Re: Autonomous Longer????
 

I think the problem this year is that whoever thought of the keeper did not think through the real bonus of using it. In a row of 7 if a single keeper is in the middle the other team has to spoil right next to it.

RRRRSRR which is worth 20 points.

Without the keeper it is:
RRRXRRR which is worth 16 points.

So essentially a keeper saves you four points during the few times that spoilers actually come out.

On a totally different note I think that FIRST should make the camera a little easier to use. Just getting the camera to track the light consistently at a regional is tough. In week 4 they had someone with light values at our regional, but he had no idea how to use them with our code (We use Kevin Watson's code base) so we had to get our own values with an IFI program we hacked. FIRST should make one standard way of using the camera in EASYC and in regular C and then give camera values specifically for that and give much more support for using the camera.

A longer autonomous mode will just punish teams without good programmers. Making autonomous a larger portion of the game just hurts teams at high schools without programming classes. I am lucky to be at a high school with programming classes so we have plenty of programmers with experience, but not many other high schools have that. In my mind autonomous mode should be short (10-15 seconds) and be helpful in the game, but not dominant.

Re: Autonomous Longer????
 

Autonomous longer? I don't want to have to watch an even longer period of nothingness.

Make it worth a lot more points, or it's not gonna happen.

Re: Autonomous Low/Med/High Scoring
 

My two cents worth,
I like the idea of having low/med/high difficulty objectives in the autonomous mode. This rewards/challenges the programming team and makes it an integral part of the team (this year, why bother with auto when only worth "golden" two points - the mechanical drivetrain team trumphed the programming team for "robot time"
at the regionals - should have gone home and not spent two vacation days in preparation for limited "robot time")?

I was part of a small team and as a result had no operating robot until the regionals and very little "robot face time". Lesson learned: Only more experienced teams
that finish robot early , before regionals, can really utilize the programming skills necessary for autonomous operation.

Lastly, in 2005 competition with Tetra's, at the end of auto
mode you could continue to score with the Tetra, so having possession and positioning a tetra for quick scoring in tele-operated mode was useful. My recommendation is that this feature be re-introduced into the 2008 game play. This
year the auto tube was worthless and needed to be discarded.

Re: Autonomous Low/Med/High Scoring
 

The tetra year there were objectives of varying difficulty. You could start with a tetra to place, you could knock down the tetra that was hanging, or you could go for the green band tetra.

Re: Autonomous Low/Med/High Scoring
 

And something that I'd like to see again:
If you had a really fast robot and an excellent programming team, you could've theoretically accomplished multiple objectives in autonomous!

Re: Autonomous Longer????
 

Really autonomous needs one thing to get better; incentive. Without a reasonable incentive to try, autonomous will not get the resources over teleoperated mode, and it will not improve. There needs to be multiple tasks that can score or otherwise contribute to your alliance, and they need to be of varying difficulty (and of course, varying reward as well). 2006 proved that if the autonomous mode is made important, teams will at a minimum try to move (the "254" for 15 seconds auto), far better than the sit and do nothing that has become customary in 2007 and was back in 2005 as well.

Autonomous must be given more point potential to takeoff, and if the games are designed with weak autos because it is believed too many teams don't have the programmers to try it, then the autonomous mode will continue to remain unused and pointless. If there are really worries that few teams will be able to do it even if they try, then I advocate doing part of the game as an "optional" autonomous; teams can complete the task(s) during a certain part under human control, but if they elect to instead throw an "auto" switch and do the task autonomously, they will receive a 2X (or even a 3X) multiplier for all the points scored. This is not insurmountable, since a team can focus on driving only and try to outscore the autonomous machines in the normal match, but it gives a substantial bonus to those willing to try it, and eliminates the "waiting period" that we've grown accostomed to at the beggining of matches.

I think all of the time periods so far have been more than enough for the tasks, if FIRST comes out with a very time consuming task for an auto they could extend it more, but I wouldn't push it too long (either that or I'd mix it in with teleoperated mode, as I stated above).

Re: Autonomous Longer????
 

I think for more teams to attempt autonomous it needs to have a greater affect on the game oucome. Like last year you were able to win autonomous for a bonus and continuing scoring into the last round. i fell with larger rewards more teams will create and devote more time to at least attempting autonomous.:D

Re: Autonomous Longer????
 

I'd love to see something like 90 seconds of autonomous followed by 90 seconds of teleoperation. Or maybe 75/75 or 60/60 if 3 minutes is too long.

Or have an autonomous period at both the start and the end of the match.

--AJY

Re: Autonomous Longer????
 

In my opinion, teams that thought the autonomous mode wasn't worth it this year, especially those that see it as only 2 points, really didn't examine the strategy and potential game play well enough. A keeper is worth anywhere between 2 and 132 points. It doubles the value of the row(s) it is in, just like any other game piece. And with this exponential scoring, a single piece plays a MASSIVE role in closely contested matches. http://www.chiefdelphi.com/forums/sh...9&postcount=16
Given the maximum potential score of the game (assuming nobody has found a way to elevate 3 robots, and that no tubes were de-scored and then re-scored elsewhere) is 656 points, the possible 132 point swing of each keeper is massive. In that 656 point configuration (2 rows of 8, 1 row of 5, 5 rows of 3, 3 rows of 2, 2 robots elevated), the value of the 3 keepers ranges from 40 points (6.1%) to 396 (61.3%) depending on placement. Autonomous potentially worth MORE THAN HALF of the max score, I don't think that's an worthless autonomous.
Sure, that situation will likely never happen (I don't think a "max score" has happened since 2001), but let's put it into a more realistic scenario. Redabot is an excellent scorer when left alone, but struggles when defense is applied. BlueAlliance is comprised of rampbots who play defense (and/or fail to score ringers). Redalliance has no ramps/lifts/platforms to score any bonus points. Redabot gets held to a single ringer until the last 30 seconds, when BlueAlliance goes back to score their "guaranteed" 60. During that time, Redabot quickly puts up 3 more tubes, but then the clock expires. One of the Blueabots doesn't make it up the ramp (or falls off, or is supported by a ringer, or is touching the wall, or..), giving the blue alliance 30 points. The Red Alliance has 16. If a keeper had been scored, the final score would have been 32-30 in favor of the other alliance. Well, what if the blueabot didn't fall off? Well, it would have only taken 1 more tube for red to have a 64-60 victory.
Look at the experiences of teams like 1902, 67, and 1114, and see the results they had when they scored and missed keepers. In GLR SF 2-2, if 1114 hits their keeper on the bottom, it would have extended their bottom row to 3, and inhibited the blue alliance from scoring a row of 6 on the bottom. If that happens, it suddenly switches from 68-56 Blue, to a 64-36 Red (the would also have been under a red tube), from -12, to +28, a swing of 40 points. 1902 has already attested how a keeper allowed them to reach the finals, and missing one kept them from winning the regional. Keepers are valuable people, it's sad so few teams saw that.

Re: Autonomous Longer????
 

True true. Far too many have thought one tube won't matter, when in fact one tube has decided a number of matches. Of course, this seems to have happened every year in autonomous; everyone thinks it will be of little value, and only once competition starts does the full potential of it get demonstrated (last year most teams still didn't have an auto at the beginning of week 1 and 2 regionals; as the regionals played out though, most robots at least tried to block at the end, because everyone recognized the advantage autonomous gave teams).

Also, even if you're looking at just raw tube value (not the potential value if you put up all the other tubes on the field), autonomous still has the potential to score your alliance up to 24 points; there are 3 keepers per alliance, and 9 ringers on the opposite end of the field, all of which can be legally scored in the autonomous period. I would greatly like to see some of the teams scoring keepers consistently go pick up tube to prepare for tele-operated mode, or even try to score that tube (most teams are done scoring keepers in less than 10 seconds anyways). I know we'll be trying it, and for teams that are beyond us and already had keeper scoring working, this would probably be easy work for their programmers.

Re: Autonomous Longer????
 

But all your examples involve robots that are already proficient at scoring. If you can put up enough ringers to make a keeper very valuable, you're probably the calibre of team that has time during the build period to actually do an autonomous mode without sacrificing other elements of the robot. If a team is struggling to make an arm that can do ANYTHING, spending time to work out an autonomous mode is counterproductive.


Plus, keepers are almost exclusively placed on the near side of the rack (or at least, they would be for a basic autonomous mode). For a team that is capable of scoring reliably, this might save them 10 seconds, but they'll need much more than that to score on the opposite side coping with increased defense and decreased visibility.

I think it is less of an active decision to ignore autonomous, and more of a question of effort allocation: If you've got an arm that you can expect to score 4 ringers per match, and even assuming that a keeper will make it 5 per match, the effort to mount all the autonomous sensors, write, and test autonomous mode might be more than the effort required to simply upgrade the arm or arm user interface. Moving from a driver-controlled arm to a PID-controlled arm is probably worth quite a few ringers per match, and might take the same effort as making an autonomous mode (it's also easier to test).

I suppose with a LOT of good scouting information or SOAP-watching you could determine the expected value of a keeper and determine what the real-world expected value of autonomous is. That'd be an interesting project.

Re: Autonomous Longer????
 

Which than comes to the question of "Why are there so few proficient teams?" and that leads to the question of the game design.

Re: Autonomous Longer????
 

Or manipulator design....

Re: Autonomous Longer????
 

Or the fact that you are trying to build something in six weeks that usually takes monthes, multiple prototypes, and lots of mistakes to actually come out with a suitable design.

Re: Autonomous Longer????
 

I dont think that last years autonomous mode had a big impact on the game. Sure you couldnt spoil a keeper, but I rarely say a spoiler placed. I also did notice that most of the teams that won awards at the championship had autonomous mode(s).