The Long Shot

I think based on the amount of teams that want to make that shot there will be a considerable number of teams that will actually make it, our team has tested a prototype shooter that can make a consistent 4/5 to 5/5 outer goal shot from 35 ft away. Not only that but about 30% of them go in the inner goal.

6 Likes

In 2017, there were many fuel balls that could’ve gone in that didn’t because another teams’ shot deflected one another. I suppose this may happen again.

For a ball to bounce out of the goal, you’d have to fire it pretty hard, and have it hit the back wall. Of course, if the ball doesn’t make it in the goal in the first place, that’s a missed shot.

If you read the whole comment, which i will repost again, I said “An OK driver can shut down a robot entirely, just by blocking trench, the only safe long shot zone, or initiation line, stopping the flow of balls”. You can see that i mentioned the initiation line, covering the middle. A bot can move back and forth between the trench and line to block them out. Try to read the whole comment next time, please.

I read the whole comment and disagreed with your analysis. Sorry if you can’t take that.

Quoting a specific part of my comment and then “rebutting” against it with an argument i covered in the sentence right after it is not evident of reading the whole comment. I can take disagreement but when someone is trying to come off serious when obviously not reading what I said, I get disappointed.

1 Like

We’ll see who’s right in a few short weeks, won’t we? :smiley:

Point of clarification - when you refer to the ‘long shot,’ are you referring to an inner port shot or an outer port shot?

I assume you’re not referring to a low port shot, although I’d agree that very few teams would pull off a 1-pointer from that location.

Greater than 80% accuracy is plausble if a team has some experience with tuning a flywheel, its azimuth, and tracking a goal. Even 20%+ in the inner goal is plausible.

What does it take? I really have no idea. We just discovered a few things today for ourselves and don’t really know what to make if it quite yet so we’ll test some more. It definitely is >80% accuracy though. Flywheels are a bit of black magic :confused:.

2 Likes

I’m referring to the outer port shot.

1 Like

We’re going all the way for this, mostly because of the cycle time and defense advantage. We’ve tested vision and our wheels (two vertical flywheels) and it certiainly seems we can make the outer goal (inner goal practically impossible) with high accuracy.

Where do your misses go? Do they bounce back across the field, toward where you got them?

1 Like

The long shot is only a slightly different case of the ramp bot paradox from 2018 and 2019.

The Ramp Bot Paradox
If a team is skilled enough to build an effective ramp to let robots climb on them, they were strong enough to build something better than a ramp bot. Thus you get very few good ramp bots.

The Math
Consider how much time is really gained from a long shot over a short shot.
Let X be your cycle time. Let R be the amount of time you save from taking a long shot instead of a short shot. Let S and L be the percent chance you make a short or long shot, respectively. Therefore the number of cycles per strategy is 120/X and 120/(X - R) for short and long shots, respectively. Now, let:

2 * 5 * S * 120/X = 2 * 5 * L * 120/(X - R)

This equation is true if there is gain in choosing either long or short. We can simply it to:

S/X = L */(X - R)
-> X - R = L * X / S
-> R = X - L * X / S
-> R = X( 1 - L / S)

Let’s consider the following cases:
L = 60%, S = 100%

  • If cycle time is 40 seconds: X = 40, R = 16
  • If cycle time is 20 seconds: X = 20, R = 8
  • If cycle time is 10 seconds: X = 10, R = 4

L = 80%, S = 100%

  • If cycle time is 40 seconds: X = 40, R = 8
  • If cycle time is 20 seconds: X = 20, R = 4
  • If cycle time is 10 seconds: X = 10, R = 2

L = 80%, S = 90%

  • If cycle time is 40 seconds: X = 40, R = 4.44
  • If cycle time is 20 seconds: X = 20, R = 2.22
  • If cycle time is 10 seconds: X = 10, R = 1.11

By observing these numbers, I propose that 60%/100% isn’t good enough to avoid moving closer.

However! What about a 10 or 20 second cycle 80%/90%? or a team that can has the same accuracy from afar or close? Wouldn’t it be beneficial to cut cycle time?

The Paradox in Effect
This is were the ramp bot paradox comes in. If you can create a strong enough robot to have very high distance accuracy, why don’t you just make 3 pointers instead? If you can make 2 out of 5 shoots into the inner goal, The equation becomes:

(2 * 3 + 3 * 2) * S * 120/X = 2 * 5 * L * 120/(X - R)
-> R = X( 1 - L / S * (5 / 6)) (Proof left as an exercise for the reader : P )

With an 80%/90% error, we suddenly get that:

  • 20 second cycle goes from 2.22 seconds to 5.81 seconds
  • 10 second cycle goes from 1.11 seconds to 2.59 seconds

With 4 out of 5 balls in the inner goal with 80%/90% error, we get that

  • 20 seconds cycles need to shave off 7.3 seconds
  • 10 second cycles need to shave off 3.65 seconds

Driving 20 less feet wouldn’t save 3.65 seconds. I very much doubt in 10 second cycles are possible by anyone except the top 1% at worlds.

Closing Message
Long shots will still exists. They can be important for strategical reasons, such as preventing your feeder station from overloading or keeping robots from crossing lanes while cycling. I am still skeptical we will see many long ballers on the field.

TLDR:
Few ramp bots were good.
Do the Math, Save the world.
Just take the time to read it darn it.

17 Likes

You really, really, really don’t want to play between my robot and the trench. It’s a very risky strategy and will likely make teams stay away from you when picking. The last things teams want when picking a defensive bit is one that has given away foul points. Most track it on their scouting.

7 Likes

Thanks for taking the time to do these calculations. I am a bit confused about what are the underlying assumptions though. Do these calculations assume a “low bot”? There seems to be more strategic incentives for “high bots” to take the long shot, considering that otherwise they can only go through the shield generator to cycle–meaning a chokepoint that can easily be utilized to shut them down. I feel that speed bumps + tight space + a defense bot with pneumatic wheels could mean total chaos in the generator :fearful:

1 Like

This is not accounting for miss shots and match game flow. The whole reason ramp bots became obsolete was because they didn’t fit into the game development easily compared to traditional climb due to setup. The whole point of the long shot is to allow for a smoother game flow strategy which is different that running just raw calculations.

Thanks for asking! My assumption are based upon the following:

Cycle times:
I choose these number to satisfy a large range of possible cycle times. 40 seconds seems like a fair low number, being only 3 cycles. 10 seconds seems like a high optimistic number, with it being 12 cycles. I wanted to satisfy people on both ends of the spectrum.

Percent of shots:
I choose 60%,100% based on OP’s comments. 80%/100% felt like a more realistic number for percentage of long shots of a strong team. 80%/90% was a “no-one’s perfect” type guess.

Number of inner goal shots:
I choose two numbers. One low and one closer to the 1% of teams.

I purposefully tried to push numbers in favor of the long shot so it was harder to refute my argument once it was completed.


I left the math up so anyone who disagrees with my assumption can do that math themselves. I really considered making 4d models, but I decided that would get confusing. As for defense, it’s such a toss up how much it will play into these factors that I decided not to put in it into the discussion. You can decide for yourself how long defense will slow you down.

I agree! Just pure calculations are not perfect, there is tons of other factors to consider. However, math can give us lots of insights into how a match plays out. I decided to present arguments that had not yet been presented yet. No one should yell, “Don’t take long shots, the math says so!”

1 Like

Hey guys. Here’s my opinion on this topic and what my teams doing. We have a shooter that can absolutely launch the power cells, 50-60+ feet reliably. We are in a position to have it dialed in to hit those trench shots. We also have our limelight being setup on a turret. I have done research and limelight will get an image from that far away, you just need to make the screen nearly black so you only get the target reflecting back. We are planning to have the limelight pick up the target, and the turret moves with it, keeping as close to center on target, so if defense starts ramming us into a wall while we are in the protected zone, we can still score regardless. We are also tuning the shooter to do shots from every other decent position on the field, so that if worst comes to worst and we can’t make it past defense we can outdrive or push them and go score somewhere else. Our goal this year was to think of all the ways defense could stop us, and we are trying to eliminate as many as possible. That’s Just my view on it, thanks for listening m

3 Likes

There is enough room just to spin around bots trying to block the trench. It is a viable strategy but no panacea.

1 Like

More than 0.1% - I’m thinking 2-3 at most events and one in every alliance in elims at Worlds

1 Like

insert 1323 spin move

1 Like