
January Ghost Ratings Published
Forum rules
This forum is limited to topics relating to the game Diplomacy only. Other posts or topics will be relocated to the correct forum category or deleted. Please be respectful and follow our normal site rules at http://www.webdiplomacy.net/rules.php.
This forum is limited to topics relating to the game Diplomacy only. Other posts or topics will be relocated to the correct forum category or deleted. Please be respectful and follow our normal site rules at http://www.webdiplomacy.net/rules.php.
Re: January Ghost Ratings Published
Durgastly fact is high GR's are timid players who in their incestial manner play only with themselves (among themselves for mods with dirty minds). they fear to play against others of lesser games or ratings worried that someone new may get into their party. as for me i fear no one and stand ready to give anyone a sever beating in any game at anytime. 

-
- Posts: 79
- Joined: Fri Sep 29, 2017 5:09 pm
- Contact:
Re: January Ghost Ratings Published
Brad, you are one of the worst players on the site. Top-tier players prefer to play with each other for the same reason that top-tier chess players do. It's more interesting.
Re: January Ghost Ratings Published
If you look at my games list you'll know I'm open to playing with everyone reliable at least once.
-
- Posts: 4305
- Joined: Fri Sep 29, 2017 4:16 pm
- Location: The Five Valleys, Gloucestershire
- Contact:
Re: January Ghost Ratings Published
I disagree. By far the most entertaining games involve people with a mix of abilities and experience. As long as they're willing to talk, that is.Condescension wrote: ↑Sat Jan 06, 2018 7:43 pmBrad, you are one of the worst players on the site. Top-tier players prefer to play with each other for the same reason that top-tier chess players do. It's more interesting.
-
- Lifetime Site Contributor
- Posts: 1099
- Joined: Fri Sep 29, 2017 4:20 pm
- Contact:
Re: January Ghost Ratings Published
Perhaps nobody likes to play with you because you're just all around unpleasant to everyone on the forum?CAPT Brad wrote: ↑Sat Jan 06, 2018 7:12 pmDurgastly fact is high GR's are timid players who in their incestial manner play only with themselves (among themselves for mods with dirty minds). they fear to play against others of lesser games or ratings worried that someone new may get into their party. as for me i fear no one and stand ready to give anyone a sever beating in any game at anytime.![]()
-
- Posts: 79
- Joined: Fri Sep 29, 2017 5:09 pm
- Contact:
Re: January Ghost Ratings Published
+1. I've played a few anon games where I later found out Brad was in it... and he was, every time, the most unpleasant, whiny, irrational player.jmo1121109 wrote: ↑Sat Jan 06, 2018 8:24 pmPerhaps nobody likes to play with you because you're just all around unpleasant to everyone on the forum?
Terrible players are basically random number generators who can't even act in their own self-interest and will frequently throw games or CD.
Re: January Ghost Ratings Published
I don't throw games. I don't CD. I provoke fear in those I play. As to you condensation, your description of me more matches you.
-
- Posts: 79
- Joined: Fri Sep 29, 2017 5:09 pm
- Contact:
Re: January Ghost Ratings Published
It's pretty embarrassing that you have trench GR when you don't CD. The only way other players can be as shite as you is by literally trying to lose.
Re: January Ghost Ratings Published
Condensation you fear to play me. Brain. I am kicking your butt and you know it.
Re: January Ghost Ratings Published
Another question: do Ghost Ratings include games that I have replaced a player in (and then proceeded to lose, because the player I replaced sucked)?
Re: January Ghost Ratings Published
Yes, they do.
A better solution would be nice, but would sadly require a good amount of development work on both the site and the GR scripts.
A better solution would be nice, but would sadly require a good amount of development work on both the site and the GR scripts.
Re: January Ghost Ratings Published
Ah well, that kind of sucks. I try to take over a lot of games when people CD because I hate it when it happens in my games. Do you have any idea what would need to be added to support that?
Re: January Ghost Ratings Published
The first step would be deciding how to handle it in the scoring. GR doesn't make a lot of mathematical sense to begin with, so that could probably be fudged a little.
Then, the site tracks CDs but not replacements. I'm not sure how exactly it does it, but there'd have to be some amount of effort put into extracting the latter from the former and then including it in the GR data dump script.
Then you'd need to actually edit GR, which is a hideous mess of poorly-documented and needlessly complex code.
Then, the site tracks CDs but not replacements. I'm not sure how exactly it does it, but there'd have to be some amount of effort put into extracting the latter from the former and then including it in the GR data dump script.
Then you'd need to actually edit GR, which is a hideous mess of poorly-documented and needlessly complex code.
Re: January Ghost Ratings Published
I believe that vRanking at vDiplomacy handles it the following way:
A CD position is 'worth' a certain percentage of the pot. This is the amount of
someone has to pay to take over a CD position and it is equal to some number just larger than 1/2 times the percentage of the supply centers of the board the CD'd layer controls. It is equal to 0 on this site, but it has been the same as on vDiplomacy, so it is possible to change in the code.
As someone familiar with the concept of rankings knows, GR and vRanking use expected points won in their calculations. For example, if you are a good player having bet 10
in a game, it may be that according to your ranking, the expected value of
won that game is 20. If you then end up getting more/less than 20
, your ranking increases/decreases.
I have the suspicion that vRanking simply multiplies everyone's expected points with what they have bet with respect to the other players. For example, if you are the same good player from the previous example and take over a CD worth 3
where the other players bet 10
, then you only need to win 6
for your ranking to remain unchanged.
This system is not perfect, but is better than the current one.
And yes, GR does not make much mathematical sense.
A CD position is 'worth' a certain percentage of the pot. This is the amount of

As someone familiar with the concept of rankings knows, GR and vRanking use expected points won in their calculations. For example, if you are a good player having bet 10



I have the suspicion that vRanking simply multiplies everyone's expected points with what they have bet with respect to the other players. For example, if you are the same good player from the previous example and take over a CD worth 3



This system is not perfect, but is better than the current one.
And yes, GR does not make much mathematical sense.
Re: January Ghost Ratings Published
The way to do the same in GR is by changing your expected part of the pot from (your own GR) / (Sum of the GRs of all the players) into (your own GR times your bet) / (Sum of (the GR times the bet) of all the players).
Re: January Ghost Ratings Published
Does the data dump that GR uses include data on how much each player bet?
Re: January Ghost Ratings Published
It looks like vDip's code for calculating ratings is here: https://github.com/Sleepcap/vDiplomacy/ ... rating.php . Is the code that is used for Ghost Ratings online somewhere?
Who is online
Users browsing this forum: No registered users