www.robowars.org

RoboWars Australia Forum Index -> General Chatter

Chess Ranking system
Goto page Previous  1, 2, 3, 4 ... 12, 13, 14  Next

Post new topic   Reply to topic
  Author    Thread
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

I just got my old site working again. I forgot all about it. Here's a link to the full national results.
http://www.hellfire-labs.s5.com/robot%20rankings.htm

If your robot has a green score it is above average, red for below average. The average for all the scores should be 1200, which is what new robots start at.

If that worked then this should be the excel spread sheet.
http://www.hellfire-labs.s5.com/robot%20rankings.xls
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Mon Feb 14, 2005 8:54 pm 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
Spockie-Tech
Site Admin


Joined: 31 May 2004
Posts: 3160
Location: Melbourne, Australia


 Reply with quote  

If you guys want the job, I can create a sub-section of the RoboWars site and give you FTP access for you to post your rankings tables on.. ?

Nice work on the table too. Good to see I.G. up there.. The only Vic Bot in the Top 10.. Cool

Although I expect that will change this year as our state series gets going and us Vics get some more battles under our belts..
_________________
Great minds discuss ideas. Average minds discuss events. Small minds discuss people

Post Mon Feb 14, 2005 9:06 pm 
 View user's profile Send private message Send e-mail Visit poster's website MSN Messenger
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

Well very roughly the position on the table is roughly in order of wins - looses. This does more to your score then how old the robot is. If PQ wins 4 fights at the next Qld event with no looses then it'll roughly end up in fifth place.
*edit* But of corse Ballistic has 10 more points the Hellbringer, and that counts a lot.
*edit again* And I think PQ is more likly to spray its internals around the arena then win a fight.
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Mon Feb 14, 2005 9:31 pm 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
Waddy the phoenix



Joined: 16 Jun 2004
Posts: 971
Location: sydney


 Reply with quote  

lol tineyest of beefs TSE and TSNG are the same just diff versions so wouldent that be slightly diff and also TSNG did win one or 2 matches in the previous marryong Razz umm won one against crap which bot did it beat mirraculassly ?? cant rember now ummm i think it beat well it was either plan b or basilysk cause their the only 2 robots its gone up against in the last 2 marryong Razz
*edit* we need to work out a system that intergrades with this one to work out the melees maby takeing the average score of all the bots in the melee as delta ie

((bot a: 2400 bot b:1100 bot c: 10000 bot d:100 etc...) <------ average of theese scores worked out=x amount)
x amount / amout of bots in melee = delta

nfi if that would work cause i aint no mathamatition or literature expert for that matter as you can see but it looks good to me care to test this out?
_________________
as a shadow is seen and not heard i am here
-Waddy the phoenix

Post Mon Feb 14, 2005 10:09 pm 
 View user's profile Send private message MSN Messenger
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

quote:
Originally posted by Waddy the phoenix:
lol tineyest of beefs TSE and TSNG are the same just diff versions so wouldent that be slightly diff and also TSNG did win one or 2 matches in the previous marryong Razz umm won one against crap which bot did it beat mirraculassly ?? cant rember now ummm i think it beat well it was either plan b or basilysk cause their the only 2 robots its gone up against in the last 2 marryong Razz
*edit* we need to work out a system that intergrades with this one to work out the melees maby takeing the average score of all the bots in the melee as delta ie

((bot a: 2400 bot b:1100 bot c: 10000 bot d:100 etc...) <------ average of theese scores worked out=x amount)
x amount / amout of bots in melee = delta

nfi if that would work cause i aint no mathamatition or literature expert for that matter as you can see but it looks good to me care to test this out?



I just used the results that Chris and Glen gave me. If I missed something out or got it wrong, blame them.

Also The Sub was a 2 wheeled wedge and so was TSE, so that sounds like the same robot, so I combined them. TSNG was a 4 wheeled box, which sounds like a completly different robot, so I didn't combine them.
Robots I combined:
Evil Spike and Evil Spike 2
IG and IG2
The different Basalisks
Hellbringer 2 and X
Singularity (did it change the number of wheels?)
Minotaur

Robots I didn't combine, but I wasn't sure about:
Fragmentor and Ball Peen
Mini Mower and Reboot
Mouse Wedge and SSR

Also I'll think about the melee average thing today.
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Tue Feb 15, 2005 6:39 am 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

quote:
Originally posted by Waddy the phoenix:
we need to work out a system that intergrades with this one to work out the melees maby takeing the average score of all the bots in the melee as delta ie

((bot a: 2400 bot b:1100 bot c: 10000 bot d:100 etc...) <------ average of theese scores worked out=x amount)
x amount / amout of bots in melee = delta

nfi if that would work cause i aint no mathamatition or literature expert for that matter as you can see but it looks good to me care to test this out?


Nope that won't work. If you win a melee you'll get over 1000 points. Considering Sorry is less then 400 points ahead of the worst robot it won't be a very fair system.

Actually I've been playing around with this method for dealing with melee's:

-Find the average points for the loosing robots (and for the winners if there were more then one such as in an annihilator fight).

-Instead of using 27 for K as in the one-on-one fights use
K = (number of robots in melee x 2) + 23
That way winning is worth more if there are more robots in the fight.

-So delta = (average of the loosers * K)/average of the winners

-The winners new score = the old score + (delta/the number of winners)

-The loosers new score = the old score - (delta/the number of loosers)

This way if there are a lot of robots in the fight and you loose, you loose less points from your score then if there were less robots in the fight. This makes sence because you have less chance of winning a 10 way melee over a 3 way melee. So you should be penalised more for loosing a 3 way. But you'll still loose less point then if you lost a 1-on1 fight. Also if you win a 10 way fight you get (roughly) 43 points, a 3 way would give you (roughly) 29 points and a 1-on-1 gives you (roughly) 27 points.

I tested it out on Metal Mayhem 1 which had 3 melees.
Melee 1 (Haemorrhage, Annihilation, Hellbringer and Spiney Tiger)
Haemorrhage - 1231
Annihilation - 1189.66667
Hellbringer - 1189.66667
Spiney Tiger - 1189.66667

Melee 2 (Annihilation, Hellbringer, Spiney Tiger, C4 and Frogg Stomp)
Annihilation - 1222.80999
C4 - 1191.71417
Frogg Stomp - 1191.71417
Hellbringer - 1181.38084
Spiney Tiger - 1181.38084

Melee 3 (Hellbringer, Annihilation, Spiney Tiger)
Hellbringer - 1210.88933
Annihilation - 1208.05575
Spiney Tiger - 1166.62660

See Annihilation got penalised more for loosing the 3 way then Hellbringer did by loosing the 5-way. But it still won more points from the 5-way the Hellbringer did.

Any comments would be helpful.
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Tue Feb 15, 2005 1:14 pm 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
Philip
Experienced Roboteer


Joined: 18 Jun 2004
Posts: 3842
Location: Queensland near Brisbane


 Reply with quote  

The main problem I see with melees is the winner may not have been the robot to beat the loosers. There may be two powerful robots that go hard at eachother and half wreck eachother while another, through no fault of it's own, is ignored until the less damaged powerful bot limps over at the end.

On the other hand, the points don't worry me so much. I just like to play with the robots. I enloyed loosing 0/2 as much as I enjoyed winning an event.
_________________
So even the rain that falls isn't actually going to fill our dams and our river systems

Post Tue Feb 15, 2005 1:41 pm 
 View user's profile Send private message
chrisjon65
Experienced Roboteer


Joined: 15 Jun 2004
Posts: 754
Location: blaxland


 Reply with quote  

well its pretty cool to see SORRY at number 1 Wink mitch my son was over the moon he loved it ,and now i have the added pressure from him to remain there ohh goody Wink
as for ranking melees well its tough ,we dont even place judging criteria into ours really ...because of the 7 or so we have had ,its been an odvious choice in most cases ,and thats the one still moving in the end damage has been known to wide spread and in many cases parts untracable till weeks later Wink we have had a couple of draws and thats due to controversy easily solved that way giving the 2 bots left 1 point each
as much as we do score them they are more fun than anything ,and a way to get back at a few blokes after the rounds proper ...not to mention its a favourite of jakes and nicks and who are we to let them down and not cast our bots into the shredder 'literally' Wink mmmmmm now i think about it ,with such little time on my hands this season i might hide in a safe corner this year ....so in conclusion i would suggest scoring melles seperately and build a new ranking based on them alone . just my thoughts Wink
_________________
Photos - http://community.webshots.com/album/154092733uokpXC
Photos- http://community.webshots.com/album/166819552PDWWqP

Post Tue Feb 15, 2005 7:42 pm 
 View user's profile Send private message Visit poster's website
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

Ok, I'll leave playing with melee's till the week end when I'll crunch some numbers.

In the mean time look at this equation:
Percentage chance of winning = 1/(1 + 20^(-(robot 1 - robot 2)/400))

eg, Annihilation has a 22.54% chance of beathing El Bravo
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Wed Feb 16, 2005 4:34 pm 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
Philip
Experienced Roboteer


Joined: 18 Jun 2004
Posts: 3842
Location: Queensland near Brisbane


 Reply with quote  

Contrary to my above post, I now think that you deserve recognition for winning a melee. Just being able to have your robot moving at the end of an event is an achievement.

The points being determined by the number of combatants sounds like a good system.
_________________
So even the rain that falls isn't actually going to fill our dams and our river systems

Post Fri Feb 18, 2005 8:20 pm 
 View user's profile Send private message
Waddy the phoenix



Joined: 16 Jun 2004
Posts: 971
Location: sydney


 Reply with quote  

the numbers look good
Very Happy
as for TSNG it "did" have a wedge then got flattned out to a scoop in a match as for the drive enhancements it would be the same or on similar terms be like upgrading from xu1 's to i dont know scooter motors the 4wd was what i origionally was going to do but ran out of time and instead upgraded the chassy so realy its just the upgraded version and i know its probably skiing fery close to being a new robot but when you watch its evolution from TS being a massive wedge to TSE being a wedge brick to TSNG being a brick with a wedge added you can see that its sections are overlaping Laughing all of its vital parts have remained the same (control ibc and reciever drive drill motors wheels redwheels then upgraded to jakes wheels) but if every one sees it as a new bot its fine by me
_________________
as a shadow is seen and not heard i am here
-Waddy the phoenix

Post Sun Feb 20, 2005 8:36 pm 
 View user's profile Send private message MSN Messenger
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

Actually the only time I've seen your robots have been in person. Without pictures or video's from your events I have to guess for this sort thing.
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Mon Feb 21, 2005 12:52 pm 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
Waddy the phoenix



Joined: 16 Jun 2004
Posts: 971
Location: sydney


 Reply with quote  

hmmm what about match vidios or did you like every one else skip downloading my fights because i was in it Very Happy
_________________
as a shadow is seen and not heard i am here
-Waddy the phoenix

Post Mon Feb 21, 2005 7:51 pm 
 View user's profile Send private message MSN Messenger
kkeerroo
Experienced Roboteer


Joined: 17 Jun 2004
Posts: 1459
Location: Brisbane


 Reply with quote  

Who here understands this sort of thing?


quote:
The code is all here:

http://botrank.com/~botrank/go
http://botrank.com/~botrank/z1
http://botrank.com/~botrank/z2
http://botrank.com/~botrank/z3
http://botrank.com/~botrank/z4
http://botrank.com/~botrank/z5

The source data file is here:

http://botrank.com/~botrank/data/data

The calculations are in http://botrank.com/~botrank/z2

The calculations are bgmoon's from the chess ranking for bots discussion.

but it boils down to:

Winner:
D=K * (1 - (1 / (1 + (10 ^ ((BR2-BR1) / ER)))))
Loser:
D=K * (0 - (1 / (1 + (10 ^ ((BR1-BR2) / ER)))))

D is Delta change in ranking score
K is defined in http://botrank.com/~botrank/z2 as 35
ER is defined in http://botrank.com/~botrank/z2 as 600
BR1 is winner's current ranking score before this fight
BR2 is loser's current ranking score before this fight





Its how the botrank.com site does all its calculations. Unfortuantly I am hopeless at anything like this. Is a certain program needed to run it or does the internet do it or what?

Also they are starting to add the Battlebot results the historical rankings. Biohazzard is only 4th at this point. Crying or Very sad
_________________
Get Some!!!

Secretary of the Queensland Robotics Sports Club inc.

Post Tue Feb 22, 2005 12:53 pm 
 View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger
Spockie-Tech
Site Admin


Joined: 31 May 2004
Posts: 3160
Location: Melbourne, Australia


 Reply with quote  

These are Unix shell scripts.. The unix equivalent of autoexec.bat batch files

You can run them on a home Linux system or I can set them up on the RoboWars server if you guys want to feed them data ?
_________________
Great minds discuss ideas. Average minds discuss events. Small minds discuss people

Post Tue Feb 22, 2005 3:19 pm 
 View user's profile Send private message Send e-mail Visit poster's website MSN Messenger
  Display posts from previous:      

Forum Jump:
Jump to:  

Post new topic   Reply to topic
Page 3 of 14

Goto page Previous  1, 2, 3, 4 ... 12, 13, 14  Next

Forum Rules:
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

 

Last Thread | Next Thread  >
Powered by phpBB: © 2001 phpBB Group
millenniumFalcon Template By Vereor.