Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Rules/Strategy (http://www.chiefdelphi.com/forums/forumdisplay.php?f=6)
-   -   3-3 (http://www.chiefdelphi.com/forums/showthread.php?t=66884)

Uberbots 12-04-2008 17:02

3-3
 
So in a strategy meeting last night, one of our mentors suggested that at the beginning of the match, the balls are not truly random; as the sequence of balls is always blue-red-blank, and would therefore lead to a situation where no two balls of the same color are on the outside of the field.

i dont see how his logic is working, nor i cannot see how its right/wrong. Any help?

BornaE 12-04-2008 17:07

Re: 3-3
 
That statement is not true, and here is an example to reject it
http://www.hhsrobotics.org/Videos/Vegas_Q1M1_SD.wmv

Kristian Calhoun 12-04-2008 17:14

Re: 3-3
 
That would be because of rule <G20> :
Quote:

Originally Posted by <G20>
<G20> TRACKBALL Locations - After all ROBOTS participating in the MATCH are in their starting
positions and the TEAM members are in the ALLIANCE ZONE and/or ROBOCOACH
STATIONS, four TRACKBALLS will be placed on the OVERPASS. On each side of the
OVERPASS there are three TARGET LOCATIONS for TRACKBALLS. The field
management system will randomly choose an initial starting location for the TRACKBALLS
before the start of each MATCH. One red and one blue TRACKBALL will then be positioned
in the chosen TARGET LOCATIONS on each side of the OVERPASS.
After this point in
time no ROBOT may be moved or repositioned until the MATCH starts.

Emphasis mine.

Uberbots 12-04-2008 17:18

Re: 3-3
 
Thanks (=

BornaE 12-04-2008 17:26

Re: 3-3
 
Quote:

Originally Posted by Kristian Calhoun (Post 735541)
That would be because of rule <G20> :

Emphasis mine.

this is the old version of <G20> that you are reffering to. It would be useful if you get the new one.;)
however, there is no diffrence regarding this thread's question.

waialua359 12-04-2008 17:41

Re: 3-3
 
Well, although the randomizer truly randomly sets the trackballs, here is food for thought!
Whats the chance that in a match prior to yours, the balls will be in exactly the same place?
Well, FYI, I kept track of just about every match when we were queued and watched where the balls were prior to our match. Only twice in three regionals did the ball end up in the same place. I'm talking about just our ball.

The reason we were tracking them was due to the fact that our ball knocker during hybrid could hit it if the ball was in front of us. We always tried then to put the robot in position where the ball wasn't in the previous match.
So rather than having a 1/3 chance, we were hoping for > 1/3 chance to have it in front of us.
I know you cant count on it, but what the heck, it was worth a shot, and it didnt hinder our alliance from doing what they wanted in hybrid.

Now we only tried it if our alliance partners couldnt knock the ball off and/or didnt move in hybrid.

BornaE 12-04-2008 17:52

Re: 3-3
 
Quote:

Originally Posted by waialua359 (Post 735550)
Well, although the randomizer truly randomly sets the trackballs, here is food for thought!
Whats the chance that in a match prior to yours, the balls will be in exactly the same place?
Well, FYI, I kept track of just about every match when we were queued and watched where the balls were prior to our match. Only twice in three regionals did the ball end up in the same place. I'm talking about just our ball.

The reason we were tracking them was due to the fact that our ball knocker during hybrid could hit it if the ball was in front of us. We always tried then to put the robot in position where the ball wasn't in the previous match.
So rather than having a 1/3 chance, we were hoping for > 1/3 chance to have it in front of us.
I know you cant count on it, but what the heck, it was worth a shot, and it didnt hinder our alliance from doing what they wanted in hybrid.

Now we only tried it if our alliance partners couldnt knock the ball off and/or didnt move in hybrid.

If the randomizer is truly random, the position of the ball is indepedent on eack match, but hey if it adds moral support :d go fo it.

GaryVoshol 12-04-2008 20:32

Re: 3-3
 
There are 6 possible arrangements:
RB-
BR-
R-B
B-R
-RB
-BR

So the chance that the one that comes up is the same as the previous one is 1/6.

waialua359 12-04-2008 22:07

Re: 3-3
 
Quote:

Originally Posted by GaryVoshol (Post 735625)
There are 6 possible arrangements:
RB-
BR-
R-B
B-R
-RB
-BR

So the chance that the one that comes up is the same as the previous one is 1/6.

Yes, 1/3 if your just tracking where your ball will be if you place your robot in front of one of the possible locations of the overpass.

XXShadowXX 12-04-2008 22:15

Re: 3-3
 
What i suggest you do is watch videos from several regionals and find the algorithm that generates the ball postions (if the out come is known then the forumla can be generated). This is like major cheating, and very not GP.

BornaE 12-04-2008 22:20

Re: 3-3
 
Quote:

Originally Posted by XXShadowXX (Post 735698)
What i suggest you do is watch videos from several regionals and find the algorithm that generates the ball postions (if the out come is known then the forumla can be generated). This is like major cheating, and very not GP.

I do not agree.

If the "Randomizer" is random, then no formulas can be found.
If it not actually random and somebody figures out the formula, then It is their superior scouting and pattern-finding talent, and they are welcome to use it. it is completely fair since every team could watch the matches and potentially find the pattern.

XXShadowXX 12-04-2008 22:25

Re: 3-3
 
computer can not generate random, computer have not or will never understand random, random is something that can only be generated by a intelliegnce and then it can still be predicted to a degree.

Random in a computer is a complex algorithm faking random, its really really really really complex so the answers seem random but the more data you have the closer you can build a similar algorithm, see a math genius its possible.

Quote:

If it not actually random and somebody figures out the formula, then It is their superior scouting and pattern-finding talent, and they are welcome to use it. it is completely fair since every team could watch the matches and potentially find the pattern.
Not really scouting skills just having internet and sitting on TBA for the next four days straight

Uberbots 12-04-2008 22:33

Re: 3-3
 
Quote:

Originally Posted by XXShadowXX (Post 735711)
computer can not generate random, computer have not or will never understand random, random is something that can only be generated by a intelliegnce and then it can still be predicted to a degree.

Random in a computer is a complex algorithm faking random, its really really really really complex so the answers seem random but the more data you have the closer you can build a similar algorithm, see a math genius its possible.


Not really scouting skills just having internet and sitting on TBA for the next four days straight

doesnt it combine the last couple of bits in the registers of the processor? its not completely random, but its at least (nearly) unpredictable in a dynamic system...

Vikesrock 12-04-2008 22:34

Re: 3-3
 
Quote:

Originally Posted by XXShadowXX (Post 735711)
computer can not generate random, computer have not or will never understand random, random is something that can only be generated by a intelliegnce and then it can still be predicted to a degree.

Random in a computer is a complex algorithm faking random, its really really really really complex so the answers seem random but the more data you have the closer you can build a similar algorithm, see a math genius its possible.


Not really scouting skills just having internet and sitting on TBA for the next four days straight

If the pseudo-random algorithm uses the system clock as a seed it would be extremely difficult if not impossible to discern any type of pattern from watching matches as the exact time that the ball randomizer was run is not known. This is an easy and common way to get a good seed for a pseudo-random algorithm.

EDIT: The method mentioned by Uberbots above is also used and would also render any amount of match watching useless.

XXShadowXX 12-04-2008 22:36

Re: 3-3
 
Quote:

Originally Posted by Uberbots (Post 735716)
its not completely random, but its at least (nearly) unpredictable in a dynamic system...

all dynamic systems are predicatable see choas theory...


Quote:

If the pseudo-random algorithm uses the system clock as a seed it would be extremely difficult if not impossible to discern any type of pattern from watching matches as the exact time that the ball randomizer was run is not known. This is an easy and common way to get a good seed for a pseudo-random algorithm.
more complex but it can be found, everything has answer and math hasn't falled us yet (or me)


All times are GMT -5. The time now is 04:34.

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