Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   General Forum (http://www.chiefdelphi.com/forums/forumdisplay.php?f=16)
-   -   Fetcher/Strategy (http://www.chiefdelphi.com/forums/showthread.php?t=89421)

XaulZan11 19-01-2011 12:00

Re: Fetcher/Strategy
 
I think that even if you have a fetcher bot, the scoring robot(s) should still get their tubes, but instead of scoring 1 per trip, they can now score 2. Assuming the scoring robot can pick up and place quickly, it is a waste for them to sit and wait for their partners to bring the tubes to them.

Molten 19-01-2011 12:10

Re: Fetcher/Strategy
 
Quote:

Originally Posted by Chris is me (Post 1003337)
Because there will be a ton of scoring bots that are completely inept at getting their own tubes...

If you're playing defense on the worst robot on the alliance, the alliance has already won. It's lose - lose: Either play a disadvantageous 2v2 matchup, or let two very good robots score even faster than you.

I think we're just envisioning the setup differently. The ideal setup for this strategy would be 2 robots scoring while 1 zooms up and down the field providing tubes. Defense comes in and stops the speed bot. The 2 robots must then leave the scoring area and head down field in order to get their own tubes, when they get to the other side of the field, you have 3 robots all in close proximity. Its a defensive dream. You can keep all 3 in that area for 20 seconds leaving your alliance free to score. I know this is an exaggerated scenario. I just want to make one thing clear. If your going to be a fetch bot, be ready for this sort of thing happening. Figure out what is the best move if your opponent does use defense. There are ways to nullify defense if you know what your doing and to take them away from the better scorers. However, there are few that can do it if the defense driver knows what they are doing. From a defensive perspective, I want all the robots on the opposing alliance doing the same thing at the same time or to all be in the same area. I could see potential for said bot, but you better be a master strategist if your planning on consistently winning with the fetch bot. 3 offensive robots make strategy simple, 3 defensive robots make the strategy simple, 2 defense/1 offense or 2 offense/1 defense make strategy simple. Having a robot that is primarily for fetching makes things complicated. I'm not saying it can't be successful. I think it can be, its just going to be a hard card to play well all the time.

Chris is me 19-01-2011 12:17

Re: Fetcher/Strategy
 
That's definitely a good point. A "pure" fetchbot with zero ability to score would need to worry, though I do think the problems are exaggerated. As breakaway showed, once you get to a 2 offense versus one defense scenario, one robot quickly wins.

The best thing to do would be to design a robot that can score well, and if it turns out you suck at that, become a great feeder.

Why do I feel like this is totally what's going to happen to Shaker?

XaulZan11 19-01-2011 12:19

Re: Fetcher/Strategy
 
Quote:

Originally Posted by Molten (Post 1003494)
I think we're just envisioning the setup differently. The ideal setup for this strategy would be 2 robots scoring while 1 zooms up and down the field providing tubes. Defense comes in and stops the speed bot. The 2 robots must then leave the scoring area and head down field in order to get their own tubes, when they get to the other side of the field, you have 3 robots all in close proximity...

I don't think you will ever see 2 robots sit in the scoring zone while 1 robot gets them tubes. That must be one VERY fast robot!

Instead, I think you will see all 3 robots get tubes with only 2 of them placing. As a defender, who are you going to stop? Probably their best placer, while letter the fetcher go free because it cannot score.

Molten 19-01-2011 13:20

Re: Fetcher/Strategy
 
Quote:

Originally Posted by XaulZan11 (Post 1003502)
I don't think you will ever see 2 robots sit in the scoring zone while 1 robot gets them tubes. That must be one VERY fast robot!

Instead, I think you will see all 3 robots get tubes with only 2 of them placing. As a defender, who are you going to stop? Probably their best placer, while letter the fetcher go free because it cannot score.

Why not both? With 6 robots on the field, maneuvering can often become a challenge. Its completely possible for 1 robot to stall(not completely stop) 2 robots to cut their opponents score in half. I've seen one robot take on 3 and held them in a corner for a full 30 seconds without ever pinning one of them. Its actually easier with more robots if you can just get them in a semi-enclosed environment(ie the corner).

XaulZan11 19-01-2011 14:55

Re: Fetcher/Strategy
 
Quote:

Originally Posted by Molten (Post 1003543)
Why not both? With 6 robots on the field, maneuvering can often become a challenge. Its completely possible for 1 robot to stall(not completely stop) 2 robots to cut their opponents score in half. I've seen one robot take on 3 and held them in a corner for a full 30 seconds without ever pinning one of them. Its actually easier with more robots if you can just get them in a semi-enclosed environment(ie the corner).

I'm not exactly sure what you are getting at.

I agree that an alliance can get it its own way and thus easier to defend, but I don't think that is unique just for alliances with a feeder/fetcher. All tube placing alliances will need to spread out and not get in each others way.

Molten 19-01-2011 16:56

Re: Fetcher/Strategy
 
Quote:

Originally Posted by XaulZan11 (Post 1003594)
I'm not exactly sure what you are getting at.

I agree that an alliance can get it its own way and thus easier to defend, but I don't think that is unique just for alliances with a feeder/fetcher. All tube placing alliances will need to spread out and not get in each others way.

My only real point was that in defense you don't have to choose one person. You said that the defense has a choice between either the fetcher or the main scorer...the truth is you can get both. You are completely right that this is true for all alliances. I just don't want teams to think the strategy this year is going to be as simple as it has been in the past.

XaulZan11 19-01-2011 17:02

Re: Fetcher/Strategy
 
Quote:

Originally Posted by Molten (Post 1003689)
My only real point was that in defense you don't have to choose one person. You said that the defense has a choice between either the fetcher or the main scorer...the truth is you can get both. You are completely right that this is true for all alliances. I just don't want teams to think the strategy this year is going to be as simple as it has been in the past.

Ok, I see what your saying and I agree. With more than one offensive threats, I forsee defenders playing a zone defense, or staying in the same region of the field and playing defense on whoever is in that region. Of course, if your playing 1114 and box bots 3xxxx and 3xxxx, then I would stick on 1114.

Shaf2909 19-01-2011 17:41

Re: Fetcher/Strategy
 
Quote:

Originally Posted by XaulZan11 (Post 1003488)
I think that even if you have a fetcher bot, the scoring robot(s) should still get their tubes, but instead of scoring 1 per trip, they can now score 2. Assuming the scoring robot can pick up and place quickly, it is a waste for them to sit and wait for their partners to bring the tubes to them.

Okay here is a break down of one way I was thinking:
  1. One robot racing back and forth getting its own tubes and hanging them.
  2. One robot Stays in the scoring zone.
  3. Zebrabot race to its feeding station and run from the feeding station to mid field and passes to the robot staying in the scoring zone and back..so on Fetch pass, fetch pass ect.
.

Zebra bot and scoring zone bot stay on the same side so that the bot racing back and forth can have a vertical peg row all to its self..

This is if the defense they are throwing at us is avoidable.

Molten 19-01-2011 23:05

Re: Fetcher/Strategy
 
Quote:

Originally Posted by Shaf2909 (Post 1003722)
Okay here is a break down of one way I was thinking:
  1. One robot racing back and forth getting its own tubes and hanging them.
  2. One robot Stays in the scoring zone.
  3. Zebrabot race to its feeding station and run from the feeding station to mid field and passes to the robot staying in the scoring zone and back..so on Fetch pass, fetch pass ect.
.

Zebra bot and scoring zone bot stay on the same side so that the bot racing back and forth can have a vertical peg row all to its self..

This is if the defense they are throwing at us is avoidable.

Ok, I think this is the part that alot of people aren't getting and that many others have tried to make.

To anyone building a fetch bot:
A. How long does it take for you to get a tube from the feeding station to the robot waiting on you and back to the feeding station?
B. How long do you think it takes for a team to pick up that tube and place it on the rack and be prepared to pick up another tube?
C. If A>B then the strategy is somewhat wasteful because it has a team sitting there waiting for your robot to bring the next piece when they could be getting their own piece.

Lets say it takes 15 seconds for the average team to pick up the tube, place it, and go back down to pick up the next one. For a fetch bot to simply drive the length of the field and back(no time needed to pick up/release tube) it would have to drive 7.2 feet per second. This seems reasonable if there isn't any other robots in your way the entire trip. If you want to consider they won't be traveling in a straight line and will probably run into opposition along both ways, and that they have to slow down to grab the tube and slow down to drop the tube and the time to simply turn around, I think your talking more like 10 feet per second necessary even if they only drive 2/3 of the way. This is possible. But all of that is assumed 15 seconds to pick up the tube. What if they take less? They sit and wait. That isn't very productive, instead they could be getting their own tubes. That is why alot of the veterans aren't really opting for the idea of a fetcher as their sole means of getting tubes. Sure, having a robot that fetches and leaves the tube there for you to pick up when you get back with your own tube is somewhat helpful, but a robot that can score that tube on the bottom row is going to be far more of an asset because they can continue to get their own tubes and so could you. No need to pick up the same tube twice.

I'm not saying the idea of a fetcher is bad. In theory I like it, but I just don't see it being a viable contender unless your robot is going very fast. In fact, if you do manage getting it to go fast enough, build to dodge well and prepare for some serious collisions. I'd be surprised if a defense bot wouldn't be willing to "take one for the team".

ThirteenOfTwo 20-01-2011 02:28

Re: Fetcher/Strategy
 
@Molten: But if the time to score is less than the time required to fetch, the team whose job it is to score can afford to take part off the match--say, the first 20 seconds, when defense is the most important since the top goals aren't full--off to play defense.

Also, keep in mind that with the two-feeder strategy it's the sum of the two feeding rates that matters. I don't think the bottleneck will be the scorer running out of tubes; rather it'll most likely be the feeders giving too many.

Molten 20-01-2011 10:35

Re: Fetcher/Strategy
 
Quote:

Originally Posted by ThirteenOfTwo (Post 1004087)
@Molten: But if the time to score is less than the time required to fetch, the team whose job it is to score can afford to take part off the match--say, the first 20 seconds, when defense is the most important since the top goals aren't full--off to play defense.

Also, keep in mind that with the two-feeder strategy it's the sum of the two feeding rates that matters. I don't think the bottleneck will be the scorer running out of tubes; rather it'll most likely be the feeders giving too many.

Perhaps a 2 feeder/1 scorer system would make more sense then a 1 feeder/ 2 scorer system. I also agree that then the bottleneck would happen on the scoring end, but then that leaves the scorer 7.5 seconds to place each tube in order to fill up the rack. That assumes that they start immediately which is reasonable considering they all get to start with a game piece and assuming the fetchers can't score the ubertube it could start with 2 in the immediate vicinity. Honestly, this setup is the first I've heard that I see potential for. It could work amazingly well versus some alliances. I don't see any obvious flaws in the strategy as long as you can get the timing down. I think you'll definitely have to wow people with your drive train or do some major networking if you want picked with this strategy for elimination though.

IndySam 20-01-2011 11:27

Re: Fetcher/Strategy
 
History shows us that few teams will be able to effectively quickly score on the top pegs at regionals especially in week one. So strategy will constantly evolve as always. Most matches will have only one or even no effective top peg scorer that can also pick up from the floor. In that case a fetcher would be useless. But that same useless fetcher will be scouted by the top teams and can become very valuable as the second pick for eliminations.

A good early regional finals alliance might consist of one Scorer and one feeder working together with a third robot running defense or interference.

The scorer gets their own tubes and also places the tubes provided by the fetcher.

Defensive robot is either providing picks for the scorer/fetchers or playing defense against other alliance.

Having said all that, I never want to design a robot that's goal is to be that second pick.

JesseK 20-01-2011 12:44

Re: Fetcher/Strategy
 
I don't think a fetcher would work well in a fetcher+2xhangers alliance setup. There's just not enough flexibility to respond to situations on the field, and the fetcher's utility is diminished towards the end of the match if it has done its job very well.

For a strictly 'fetcher' bot (good name btw), I would say a reliable fetch and a reliable minibot are a must. The allliance would be composed of a good hanger, a good fetcher, and a defensive robot (who should also have a minibot). The Hanger and the Fetcher gather tubes for the first minute minute or so (2-3 trips each) at which point the Hanger begins hanging and the Fetcher keeps fetching. At that point we're looking at 6-9 tubes available for hanging (more under no defense), with minibot options for endgame. The idea here is to prevent clutter in the scoring zone by only keeping 1 robot in the scoring zone during Hanging, thus freeing up the other 2 robots to do other tasks (fetch, defense, minibot) and still having a viable option to respond to the opposing alliance.

Thus, to incorporate it all into requirements for a fetcher bot:
-- Fast, as in, able to minimize the time traversing from one safe zone to the other (12+ fps for heavy bot, 14-16 fps if it's like 148's 2008 bot)
-- Able to avoid defense: either a small footprint + agile drive train (tumbleweed) OR a heavy robot + drive train shifting to push through defense
-- Effective pickup mechanism from the far side (Quick, "smart alignments" etc)
-- Can signal to the HP which game piece it wants
-- Can snipe opponent tubes from the floor
-- Effective drop-off mechanism on the close side ("drive-by" delivery desired)
-- An effective MINIBOT that always gets at least 10 points
-- Competent drive team that can place tubes in the position desired by the Hanger bot (e.g. all triangles in 1 area while all circles in another; yet directly deliver the squares to the Hanger bot so they're placed immediately)

XaulZan11 20-01-2011 20:00

Re: Fetcher/Strategy
 
Quote:

Originally Posted by JesseK (Post 1004269)
The Hanger and the Fetcher gather tubes for the first minute minute or so (2-3 trips each) at which point the Hanger begins hanging and the Fetcher keeps fetching.

I wonder if it would be better for the hanger just to start placing tubes right away, instead of taking 2-3 trips to get tubes. Recalling our fairly sucessful robot in 2007, it took us a fair amount of time to pick up tubes; enough that I wouldn't want to drop a tube just to pick it back up again later in the match. Plus, I would hate to spend too much time gathering tubes to run out of time to hang them all at the end. Of couse, with practice you would be able to determine when to make the switch from fetching to placing.

Its probably not going to make a huge difference but in a 2 minute match, you have to be as effecient as possible.


All times are GMT -5. The time now is 15:51.

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