Go to Post You can never be done too early for more testing and practice! - Donut [more]
Home
Go Back   Chief Delphi > Competition > Rules/Strategy
CD-Media   CD-Spy  
portal register members calendar search Today's Posts Mark Forums Read FAQ rules

 
Closed Thread
Thread Tools Rate Thread Display Modes
  #16   Spotlight this post!  
Unread 18-01-2009, 21:00
Herodotus's Avatar
Herodotus Herodotus is offline
Mountain Dew Bandolier Man
AKA: David Resowski
FRC #0910 (Foley Freeze)
Team Role: Alumni
 
Join Date: Apr 2007
Rookie Year: 2006
Location: Michigan
Posts: 428
Herodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond reputeHerodotus has a reputation beyond repute
Re: A statistical look at G14

I think our team's plan is to make G14 a non-issue by just attempting to dominate every match, even the ones we have no super cells in.Hopefully we'll never have a supercell available to us for just that reason.

Doesn't change the fact that it's a horrible rule though.
__________________
I am always doing that which I cannot do, in order that I may learn how to do it. - Pablo Picasso
--My Life in FIRST--
2009 Detroit Xerox Creativity Award
2009 Detroit District Finalists - Thanks to 1856 and 2620
2009 Kettering District Quarter-Finalists - Thanks to 1504 and 1025.
2008 Kettering Kickoff Champions - Thanks to 67, 1075 & 2619
2008 MARC Finalists - thanks 67 & 226
2008 Great Lakes Finalists - thanks 66 & 217
2008 Western Michigan Semi-Finalists - thanks 2337 and 1504
2008 GLR Judges Award Winner
2007 Curie Division Champions - thanks 330 and 1270
2007 ARC Champions (13 and 0, plus scoring a double-keeper!) - thanks 1625 and 313
2007 MARC Champions - thanks 1732 and 1023
2007 Xerox Creativity Award - Western Michign
2006 I.R.I. Champions - thanks 71 and 1625
  #17   Spotlight this post!  
Unread 18-01-2009, 21:39
RobotDevil1985's Avatar
RobotDevil1985 RobotDevil1985 is offline
Registered User
FRC #0533 (PsiCotics)
Team Role: Mentor
 
Join Date: Jan 2009
Rookie Year: 1999
Location: Lindenhurst
Posts: 6
RobotDevil1985 is an unknown quantity at this point
Re: A statistical look at G14

Quote:
Originally Posted by Lil' Lavery View Post
This data isn't the proof, the proof is the fact that G14 penalizes teams who have done nothing wrong. Why should my team be penalized for the actions of my alliance partner the match before?


This shows the upper bound of the scale of this problem.

<G14> is, by far, the single worst rule written since I've been involved with FIRST. There is no justification for this abomination of a rule.
If you are the team which keeps winning by a landslide then yes you are going to feel like you are being penalized for no reason, but for those teams who were not capable, either financially or technically, to design a robot which is perfect at all the tasks, this is a godsend.

There have been plenty of years where walking through the pits all you hear is "We're matched up against (insert amazing performing team here) 3 times today, we can't win against them." This rule may even the odds a little bit.

It's similar to when your score was based on a multiple of the losing score, it brings the score a little closer together without implementing a mercy rule.
  #18   Spotlight this post!  
Unread 18-01-2009, 23:22
Lil' Lavery Lil' Lavery is offline
TSIMFD
AKA: Sean Lavery
FRC #1712 (DAWGMA)
Team Role: Mentor
 
Join Date: Nov 2003
Rookie Year: 2003
Location: Philadelphia, PA
Posts: 6,634
Lil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond reputeLil' Lavery has a reputation beyond repute
Send a message via AIM to Lil' Lavery
Re: A statistical look at G14

Quote:
Originally Posted by RobotDevil1985 View Post
If you are the team which keeps winning by a landslide then yes you are going to feel like you are being penalized for no reason, but for those teams who were not capable, either financially or technically, to design a robot which is perfect at all the tasks, this is a godsend.

There have been plenty of years where walking through the pits all you hear is "We're matched up against (insert amazing performing team here) 3 times today, we can't win against them." This rule may even the odds a little bit.

It's similar to when your score was based on a multiple of the losing score, it brings the score a little closer together without implementing a mercy rule.
No, it's just the opposite. Now when I face off against these elite teams, there's a chance that my alliance partners will put me at a disadvantage because they did well their last match.

My problem with G14 is not what I do, but what my alliance partners do. I can control my team, but I can't control what my randomly paired partners did the match before.
__________________
Being correct doesn't mean you don't have to explain yourself.
  #19   Spotlight this post!  
Unread 18-01-2009, 23:30
pacoliketaco's Avatar
pacoliketaco pacoliketaco is offline
Alumni
FRC #1807 (Redbird Robotics)
Team Role: Alumni
 
Join Date: May 2005
Rookie Year: 2005
Location: Pittsburgh, PA
Posts: 119
pacoliketaco is a splendid one to beholdpacoliketaco is a splendid one to beholdpacoliketaco is a splendid one to beholdpacoliketaco is a splendid one to beholdpacoliketaco is a splendid one to beholdpacoliketaco is a splendid one to beholdpacoliketaco is a splendid one to behold
Send a message via AIM to pacoliketaco Send a message via Yahoo to pacoliketaco
Re: A statistical look at G14

Quote:
Originally Posted by Lil' Lavery View Post
We can't afford to go with "let's look at Week 1" method, because doing so is essentially screwing over week 1. We need to fix the problem BEFORE the competitions begin, rather than tell all teams competing in week 1 that they are just the test subjects.
lol. as a member of the week 1 regionals for the fourth year this year, i have to comment that the same will probably be true. it always seems like the NJ regional runs a lot less smoothly than NY (last week in previous years). im not sure how this could be remedied, as FIRST wont have 50+ teams ready to play before week 1 to try out the game.

but in general, i think the idea for G14 is a good one, but it will certainly not be as much fun as trying to get the highest possible score.

another bad thing about week 1 this year is the open "fix-it" windows, which seems to allow all teams unlimited time to work on parts for our robots, at least that is how i read it.
__________________
Pace Nalbone
Carnegie Mellon University
Mechanical Engineering Class of 2013

2009 NYC Winner - 56 and 1796
(2009 Regional Record: 18-0-7 wins/ties/losses)
  #20   Spotlight this post!  
Unread 19-01-2009, 07:25
Bongle's Avatar
Bongle Bongle is offline
Registered User
FRC #2702 (REBotics)
Team Role: Mentor
 
Join Date: Feb 2004
Rookie Year: 2002
Location: Waterloo
Posts: 1,069
Bongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond repute
Send a message via MSN to Bongle
Re: A statistical look at G14

Quote:
Originally Posted by thefro526 View Post
point values. I don't know what game would be the best comparison because I'm not familiar with the scoring prior to 2006.

If the Statistics for Overdrive were true though and I'm Reading the Graph correctly then my understanding is that 75% of matches would have at least one ball missing?
A more accurate statement to draw from this (since it is a worst-case scenario) is that at least 25% of lunacy matches are likely to have their full super cell/empty cell allotment. Granted, the first couple matches will have their full set anyway because nobody has drawn a G14 yet.
  #21   Spotlight this post!  
Unread 19-01-2009, 08:45
JesseK's Avatar
JesseK JesseK is online now
Expert Flybot Crasher
FRC #1885 (ILITE)
Team Role: Mentor
 
Join Date: Mar 2007
Rookie Year: 2005
Location: Reston, VA
Posts: 3,694
JesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond repute
Re: A statistical look at G14

In order to truly align the statistical data with what we should expect from this year's game, we need to figure out which game had similar point values for the game pieces and for the endgame. 2008's scoring happened in one spot for each alliance (except for hybrid mode), so I don't believe 2008's matches provide a valid anaylsis for 2009's midgame. 2008's endgame only provided a maximum of 24 points and also wasn't even an option if the alliance didn't already have great midgame execution potential.

In all honesty, 2007 has the correct structure. Even though the game pieces could be worth more individually, there was in fact a sort of scarcity of them relative to this year. This scarcity combined with the 'factor of two' multiplication will provide a more direct correlation to what we can expect in 2009 midgame results. Also, the endgame point values are identical -- in the last 20 seconds, teams have the opportunity to score 60 points without having to rely on mistakes made by their opponents.

Then, do a statistical analysis on teams who would have lost a match had their endgame point values been slashed in half (alot of research, yes). I know for a fact that the '07 VCU finals and '07 Einstein finals would have had a different outcome. Overall, I'm pretty sure that what we'd find is that alliances who put more emphasis on midgame active robot scoring in teleop would have prevailed over alliances who went for defense/endgame strategies.
__________________

Drive Coach, 1885 (2007-present)
CAD Library Updated 5/1/16 - 2016 Curie/Carver Industrial Design Winner
GitHub
  #22   Spotlight this post!  
Unread 19-01-2009, 11:42
Bongle's Avatar
Bongle Bongle is offline
Registered User
FRC #2702 (REBotics)
Team Role: Mentor
 
Join Date: Feb 2004
Rookie Year: 2002
Location: Waterloo
Posts: 1,069
Bongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond reputeBongle has a reputation beyond repute
Send a message via MSN to Bongle
Re: A statistical look at G14

Quote:
Originally Posted by JesseK View Post
In all honesty, 2007 has the correct structure. Even though the game pieces could be worth more individually, there was in fact a sort of scarcity of them relative to this year.
I disagree. 2007's exponential scoring makes the 2x/3x thresholds much easier to hit. You could have one team score 6 ringers and another team score 5 ringers, and have the winning team get double the score. Scoring just 2 ringers less than your opponent's team could result in you having 1/4 of their score. 2005 and 2007 both had strategic scoring: where you put something (and where you had put things earlier) had a great deal to do with how much it was worth. 2006, 2008, and 2009 all have what I'll call 'speed' scoring: you have to do similar things as fast as you can, regardless of location. The built-in lopsidedness of 2007 scoring would make it a poor candidate as a stand-in.

2008 or 2006 would be better choices than 2007 as proxies because your score was approximately proportional to what you did in the game and how many times you did it, much like Lunacy. Lunacy will have less lopsided scoring than either because even if your alliance's robots are all horrible and can't score, your human players can still rack up some points.
  #23   Spotlight this post!  
Unread 19-01-2009, 13:22
gorillamonky gorillamonky is offline
Registered User
FRC #0696
 
Join Date: Jan 2009
Location: la canada
Posts: 33
gorillamonky is an unknown quantity at this point
Re: A statistical look at G14

regardless of the statistics using previous games to determine the quality of a rule, the rule isn't going to change. however, if we use practice games to gain the data, it would at least give us a reasonably estimate of how often the rule will come into effect
  #24   Spotlight this post!  
Unread 19-01-2009, 14:08
JesseK's Avatar
JesseK JesseK is online now
Expert Flybot Crasher
FRC #1885 (ILITE)
Team Role: Mentor
 
Join Date: Mar 2007
Rookie Year: 2005
Location: Reston, VA
Posts: 3,694
JesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond repute
Re: A statistical look at G14

Quote:
Originally Posted by Bongle View Post
2008 or 2006 would be better choices than 2007 as proxies because your score was approximately proportional to what you did in the game and how many times you did it, much like Lunacy. Lunacy will have less lopsided scoring than either because even if your alliance's robots are all horrible and can't score, your human players can still rack up some points.
This is a good point to an extent. I agree that 2006 would be better for a midgame score sampling, though my statements also took into account the endgame and the big picture... which takes additional research and time to get right, so maybe it's inapplicable.

2008's endgame was the deciding factor in some matches, but didn't have as much impact as 2007 or 2009. 2006's endgame was somewhere in between, though many teams incorporated it into the way they scored to begin with...so it then was no longer an isolated 'endgame' strategy like 2007 was or 2009 will be, which rules it out. None of the three years had direct and, consistent human player scoring so it's difficult to gauge where that plays into statistics.

While I agree that 2007's exponential scoring and where you placed ringers had more effects than quantity of ringers, I could still argue that 2007's scoring is more directly related to 2009's than any other because 2009 has similar potential for spikes in scoring. For dumper bots, the scores spike relative to the maximum score just as much as placing ringers in a row did in a typical match -- remember, in 2007 rows of 6 or more were fairly rare. However, my point was more to the fact that 2007's game strategies were more inlined with 2009's spectrum of strategies: on one end you are a bot who can score the game piece magnificently, and on the other you are a defensive bot who's primary strategy is to hold out for the endgame. <G14> makes the latter less attractive as a primary strategy for a game like Lunacy or Rack 'N Roll, and I believe we would immediately see that if <G14>'s implications were put into effect for 2007 rather than 2008/2006. Though I don't want to do that much work, and neither does anyone else I don't think.
__________________

Drive Coach, 1885 (2007-present)
CAD Library Updated 5/1/16 - 2016 Curie/Carver Industrial Design Winner
GitHub
Closed Thread


Thread Tools
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
<G14> Shenanigans? Team1710 Rules/Strategy 123 12-01-2009 12:42
My case against <G14> bduddy General Forum 58 07-01-2009 15:20
Rule G14 KE5WGE Technical Discussion 3 03-01-2009 17:36
Statistical Analysis of Regional Competion Scores rourke Regional Competitions 9 08-04-2004 01:05
Statistical ? aman Regional Competitions 0 09-03-2003 11:23


All times are GMT -5. The time now is 16:29.

The Chief Delphi Forums are sponsored by Innovation First International, Inc.


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