We did find that the diamond plate reflection can be used when you are behind a tall defense to help you lineup if you can see it.

I am sure that with practice on a real field, teams will get better over time but just a word to the wise about translating practice field vision to the competition field vision. It is substantially different.

Re: Reality of low-bar shooters
 

However, with the low goal though as long as you have a camera it doesn't matter if you have defenses in front of you, you should still be able to see the low goal.

Re: Reality of low-bar shooters
 

Yup..biggest issue we had with our driving camera was lag... in fact that is one way we broke our shooter was our driver was relying on the camera to drive and slammed into that divider multiple times. Thinking he was heading to LG.

Afterwards he learned to use it to line up goal shots and wait for it to settle down frame wise. I asked the team to order a faster framerate camera for CV we'll need it up there.

Re: Reality of low-bar shooters
 

This question doesn't make much sense. Most breaching bots* will be low bar robots. It is far easier to accept the constraint of a low bar passage than to design manipulators to handle the sally port and draw bridge - even in the plywood versions, much less the reality of polycarbonate.

3946's primary strategic point was being a sapper/breaching robot, so we could get that RP by ourselves even when we draw the short straw on alliance selection. The ability to intake and score high goals was second, because we realized that an alliance can only score on ten defense crossings, and boulders are the only "unlimited" source of offensive points in the game. We expect to have an autonomous high goal in the next week. Due to the vision obstruction of the defenses and the small goals, I expect most successful high-goal shots to either involve computer targeting or to be taken from the batter.

* by this, I mean a robot designed to effect a solo breach without assistance (or at least specific assistance) from its alliance partners to ensure at least 1 RP per match.

Re: Reality of low-bar shooters
 

AGREED. 359 sums it up the best from a strategy pov
Thanks 2502! We'd love to finally play some matches at your side

Re: Reality of low-bar shooters
 

Low goals are still extremely valuable. Most captures I have seen have had at least half of the boulders scored in the low goal. Seeing almost everyone focus on breaching during qualifications is just a matter of it being impossible to get the additional ranking point for a capture without all 3 robots being on mission and reasonably functional.

There are lots of low bar/low goal robots. A particularly good one in my opinion is 5624. They are fast and nimble being able to get under the low bar quickly and navigate around defenders. At the Mount Olive District playoffs, they would score 6 or more boulders in the low goal almost every match. Just because capturing was rare does not mean that being able to capture was rare. Most teams just focused on breaching since it was the most reliable way to score the ranking point.

One of the central strategic components to this game is the interaction between boulder points and defense points. Defense points are easier to get, and you have to get them (and breaches) if you want to do well in quals. But defense points are capped, so to go on, you need boulder points. The balance between easy defense points and uncapped boulder points is the key to understanding Stronghold

Re: Reality of low-bar shooters
 

Our shooter sucked... (Balls continually stuck, and burnt 4 775pros)
Our NavX failed to tell us telemetry... All Autonomous turns found bot doing donuts.
Our climber sucked...
Our vision sucked....
Never made a high goal...

Still ranked 2nd at Waterford mainly due to understanding the game, and plying to win, along with posting the highest auto

Think we are sitting around doing nothing this week. . .
Think Hot is. . . (they didn't move in auto)
Think Bees are. . . (they didn't move in 3 matches)

Our performance at Waterford, will probably not even be picked for eliminations by week 4, in Marysville.

The only conclusion you should draw from week 1, is that it's over...

Re: Reality of low-bar shooters
 

We built a low bar capable robot with an extremely robust drivetrain and a shooter designed for the low goal. We competed in Week 1 at GTR Central.

Our shooter did not perform to our satisfaction but we discovered we were very good at doing breaching, we were capable of breaching most defense configurations on our own.

Our best matches were with two alliance members both capable of doing low/high goal shots - 4 boulders each for the capture and we took care of the breaching.

Some of the best robots at the event were low bar, low goal capable robots who had worked to optimize their cycle time. We didn't see a lot of high goal shooters because they took so long to line up their shot, and they were vulnerable to defense.

Of course, that was week 1...

Re: Reality of low-bar shooters
 

It sounds like you are looking for an explanation for why capturing the tower was so rare. Several reasons:
- You need to walk before you can run. Even if you have designed a robot to be a fast defense crosser and a low goal scorer, until you play the game, you don't know how quickly you can do low goal cycles. The smart thing to do is start by establishing your breaching skills first, then add in scoring on the tower.
- Despite best laid plans, probably half the robots at a given event have trouble executing a drivetrain capable of quickly traversing most of the defenses. That's just the reality of FRC. When we do our kickoff weekend planning, we consider three classes of robots. Class A bots are alliance captains. Class B bots are pretty good. And the rest are class C bots, which struggle to accomplish the basic tasks of the game.
- This year, it takes at least 2 class B bots to breach. That combination, or better isn't that rare. To capture, you need at least a class A, a class B, and a class C bot, plus a smart strategy which is well executed. Statistically, this combination is hard to come by in the qualification rounds.
- At their first event, most teams are learning to walk. Out of our 19 matches in our first event, we failed to breach only in our first match and in one other, where a robot broke down in front of the last defense, with one crossing left. We also captured 3 times in the elimination matches. We now know when it's appropriate to make an attempt to capture, and what is needed from our alliance partners to do so. We're looking forward to playing our next event, where there will be other teams who have also learned these lessons. I expect there will be many more captures at that event than at the first.

I should also mention that at our first event we were essentially a low bar capable box on wheels with the ability to carry balls and score them in the low goal. Despite this we were the #1 alliance captain, and won the event. This was due to a great drivetrain, hours of driver practice using our practice bot, a terrific scouting team, a terrific pit crew, and well thought out match strategy.

Over focusing on the abilities of your robot starts you down a path to under-performance. I continually preach the mantra "It's far better to have a good robot quickly, rather than the perfect robot right before bagging." When we opened our bag, we were able to proceed directly to inspection and right into practice matches, before more than half the robots at our event had even passed inspection. That alone is a huge advantage.