Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Rules/Strategy (http://www.chiefdelphi.com/forums/forumdisplay.php?f=6)
-   -   Observations after week ! (http://www.chiefdelphi.com/forums/showthread.php?t=75309)

Tom Line 02-03-2009 09:04

Re: Observations after week !
 
Quote:

Originally Posted by jblay (Post 829856)
another thing that really stuck out to me was that basically any shooter that was effective was in essence a dumper in that they would try and pin the trailer and would only shoot from close range. i don't think i saw anyone actually shooting the entire weekend.

We spent the weekend working on our turret aiming. A realistic framerate on the camera with the given Labview Find Color code is between 10-15 FPS. We are using 160x120 decimated twice to bring it up to 15 FPS pretty consistently, but it dips to 10 on occasion. We simply didn't have the labview knowledge or the time to further optimize the supplied routine.

10 Frames per second. Compare that with a robot speed, of, say 10 feet per second. At that speed, you get one frame every robot foot. You finish processing that frame in about .1 seconds. So by the time you tell your turret to turn, your target is already 1 foot from where it was before.

Unless you're using a targeting procedure that calculates absolute position then figures out how to lead the target, you're going to have a VERY hard time hitting that trailer when you include travel time of the ball.

There are ways of fudging it that work to varying degrees, but adding in the new control system this year and any team that pulls it off in the first couple weeks will be superhuman in my book. I have no doubt that the teams who have the funding to run a second Crio and practice robot may have it working pretty well in a couple weeks - see simbotics and chickens, etc.

thefro526 02-03-2009 09:19

Re: Observations after week !
 
Quote:

Originally Posted by johnr (Post 830135)
Did any teams get caught trying to avoid a g-14? You know, keep the score close and then the other teams toss fuel cells. Or worse, you over fill your own trailer.Just wondering if anyone is now saying,"We will take the 2x but try to avoid 3x"

We did this in our first Quarter Final Match. We were up 80 something to 30 something and everyone stopped scoring and just moved around. The final score ended up being 36-90 in our favor.

The big thing about avoiding a G14 is that you should always be in a scoring position and chose not to score, so then if the other alliance uses a Super Cell or opens out into one of your trailers you can be ready to counter-attack. You don't want to be stuck trying to avoid a G14 and then losing because you couldn't score in time.

Luke Pike 02-03-2009 10:02

Re: Observations after week !
 
Quote:

Originally Posted by Tom Line (Post 830155)
We spent the weekend working on our turret aiming. A realistic framerate on the camera with the given Labview Find Color code is between 10-15 FPS. We are using 160x120 decimated twice to bring it up to 15 FPS pretty consistently, but it dips to 10 on occasion. We simply didn't have the labview knowledge or the time to further optimize the supplied routine.

10 Frames per second. Compare that with a robot speed, of, say 10 feet per second. At that speed, you get one frame every robot foot. You finish processing that frame in about .1 seconds. So by the time you tell your turret to turn, your target is already 1 foot from where it was before.

Unless you're using a targeting procedure that calculates absolute position then figures out how to lead the target, you're going to have a VERY hard time hitting that trailer when you include travel time of the ball.

There are ways of fudging it that work to varying degrees, but adding in the new control system this year and any team that pulls it off in the first couple weeks will be superhuman in my book. I have no doubt that the teams who have the funding to run a second Crio and practice robot may have it working pretty well in a couple weeks - see simbotics and chickens, etc.

The trick is in positioning yourself. Just like a dumper, you can't just shoot anytime you want. The best position is following behind a trailer and matching speed. Another good time to shoot is when a robot is switching directions and is not moving fast.

Dumpers have to be right up against a trailer to score, shooters need to have a good line of fire.

wcamp1742 02-03-2009 10:07

Re: Observations after week !
 
Quote:

Originally Posted by spc295 (Post 829534)
were there any turreted shooters that did well, if so how were they tracking?

Ours is a turreted shooter and we qualified as the #1 seed and won the regional. During teleoperated our gunner controlled the movements of the turret with his joystick, without a tracking code, and shot a fairly hghi percentage. However, we had different speeds we could shoot at and ended up basically dumping at point blank range. We were able to follow trailers well enough we didn't have to shoot many long range shots unless we had to shoot over a bot.

Adama 02-03-2009 14:34

Re: Observations after week !
 
Quote:

Originally Posted by XXShadowXX (Post 830137)
Defense can make a big difference, by pinnign the moust powerful robot on the other alliance you can make a massive difference.

There were several instances where we were driven into a corner and held there when our alliance mates were incapable/ very bad at scoring. Then the remaining robots on the opposing alliance had their way with the rest of our alliance. It wasn't pretty though it was a sign of respect in a twisted sorta way.

JesseK 02-03-2009 15:30

Re: Observations after week !
 
Save the information on your wireless receiver to disk. There is a 'download configuration' button on the configuration page. That way, if you ever replace the receiver and also misplace your WPA key, you have something to load into a new receiver. We were the guinea pigs for this, and it worked like a charm.

Move in autonomous. Put purpose behind your movement if you can; if everyone crashes into each other in the middle there's really no way you can effectively begin executing a strategy in teleop.

Talk with other teams' human players. Show the Outpost person where the empty cell must go. Remind everyone if they should load your bot's hopper as you pass by. Help them out by waving wildly during the match if you have to.

The hardest part to get used to in this game is its ever-changing nature. Your strategy MUST change in response to your opponents if they get the upper hand. Sticking to a plan that is being effectively countered is what got us in a bind in too many matches. Of course, your alliance partners have to agree and be willing to cooperate with a new plan -- not so easy in Qualifications :rolleyes:.

Do not be afraid to go on defense in a split second. Do not be afraid to take your super-star bot and pin for 10-20 seconds if the other 4 bots are on the opposite side of the field.

Generally, fan drive trains > swerve > skid steer with traction control. Yet for defense, crab/omni bots conquer all :cool: Note that this has nothing to do with what's above the drive train.

If you're in a traffic jam, figure out how to get out of it. If you're in open field, there's some action going on and you're obviously not part of it ;) at that moment. Choose those moments wisely.

Trailers without robots mean death -- you can have a great strategy for dealing with it, yet even if you properly execute them there are simply too many counters to those strategies. Go on Thursday and make sure everyone's working and inspected!

TigerDrive09 03-03-2009 21:05

Re: Observations after week !
 
Quote:

Originally Posted by Tottanka (Post 829672)
And what was that?


Apparently we had our modules switched and we were supposed to have power running to one of them, it was a quick fix but no one was able to find it until our 4th qualifying match.

nukelear1991 03-03-2009 22:25

Re: Observations after week !
 
We have a good strategy that fits the outpost. Not many teams at traverse city wanted to start with robot or human player on the outpost, but after a ("NO COMM") day of competition on Friday, we had lots of time to figure out a game plan. We start autonomous and pin the other outpost player to the wall and then have our human player fill up their trailer. During the match we pin the heck out of other teams and try to get one decent dump. I know we aren`t AMAZING like some teams, but what kind of things could a team like us do to improve our strategy?

Thanks Guys,
Mike P.:cool:

longboardlover 04-03-2009 17:26

Re: Observations after week !
 
So to those of you who have already attended a competition, how important is hopper size? Our robot can safely (without jamming) hold 7-8 balls, but we are capable of a fairly quick turnover and have great accuracy (90%-ish), but after seeing all of these teams with robots can hold 20 balls it makes we worry.

Alan Anderson 04-03-2009 17:58

Re: Observations after week !
 
Quote:

Originally Posted by longboardlover (Post 831791)
So to those of you who have already attended a competition, how important is hopper size?

Hopper size is only one part of the scoring equation. Even if you can carry all the moon rocks on the field at once, you still need to be able to deliver them quickly or risk your target getting away while you're still trying to dump on it. A larger hopper can make a big difference if you can empty it effectively.

Vikesrock 04-03-2009 18:01

Re: Observations after week !
 
Quote:

Originally Posted by longboardlover (Post 831791)
So to those of you who have already attended a competition

From what I saw on the webcasts it looked like hopper size gives you strategic flexibility. Many robots that had giant hoppers would score small bunches of 5-7 balls at a time throughout the match instead of counting on the 1 giant dump. I think it is definitely possible to be successful with a robot with a capacity of 7-8 balls, you just have to make sure your strategy fits the capabilities of your robot.

thefro526 04-03-2009 18:05

Re: Observations after week !
 
Quote:

Originally Posted by longboardlover (Post 831791)
So to those of you who have already attended a competition, how important is hopper size? Our robot can safely (without jamming) hold 7-8 balls, but we are capable of a fairly quick turnover and have great accuracy (90%-ish), but after seeing all of these teams with robots can hold 20 balls it makes we worry.

A large hopper's effectiveness depends on strategy. We can hold 20+ balls in our scoring mechanism. We never held this many though because it wasn't how we chose to play the game. If you only hold 7-8 but you can deliver them in a fast or even near instant punch with deadly accuracy you'll be fine. A lot of robots that have high capacity take a while to offload (>3-4 seconds) or can't offload with high accuracy so that's part of the equation too.

Really, high ball capacity is only effective if the robot has the mechanism to back it up.

GillSt.Bernards 04-03-2009 20:16

Re: Observations after week !
 
I noticed
- most of the scoring was being done by the human players
- watch the static
- do not throw the obit balls over the wall
- most of the teams are doing an offensive strategy and it is good to have at least one defensive robot on your alliance

BradMello 05-03-2009 23:36

Re: Observations after week !
 
Quote:

Originally Posted by longboardlover (Post 831791)
So to those of you who have already attended a competition, how important is hopper size? Our robot can safely (without jamming) hold 7-8 balls, but we are capable of a fairly quick turnover and have great accuracy (90%-ish), but after seeing all of these teams with robots can hold 20 balls it makes we worry.

It also depends on the pickup speed of your robot. Many robots that have a high hopper capacity naturally take longer to fill up their hopper compared to medium sized hoppers like yours. Many dumper teams that I observed unloaded their hopper all at once, regardless of their accuracy. This means that if they miss, they miss big. A benefit having a medium sized hopper (10-11 moon rocks) is that if you miss a few of your balls, you can easily recover. A few times that we missed dumping the majority of our balls, 20~~ it was devastating


All times are GMT -5. The time now is 19:10.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi