Clan System Request for Proposals

If everyone on this site just played to try and win every game they play, we wouldn't even need this thread. No engines, no sandbagging, no collusion, just an honest game of chess, Some people need to take a long hard look at themselves, and then we can make 2017 a good year on RHP.

Agreed Steve.I think most of us do try to win every game.I still try to win remaining games for instance in a tournament where I am already out.
As in most things ,like football hooligans for example , it's the minority who ruin it for everyone else

collusion is a symptom - a correction to a situation deemed unfair

the folks responsible did this deliberately ...with full knowledge that they might be penalized and or kicked off the site


in the context of this forum

some suggestions are designed facilitate fair play (reducing the need for the perceived corrective action of collusion)

some suggestions are designed to address and penalize collusion directly


we're here to review all of them

we suggest the one's that we think have the potential to improve the site to Russ

we discover from his evaluation, the programmable options

Russ is paying attention, and he's being kind enough to let us vote

we cannot disappoint him


so ...is this idea useful?

Originally posted by Giannotti
luv you but ...your response not in keeping with the questions

so i'll ask again, could you manipulate it (if you were so inclined), and is it scalable

this is a new idea, please open your mind and consider the possibilities

I think manipulation can maybe come in when the starting rating is not the true rating....

settl 930 v suzzianne 1000
in that example settl won... but what if settl was once a 1800.... that is why he won.
so in this example if i understand it correctly...
settl will move up slowly... and extremely slowly back towards his true rating of 1800... with many wins on the way there.

romas collusion points comes in as well... say settl wants to stay down at 930... will lose many if he does not care for his clan...

hence my hope to bring all time high rating back into the mix...
settl with an all time high of 1800 should never be allowed to play suzzianne whose all time high is only 1055...

I just want to add also. All the reasonable points put forward by others are still good points.
lets not all get bickering rather than seeking the solution...

Originally posted by Giannotti
collusion is a symptom - a correction to a situation deemed unfair

the folks responsible did this deliberately ...with full knowledge that they might be penalized and or kicked off the site


in the context of this forum

some suggestions are designed facilitate fair play (reducing the need for the perceived corrective action of collusion)

some ...[text shortened]... e's being kind enough to let us vote

we cannot disappoint him


so ...is this idea useful?

totally disagree collusion is no symptom its the cause of all the trouble

only a points removal and a warning to every clan on here will stop it

benefits
1] put integrity back to RHP
2] act as a warning to all clans on here
3] stop players leaving this site

there was no protest, a very poor 11th hour excuse, certain players set out to ruin the clan system and in the process ruined many tournaments they entered to resign in order to sand bag a real protest would be to stop subs that is what is happening now,

if this was a new problem fair enough but the same happened last year, clans suspended but the same player do it again on a much larger scale but nothing happens for a year

to me there is no way forward until those involved in the collusion are punished, no ban required just remove collusion points.

...so maybe the answer is a more dynamic equation, including the five year high, as well as the yearly high, and the current average

...very good point

Originally posted by Costad
I think manipulation can maybe come in when the starting rating is not the true rating....

settl 930 v suzzianne 1000
in that example settl won... but what if settl was once a 1800.... that is why he won.
so in this example if i understand it correctly...
settl will move up slowly... and extremely slowly back towards his true rating of 1800... with man ...[text shortened]... by others are still good points.
lets not all get bickering rather than seeking the solution...

Yes this is a great point. Some mechanism would need to be used to make sure that players do not enter a clan with a vastly reduced rating to take advantage of the system. Perhaps the tournament entry rating could be used to prevent the player from playing anyone below that? In the case of players with a provisional rating they would need to complete their six games before their clan games counted.

Originally posted by Giannotti
...so maybe the answer is a more dynamic equation, including the five year high, as well as the yearly high, and the current average

...very good point

Actual site formula utilised to project ratings. Thanks to moonbus.

Hypothetical scenario.

Clan A v Clan B

1) robbie 1798 v mghrn55 1810: result 1/2 - 1/2
2) tomeasyrider 1600 v hetrz van rental 1610: result 0-1
3) settl 930 v suzzianne 1000: result 1-0

The following formulas were copy-pasted from the RHP FAQ:

New Rating = Old Rating + K * (Score - Win Expectancy).
K is a constant (32 for 0-2099, 24 for 2100-2399, 16 for 2400 and above).

Score is 1 for a win, 0.5 for a draw and 0 for a loss.
The Win Expectancy is calculated using the following formula:
Win Expectancy = 1 / (10^((OpponentRating-YourRating)/400)+1)
Note: ^ = "to the power of", e.g. 2^3=8.

The following calculations were executed according the above formulas and the hypothetical challenge results above:

1) new rating robbie = 1798 + 32 * (0.5 - X).
where X = 1 / (10^((1810-1798)/400)+1).
solving for X first = 0.48273747.
now solving the new rating equation = 1798 + 32 * (0.5 - 0.4827) = 1799.

2) new rating tomeasyrider = 1600 + 32 * (0.0 - X).
where X = 1 / (10^((1610-1600)/400)+1).
solving first for X = 0.4856.
now solving the new rating equation for tomeasyrider = 1600 + 32 * (0.0 - 0.4856) = 1584.

3) new rating settl = 930 + 32 * (1.0 - X).
Where X = 1 / (10^((1000-930)/400)+1).
solving first for X = 0.400.
now solving the new rating equation for settl = 930 + 32 * (1.0 - 0.400) = 949.

Based on these individual ratings changes, the collective net rating change for this challenge is calculated as follows: +1 for game one, -16 for game two, + 19 for game three = +4 points for Clan A. (Do the math for Clan B.)

Hope this helps.

Originally posted by robbie carrobie
Yes this is a great point. Some mechanism would need to be used to make sure that players do not enter a clan with a vastly reduced rating to take advantage of the system. Perhaps the tournament entry rating could be used to prevent the player from playing anyone below that? In the case of players with a provisional rating they would need to complete their six games before their clan games counted.

Yes, something does need to be done.
There's a chap near the top of the clan ladder who has a all time high of 1647. Tournament entry rating of 1306 and actual rating of 1149.
Manipulated by playing lowly ranked non subscribers and resigning after a few moves. Then playing clan games against players of the same (1149) ranking who are of course inferior players.
I think the tournament entry rating is a much better guide and should be used instead.

Originally posted by The Postman
Yes, something does need to be done.
There's a chap near the top of the clan ladder who has a all time high of 1647. Tournament entry rating of 1306 and actual rating of 1149.
Manipulated by playing lowly ranked non subscribers and resigning after a few moves. Then playing clan games against players of the same (1149) ranking who are of course inferior ...[text shortened]... players.
I think the tournament entry rating is a much better guide and should be used instead.

Yes the situation really calls for reform.

realize that the changes we make may have a profound impact


one of my favorite players is DepecheMode, he plays for The Quantum Mechanics

1300 level player, sure... but his average opponent rating is 1740

this guy doesn't care about his rating, he's trying to improve by playing his betters

these challenges happen outside the clan system, there's no thought of foul play

as a clan leader, i avoid him like the plague ... as a player i admire him


as we get into this, please keep this in mind

we have to separate clan ratings from individual ratings


it has to happen so guys like DepecheMode can do their thing and still play for a clan


(for my part, i do this too, but my invites are unrated)

(i wonder if i'd play better if they weren't)

Originally posted by Giannotti
realize that the changes we make may have a profound impact


one of my favorite players is DepecheMode, he plays for The Quantum Mechanics

1300 level player, sure... but his average opponent rating is 1740

this guy doesn't care about his rating, he's trying to improve by playing his betters

these challenges happen outside the clan system, ther ...[text shortened]... art, i do this too, but my invites are unrated)

(i wonder if i'd play better if they weren't)

Interesting development here.

A clan leader evaluating a challenge will look at any player's rating.
A good clan leader will look at both ratings to get some idea of this player's actual strength.
And rating history.
And may toss the challenge based on tournament rating history.
Especially early on in the new system.
Not saying this is good or bad.
But we will all learn new things in the new world.

And some clan leaders will be better than other clan leaders at exploiting the new system.

And the complaining of system manipulation will start all over again.

Originally posted by robbie carrobie
Actual site formula utilised to project ratings. Thanks to moonbus.

Hypothetical scenario.

Clan A v Clan B

1) robbie 1798 v mghrn55 1810: result 1/2 - 1/2
2) tomeasyrider 1600 v hetrz van rental 1610: result 0-1
3) settl 930 v suzzianne 1000: result 1-0

The following formulas were copy-pasted from the RHP FAQ:

New Rating = Old ...[text shortened]... two, + 19 for game three = +4 points for Clan A. (Do the math for Clan B.)

Hope this helps.

too complicated meaning open to collusion by clever manipulators such as yourself.

We need a simple straightforward system with a clan review board in place to address issues such as collusion, sandbagging, etc.

Originally posted by my2sons
too complicated meaning open to collusion by clever manipulators such as yourself.

We need a simple straightforward system with a clan review board in place to address issues such as collusion, sandbagging, etc.

Its not too complicated at all, you are simply too dim. Can you tell us why such a system is open to collusion for it cannot be because its too complicated, infact its very simple. Your suggestion of a review board will not address the issue of those who sandbag and do not get caught, will it. We cannot constantly police the system, we need a system that is self regulating and that rewards clans for good play.

Originally posted by mghrn55
Interesting development here.

A clan leader evaluating a challenge will look at any player's rating.
A good clan leader will look at both ratings to get some idea of this player's actual strength.
And rating history.
And may toss the challenge based on tournament rating history.
Especially early on in the new system.
Not saying this is good or bad. ...[text shortened]... loiting the new system.

And the complaining of system manipulation will start all over again.

what changes would you make to the constant K