General Shooter Discussion - What? Where? How?
4 posters
Page 1 of 1
General Shooter Discussion - What? Where? How?
Admin Sat Jan 14, 2012 8:23 pm
So.... shooters.
What we really need to answer is: Where should we shoot from? What kind of shooter should we do? Bank shot or swish?
First of all, some physics:
Forget about 45 degree shooting. To hit the basket, we need at LEAST a 60 degree angle, and up to 80 depending on how close we want to shoot from. Varying shooting distances will necessitate a tilting shooter, a "sweet spot" wouldn't necessarily.
Next: How do we ensure consistency? The reason to go with a shooter is because we could potentially score more points than a dumper, but that would require a 67% shooting percentage or better. So: how do we get that accuracy? Like I've said, to get such accuracy requires at the minimum heavy rollers and a turreted shooter. However, running the shooter motor for 2 minutes WILL drain your battery - it happened to us in 06. So we can't run it all the time. So do we direct-drive the shooter, and wait the few seconds each time we want to shoot for the wheels to spin up, or do we make a clutch that can disengage the wheels and let then free spin? One sacrifices scoring speed, the other weight.
Additionally, should we have a sweet spot, or be able to shoot from several places? Obviously, having a sweet spot limits where you can score from, but being able to score from several places necessitates being able to tilt the shooter - which again adds more weight, moving parts and complexity.
Furthermore, each shot needs to be separated by some amount of time to allow the system to regain angular velocity. The question becomes: how long do we need between shots? Essentially, to score shoot 3 times, how long will it take?
Just a few questions to ponder. We're open to any designs at this point, and I'll weigh in as much as possible to do the physics calculations and such.
Since we decided to make a shooter, let's concentrate on making an excellent, consistent shooter. If we're going to win, consistency will be key.
What we really need to answer is: Where should we shoot from? What kind of shooter should we do? Bank shot or swish?
First of all, some physics:
Forget about 45 degree shooting. To hit the basket, we need at LEAST a 60 degree angle, and up to 80 depending on how close we want to shoot from. Varying shooting distances will necessitate a tilting shooter, a "sweet spot" wouldn't necessarily.
Next: How do we ensure consistency? The reason to go with a shooter is because we could potentially score more points than a dumper, but that would require a 67% shooting percentage or better. So: how do we get that accuracy? Like I've said, to get such accuracy requires at the minimum heavy rollers and a turreted shooter. However, running the shooter motor for 2 minutes WILL drain your battery - it happened to us in 06. So we can't run it all the time. So do we direct-drive the shooter, and wait the few seconds each time we want to shoot for the wheels to spin up, or do we make a clutch that can disengage the wheels and let then free spin? One sacrifices scoring speed, the other weight.
Additionally, should we have a sweet spot, or be able to shoot from several places? Obviously, having a sweet spot limits where you can score from, but being able to score from several places necessitates being able to tilt the shooter - which again adds more weight, moving parts and complexity.
Furthermore, each shot needs to be separated by some amount of time to allow the system to regain angular velocity. The question becomes: how long do we need between shots? Essentially, to score shoot 3 times, how long will it take?
Just a few questions to ponder. We're open to any designs at this point, and I'll weigh in as much as possible to do the physics calculations and such.
Since we decided to make a shooter, let's concentrate on making an excellent, consistent shooter. If we're going to win, consistency will be key.
Admin- Admin
- Posts : 10
Join date : 2012-01-14 -
Why is 45 deg. out?
ncwitte Sat Jan 14, 2012 9:01 pm
This article claims that 45 deg. is best for accuracy. While it's focused mostly on humans, the author also talks about automated shooting machines.
Apparently new members aren't allowed to post links for 7 days, Google "winning hoops building the perfect arc" and look at the first result.
Norm Witte
Apparently new members aren't allowed to post links for 7 days, Google "winning hoops building the perfect arc" and look at the first result.
Norm Witte
ncwitte- Posts : 2
Join date : 2012-01-14
Re: General Shooter Discussion - What? Where? How?
Admin Sat Jan 14, 2012 9:56 pm
To link:
http://www.winninghoops.com/pages/Feature-Articles---Building-The-Perfect-Arc.php
Basically, they're saying that for basketballl shooters, the optimal shot is a back rim and bounce down shot.
That's an interesting thought - shooting "deeper" into the basket. The only problem is shooting close up - you have to shoot at > 45 degrees just to get enough height to put it in. Is that an argument for a fixed-distance shooter?
From the article:
That's just for freethrows, though....
For reference, I put together simple models of balls being shot into the hoop from about the key (at 55 degrees) vs wedged between the wall and fender (70 degrees):
(all distances are in meters, and the hoop height is measured as height above 60 inches - that is, 60 inches is 0 on the graph)
http://www.winninghoops.com/pages/Feature-Articles---Building-The-Perfect-Arc.php
Basically, they're saying that for basketballl shooters, the optimal shot is a back rim and bounce down shot.
That's an interesting thought - shooting "deeper" into the basket. The only problem is shooting close up - you have to shoot at > 45 degrees just to get enough height to put it in. Is that an argument for a fixed-distance shooter?
From the article:
“We are the first group to systematically and scientifically confirm what makes great shooters, great,” Marty says. In addition to the solid shooting mechanics taught by shooting coaches, players must also master these three shooting principles, according to Marty:
• All great shooters shoot the ball straight.
• All great shooters shoot the correct distance (11 inches beyond the front of the rim).
• And all great shooters shoot a consistent, medium height arc of 45 degrees.
That's just for freethrows, though....
For reference, I put together simple models of balls being shot into the hoop from about the key (at 55 degrees) vs wedged between the wall and fender (70 degrees):
(all distances are in meters, and the hoop height is measured as height above 60 inches - that is, 60 inches is 0 on the graph)
Admin- Admin
- Posts : 10
Join date : 2012-01-14 -
Re: General Shooter Discussion - What? Where? How?
ncwitte Sun Jan 15, 2012 12:45 am
I think that the issue is whether you want to be able to vary the power. If you can vary power, you can keep your shooting angle a constant 45 deg, at least until you get up close. However, being able to adjust the power to the correct level for the distance I bet would be exceptionally difficult. So if you're going to have a variable angle, that means that at different distances we will have different angles. At some particular distance from the basket we will be close to 45 deg and that will probably be our sweet spot for getting the best shots off.
ncwitte- Posts : 2
Join date : 2012-01-14
Re: General Shooter Discussion - What? Where? How?
Admin Sun Jan 15, 2012 2:35 am
Just so everyone knows - it's a LOT easier to vary speed than angle. A LOT. To vary speed, all you need is an encoder to sense the speed of the wheels - actually, we did that on the Lunacy robot. Changing angle is a LOT more difficult - now you have to take into account the ball transport system not getting in the way and releasing the ball before the shooter grabs it, having a tilt system that will give accurate amounts of tilt, and such.
We'll need an encoder on the shooter to ensure constant speed anyway - you can't just give it a voltage or power level from the code and expect it to go the same speed every time - because it won't. And again, to be effective at all as a shooter we need accuracy and consistency - and that means very accurate speed control of the shooter.
However, if you don't vary angle and just speed, the robot won't be able to score up close. I'm going to do some more work with my graphs to look at 45 degree shots from various distances, as well as a "back rim and down" shot - the article suggests targeting 11" past the front of the rim, so that's what I'll look at.
We'll need an encoder on the shooter to ensure constant speed anyway - you can't just give it a voltage or power level from the code and expect it to go the same speed every time - because it won't. And again, to be effective at all as a shooter we need accuracy and consistency - and that means very accurate speed control of the shooter.
However, if you don't vary angle and just speed, the robot won't be able to score up close. I'm going to do some more work with my graphs to look at 45 degree shots from various distances, as well as a "back rim and down" shot - the article suggests targeting 11" past the front of the rim, so that's what I'll look at.
Admin- Admin
- Posts : 10
Join date : 2012-01-14 -
Nope, 45 degrees won't work. AT ALL
Admin Sun Jan 15, 2012 2:47 am
Actually, I just took another look at their data, specifically:
When they're talking about a 45 degree shot, they're not talking about the angle it leaves the player's hands, they're talking about the angle that it enters the hoop! Shooting at 45 degrees probably won't even be able to get the ball into the hoop - it'll bounce out!
(again, distances in meters and the dotted lines are the width of the ball as it travels over the solid line)
Also interestingly, the "key" for a robot, with the distances and heights of things, pretty much exactly represents a free throw in Basketball. Which I suppose makes sense, when you think about it!
So my original point: Forget 45 degrees. You need to shoot a LOT higher to even get to the hoop in most cases, let alone be anywhere close to accurate!
When they're talking about a 45 degree shot, they're not talking about the angle it leaves the player's hands, they're talking about the angle that it enters the hoop! Shooting at 45 degrees probably won't even be able to get the ball into the hoop - it'll bounce out!
(again, distances in meters and the dotted lines are the width of the ball as it travels over the solid line)
Also interestingly, the "key" for a robot, with the distances and heights of things, pretty much exactly represents a free throw in Basketball. Which I suppose makes sense, when you think about it!
So my original point: Forget 45 degrees. You need to shoot a LOT higher to even get to the hoop in most cases, let alone be anywhere close to accurate!
Admin- Admin
- Posts : 10
Join date : 2012-01-14 -
Re: General Shooter Discussion - What? Where? How?
MikeMac Sun Jan 15, 2012 10:24 am
From http://discovermagazine.com/2000/oct/featphysics:
Using lots of trigonometry, [Peter] Brancazio calculated the optimal angle of the arc from the free-throw line. If tossed at 32 degrees or less, the ball will likely hit the back of the rim. "That doesn't mean it won't go in, but it will certainly bounce off the metal and reduce the chance of success," he says. At angles greater than that, the ball has a chance of making a nice swish. The optimum angle, he calculated, is 45 degrees—plus half the angle from the top of the player's hand to the rim. "The shorter you are, the steeper that angle has to get to give you the best chance of making the shot," he says. Of course, lobbing a ball very high so that it comes down nearly straight into the basket would be the most efficient technique, but a shot like that "is almost impossible to aim," says Brancazio. Instead, he says, his formula makes it possible for a player to shoot with the largest possible margin for error.
Also see the more mathematical discussion in this paper by Howard Penn:
http://mathaware.org/mam/2010/essays/PennJumpShot.pdf
Using lots of trigonometry, [Peter] Brancazio calculated the optimal angle of the arc from the free-throw line. If tossed at 32 degrees or less, the ball will likely hit the back of the rim. "That doesn't mean it won't go in, but it will certainly bounce off the metal and reduce the chance of success," he says. At angles greater than that, the ball has a chance of making a nice swish. The optimum angle, he calculated, is 45 degrees—plus half the angle from the top of the player's hand to the rim. "The shorter you are, the steeper that angle has to get to give you the best chance of making the shot," he says. Of course, lobbing a ball very high so that it comes down nearly straight into the basket would be the most efficient technique, but a shot like that "is almost impossible to aim," says Brancazio. Instead, he says, his formula makes it possible for a player to shoot with the largest possible margin for error.
Also see the more mathematical discussion in this paper by Howard Penn:
http://mathaware.org/mam/2010/essays/PennJumpShot.pdf
MikeMac- Posts : 2
Join date : 2012-01-15
Re: General Shooter Discussion - What? Where? How?
MikeMac Sun Jan 15, 2012 11:29 am
Seems to me that the Where? What? How? question shakes itself out into three different shooting paradigms of progressively increasing difficulty, relative to the three problems of angle to the basket [left-right], elevation of the release [up-down], and distance to the basket [force required].
The "Power Forward": shoots from fixed position, posted up against the fender or into the fender/backboard angle
--distance to the basket known with high accuracy without special systems (same as dumper); no ranging system required
--most forgiving error bar for L-R angle to the basket
--shooter could have fixed L-R angle, traverse aiming done with entire robot (using arena structures); no turret or complex sighting system required
--most forgiving error bar for elevation of the release; shooter could have fixed elevation angle as well
--least powerful shooting mechanism needed (could allow for types of shooting mechanisms not workable at greater distances)
The "Free Throw Shooter": shoots from protected key
--distance to basket only approximately known; simple ranging system probably required
--smaller error bar for L-R accuracy; requires fine control of robot position and orientation if no turret used
--difficult to use arena to orient robot to basket; simple sighting system probably required
The "Shooting Guard": shoots from anywhere on the floor within its range
-- must be able to determine distance to basket (ranging)
-- must be able to determine orientation to basket (sighting/aiming)
-- premium on accuracy increases with distance from basket; could operate "in the paint" (between key and fender)
-- fire-anytime-anywhere likely requires a turret-mounted shooter which can quickly orient itself to the basket
I have high confidence in our ability to build a Power Forward, but if we can't get the accuracy we need this design could most readily be changed to a dumper at the 11th hour. A Free Throw Shooter would require more powerful and more precise shooting mechanism and probably also ranging and sighting mechanisms, which would test our applied (or borrowed) creativity and our workmanship; the only reason we might be able to pull one off is that the set of all shooting solutions with the robot in contact with the key is much smaller than that for for a Shooting Guard. I think the Shooting Guard is beyond us at this point, and would recommend we focus our brainstorming on the other two paradigms with the intention of picking one of them before long.
The "Power Forward": shoots from fixed position, posted up against the fender or into the fender/backboard angle
--distance to the basket known with high accuracy without special systems (same as dumper); no ranging system required
--most forgiving error bar for L-R angle to the basket
--shooter could have fixed L-R angle, traverse aiming done with entire robot (using arena structures); no turret or complex sighting system required
--most forgiving error bar for elevation of the release; shooter could have fixed elevation angle as well
--least powerful shooting mechanism needed (could allow for types of shooting mechanisms not workable at greater distances)
The "Free Throw Shooter": shoots from protected key
--distance to basket only approximately known; simple ranging system probably required
--smaller error bar for L-R accuracy; requires fine control of robot position and orientation if no turret used
--difficult to use arena to orient robot to basket; simple sighting system probably required
The "Shooting Guard": shoots from anywhere on the floor within its range
-- must be able to determine distance to basket (ranging)
-- must be able to determine orientation to basket (sighting/aiming)
-- premium on accuracy increases with distance from basket; could operate "in the paint" (between key and fender)
-- fire-anytime-anywhere likely requires a turret-mounted shooter which can quickly orient itself to the basket
I have high confidence in our ability to build a Power Forward, but if we can't get the accuracy we need this design could most readily be changed to a dumper at the 11th hour. A Free Throw Shooter would require more powerful and more precise shooting mechanism and probably also ranging and sighting mechanisms, which would test our applied (or borrowed) creativity and our workmanship; the only reason we might be able to pull one off is that the set of all shooting solutions with the robot in contact with the key is much smaller than that for for a Shooting Guard. I think the Shooting Guard is beyond us at this point, and would recommend we focus our brainstorming on the other two paradigms with the intention of picking one of them before long.
MikeMac- Posts : 2
Join date : 2012-01-15
Page 1 of 1
Permissions in this forum:
You cannot reply to topics in this forum