Originally posted by robbie carrobieA new K value can be arrived at.
Ok great I think its a fine and valid proposal. but eight is too small, you should not be allowed to arbitrarily throw games because the entire point of a performance based system is to prevent such abuses. If you are going to throw games you must be made to suffer for it. Furthermore you might find that the margins are so small between clans that it becomes something of a lottery.
One that meets with general approval.
As to the other issue.
In clan challenges, there are 2 competitions at play here.
1 - individual games on the chess board.
2- the clan challenge.
The objective of the challenge is to win the challenge.
Once the challenge is decided, the games tend to become less meaningful.
Perhaps a suggestion in this regard is that once a challenge is won by one clan, the points can be rewarded to the winning clan while the games can continue.
Challenges that look like they end in a tie must be played out to completion.
This suggestion would only apply if the current scoring system stays in place.
Which looks unlikely at the moment.
But it would have worked.
And it would have solved game dumping.
Originally posted by robbie carrobieJust been through and counted the thumbs up and thumbs down at the moment you are losing 47 - 98 or roughly half
Actually I have received private correspondence that a number of clan leaders are really interested in it. You it seems are simply here to cause trouble.
The majority of people on these posts do not want your ideas
I would like to toss this out there.
1 - do away with the gross point total. No longer relevant.
2 - use the rating change (rounded to nearest point) for each player in the challenge. ELO system will work. In-house formula will work too. Not that much different from ELO. Agree on a suitable K-value.
3 - do not use clan average rating as starting point. Start every clan at 0 at beginning of clan season. And then use rating change for each game in challenge to update the new clan net point total. Every game counts. No game tossing to close challenges.
4 - in addition to rating change, also award/deduct challenge points for the challenge itself. I would like to keep a reward for winning the challenge itself. Today that is 2 points per player in the challenge. That can be set to 1 point per player in challenge. Example - a 10 player challenge will award 10 points to winning clan for the challenge in addition to the ratings adjustments award for each game in the challenge. For losing clan, it can be decided whether to deduct challenge loss or set to zero.
A 10 player challenge that is close, for example 6-4 may result in rating award same as a 1 player challenge won 2-0. Larger challenges should have larger risk/reward attached.
This is a hybrid system which should bridge the perceived gap between current and new system.
5 - the challenge points will be awarded when the challenge has been decided, instead of completed as it is today. If a clan is ahead 6-4 with 1 game to play, award the points. This will eliminate game dumping. Rating points can be awarded when the game is completed or when the challenge is completed. This can be discussed amongst the clan community. Russ can analyse the complexities involved.
6 - eliminate the maximum 3 challenges between any 2 clans as the new scoring system should fix the imperfections. It is in everyone's best interest that the fluidity of challenges is enhanced.
7 - maintain a refereeing system to monitor any abuses of the system and administer corrective action accordingly.
This is open for feedback of course.
And all other suggestions are on the table.
Of course.
Originally posted by mghrn55??????
I would like to toss this out there.
1 - do away with the gross point total. No longer relevant.
2 - use the rating change (rounded to nearest point) for each player in the challenge. ELO system will work. In-house formula will work too. Not that much different from ELO. Agree on a suitable K-value.
3 - do not use clan average rating as starting p ...[text shortened]... This is open for feedback of course.
And all other suggestions are on the table.
Of course.
Originally posted by mghrn55any idea if their is a closing date on this thread?
You don't understand parts of it ?
Or you don't understand any of it ?
Pretty simple I think.
It builds on what others have started.
i would add
8] collusion points removed from last year and warnings handed out
9] challenges ending in a draw no points awarded. close the one man clan in collusionwith another in handing each other points
10] get rid of the dead player tactic, not moved in 14 gays banned from being selected until they move again
- Joined
- 12 Nov 05
- Moves
- 145614
Originally posted by mghrn55Hello mghrn55,
I would like to toss this out there.
1 - do away with the gross point total. No longer relevant.
2 - use the rating change (rounded to nearest point) for each player in the challenge. ELO system will work. In-house formula will work too. Not that much different from ELO. Agree on a suitable K-value.
3 - do not use clan average rating as starting p ...[text shortened]... This is open for feedback of course.
And all other suggestions are on the table.
Of course.
My feedback (for what its worth - there seems to be a lot of white noise from the usual suspects)
Number 1 - i assumed this had been the case for a while as standings are based on net points each year.
Number 2 - i actually have no clue how the ELO system works - ive spent the last couple of weeks trying to figure it out.
Number 3 and 4 (both merge sorry) - i like the fact that every game counts - especially in a big clan match finishing 11-9 means both clans get some points. (i wouldn't have any win bonus though)
Number 5 - Maybe instead of every game counts you could have the winning margin being the total points awarded (so there isnt dump loading at the end)
i.e - in a 6 player challenge Clan A go 7-0 up on Clan B. If they mass resign then it would be 7-5 - in old system Clan A gets +12 and clan B gets -12. If winning margin counted it then it would be Clan A +2 and B -2.
If you had a close match up finishing 11-9 then it would be +2/-2 (or 11/9)
Number 6 - As an ex-clan leader, i actually liked the 3 challenge limit as it meant i had to look for other clans to play which added variety.
Number 7 - A referring system sounds good but i think admin would be cautious after past experience with Game Moderators and people's reactions to Forum Moderators.
Originally posted by roma458) Agree.
any idea if their is a closing date on this thread?
i would add
8] collusion points removed from last year and warnings handed out
9] challenges ending in a draw no points awarded. close the one man clan in collusionwith another in handing each other points
10] get rid of the dead player tactic, not moved in 14 gays banned from being selected until they move again
9) Challenges ending in draw should not be null-rounds if one clan is lower rated. Drawing a higher-rated clan should count for something.
10) In principle yes, but 14 days is too short a time limit; many people (in EU at any rate) have more than 14 days paid holiday per year.