CLA Clan Rankings [Jan 14-2010]

[Namor]

haha Eyestone, we were obviously typing at the same time and have said pretty much the same thing.

[pmchugh]

So those in the top 11 spots are going to be inundated with requests for a challenge, whilst those in the bottom half are going to have difficulty finding any interest.

That would be the case if anyone paid close attention to the laders, but it's pretty clear the ladders aren't reflective of ability and are liable to change so I wouldn't worry about it and just play whoever you think matches up well with you.

[Chariot of Fire]

The idea of running something along similar lines to the CC scoreboard makes very good sense - and would surely be a great relief for the clan administrators due to its simplicity.

Take, for example, the new ladder - and let's make that the official ranking. Get rid of all historical data and start the whole thing new (tho of course historical data is used to determine a clan's starting position on the new ladder). A clan at #1 has a game value of 2.0 and this is graded all the way down to the bottom clan that has a game value of 1.0. Thus if THOTA (at 2.0) challenged a lowly ranked Clan X (at say 1.20) in a 40 point series and won 26-14 the calculation would be THOTA (26 x 1.2 = 31.2pts) and Clan X (14 x 2.0 = 28pts). THOTA win the challenge 31.2 vs 28 (not 26-14). Note that if the games won had been 25-15 the outcome would be 30pts per team and a draw. Much fairer eh?

Adopting a handicapping system such as this makes it a much more level playing field. It is easy for a clan to work out how many games they must win to win the challenge. It eliminates farming of lower ranked clans and is easy to administer. The movement up and down the ladder would be based on one's handicap (like golf I suppose). THOTA's handicap would be adjusted by the equation of 2.0 x (31.2/28) = 2.23 and Clan X would be 1.2 x (28/31.2) = 1.08

An additional advantage of this system, adopting the mercy rule, is that once a series cannot be lost (say THOTA reached 26 games and led 26-8) then all remaining games may be 'awarded' to the losing team so that the result is officially 26-14 and THOTA's handicap isn't too adversely adjusted. Of course, this is at the discretion of the team that has secured the win already. If they need to win 30-10 for example (to get their handicap to a higher figure that moves them up the ladder) then they may wish to play until they reach that magic number of 30 and then stop the challenge.

I see this as being a far simpler method, with just few & easy stats that can be incorporated on to the chart so that anyone can quickly work out how many games in a challenge must be won to climb the ladder.

In principle it works the same as the CC scoreboard, and also makes every opponent as difficult to beat as any other due to the graded handicaps. Definitely worth considering.

[Scott-Land]

The idea of running something along similar lines to the CC scoreboard makes very good sense - and would surely be a great relief for the clan administrators due to its simplicity.

Take, for example, the new ladder - and let's make that the official ranking. Get rid of all historical data and start the whole thing new (tho of course historical data is used to determine a clan's starting position on the new ladder). A clan at #1 has a game value of 2.0 and this is graded all the way down to the bottom clan that has a game value of 1.0. Thus if THOTA (at 2.0) challenged a lowly ranked Clan X (at say 1.20) in a 40 point series and won 26-14 the calculation would be THOTA (26 x 1.2 = 31.2pts) and Clan X (14 x 2.0 = 28pts). THOTA win the challenge 31.2 vs 28 (not 26-14). Note that if the games won had been 25-15 the outcome would be 30pts per team and a draw. Much fairer eh?

Adopting a handicapping system such as this makes it a much more level playing field. It is easy for a clan to work out how many games they must win to win the challenge. It eliminates farming of lower ranked clans and is easy to administer. The movement up and down the ladder would be based on one's handicap (like golf I suppose). THOTA's handicap would be adjusted by the equation of 2.0 x (31.2/28) = 2.23 and Clan X would be 1.2 x (28/31.2) = 1.08

An additional advantage of this system, adopting the mercy rule, is that once a series cannot be lost (say THOTA reached 26 games and led 26-8) then all remaining games may be 'awarded' to the losing team so that the result is officially 26-14 and THOTA's handicap isn't too adversely adjusted. Of course, this is at the discretion of the team that has secured the win already. If they need to win 30-10 for example (to get their handicap to a higher figure that moves them up the ladder) then they may wish to play until they reach that magic number of 30 and then stop the challenge.

I see this as being a far simpler method, with just few & easy stats that can be incorporated on to the chart so that anyone can quickly work out how many games in a challenge must be won to climb the ladder.

In principle it works the same as the CC scoreboard, and also makes every opponent as difficult to beat as any other due to the graded handicaps. Definitely worth considering.

I suggested this at the inception of CLA ( with a slight variation on how the points were calculated ). I also adjusted the points won by the total games per challenge- obviously weighting a 60 game challenge vs a 20 game challenge in order to give clans the option to play however many games they wanted with a min of 20 and a ceiling of a 100. But where we differ, I would've needed the archived data to calculate the rank at the time the challenges were played.

One of the reasons was to establish the strengths of clans at the time the clans played. Let's say for an example; Thota played against Bushwackers-- they were a 'Colonel' clan when they played thota but they fell to a 'Cook' when Dragoons played them. Both challenges obviously couldn't yield the same points won/loss- at least not to the point of being accurate. The archived data would give us that necessary information-- an accurate time line of the strength of clans when they faced. All clans started at the same point level, just as a new recruit begins at 1000.

Another issue/ problem was the length of time it would take to establish a separation in 'clan ranks'. I was under the assumption Jp's Ladder would require a 2 or 3 year window for it to level. If that were the case (and acceptable), it's possible to have enough clans / challenges to inflate the rankings within that time frame (EDIT) to see a separation between clans.

At any rate it wasn't received very well and I stopped working on it. I obviously like where you're headed with your suggestion and hope it goes towards that direction.

[Chariot of Fire]

Thanks for supporting the idea. Hadn't realised you had already proposed something along similar lines (maybe I could have taken your concept and embellished it?)

If I'm reading you correctly then the way around a sudden drop is to establish the handicap weighting at the time the challenge is made, so if THOTA had two concurrent challenges (whilst weighted at 2.0) then that is what is applied when the final calculations are made for both challenges, i.e. the challenge to finish later doesn't get adjusted due to the handicap change after the other challenge finishes, so playing a 'major' clan will still yield a 'major' result, even if that clan happens to have slipped to captain in the meantime. It just forms part of the challenge criteria that is posted when a new challenge becomes 'live', i.e.

THOTA 2.0
TOFU 1.75
Best of 60 game series
20 doubs
20 trips
20 quads

THOTA 2.0
Clan X 1.3
Best of 40 game series
20 doubs
10 trips
10 quads

Regardless of which of these challenges finished first, THOTA could easily calculate that in both instances they would have to win 33 games vs TOFU and 25 games vs Clan X as the formula applied is that which was apparent at the time the challenge was made.

The formula's pretty simple. Take your clan's handicap and divide it by the sum of the two handicaps of the competing clans and multiply by number of games (or points) in the challenge, e.g. 2.0 / (2.0 + 1.3) x 40 = 24.24 (that's the number of games a 2.0 rated clan would have to win against a 1.3 rated clan in a 40 game challenge, i.e. 25 games).

I like the simplicity of the system and I don't think it would be too hard to apply handicaps to each clan in the ladder given the data we already have. And a simple ruling for new clans is that they must undertake one challenge before they can appear on the ladder - that initial challenge being used as a gauge to calculate where they enter the ladder. Every clan on the ladder must at some stage accept a challenge from a new clan - this distributes it fairly - and can be marked with a * next to a clan's name once it has done so. Alternatively a new clan is simply given a handicap of 1.5 when it joins the ladder.

[Namor]

Chariot's suggestion is brilliant in it's simplicity, so I can't understand why it was rejected when Scott-Land proposed something similar. it may have needed fine tuning, but it was worth a closer look.

Alternatively a new clan is simply given a handicap of 1.5 when it joins the ladder.

This is the fairest way of dealing with new clans. It makes sense to assume a new clan to be of average ability.

So those in the top 11 spots are going to be inundated with requests for a challenge, whilst those in the bottom half are going to have difficulty finding any interest.

That would be the case if anyone paid close attention to the laders, but it's pretty clear the ladders aren't reflective of ability and are liable to change so I wouldn't worry about it and just play whoever you think matches up well with you.

I think you missed my point here.

What I was trying to get across, is the fact that those from position 26 down can't possibly climb above 26 until they have played an A level clan and it may prove difficult for a B level clan to get an A level interested in a challenge.

Further, and more importantly, this ladder is an insult to all B level clans. As it stands any result against a B level, counts for absolutely nothing, until two clans happen to be even on A level average. As we can clearly see that is not likely to happen very often, since all clans have played only 1-4 matches and yet there are few with the same average. Give each clan another 5-6 matches and we will rarely see two clans with the same A level score.

[Eyestone]

Namor wrote:
Further, and more importantly, this ladder is an insult to all B level clans. As it stands any result against a B level, counts for absolutely nothing, until two clans happen to be even on A level average.

Very important point. I really can't understand how no one reacted to this when this new ranking was voted on. If 2 clans have played one challenge against an A-clan and gotten different results (very likely scenario) these 2 clans can play a thousand challenges against B-clans and it won't affect their rank because the only thing that matters is the A-rank And that's just crazy...

[Qwert]

for me this ranking its to much confusing-can its possible to create something much simple formula for calculation?
for example :
1.If you win,you get 1 point.
1.1 tie results-bouth clan get 0,5 points
2.If you win with diference of more then 5 games(30 games-18 wins),you get extra 0,3 points.
3.If you win with diference of more then 10 games(30 games-21 wins) you get extra 0,6 points.

no negative points.plain and simple.

[pmchugh]

Agreed about making the B level games pointless, as right now any clan who has played against an A raned clan is above any other that has not regardless of their results in any matches. Basically this is another ladder that fixes the top few clans "correctly" until a few updates down the line, no offense jp but i don't even see the point in having a ranking system.

for me this ranking its to much confusing-can its possible to create something much simple formula for calculation?
for example :
1.If you win,you get 1 point.
1.1 tie results-bouth clan get 0,5 points
2.If you win with diference of more then 5 games(30 games-18 wins),you get extra 0,3 points.
3.If you win with diference of more then 10 games(30 games-21 wins) you get extra 0,6 points.

no negative points.plain and simple.

That would encorouage farming.

[Qwert]

That would encorouage farming.

What you mean?

[pmchugh]

That would encorouage farming.

What you mean?

If you get the same amount of points for beating waters fury as you do for beating thota, why play thota? Just target the weak clans.

[jpcloet]

Farming of B clans would mean nothing since only A points matter for A division.

[Chariot of Fire]

He knows that JP. He was merely responding to qwert's suggestion that points be awarded for a win (regardless of the oppo) on a sliding scale.

[reptile]

is there like an overall standings points? i see you have us ranked and also have points per section, but not overall so we know how far away from thota we are.

[pmchugh]

is there like an overall standings points? i see you have us ranked and also have points per section, but not overall so we know how far away from thota we are.

Forget it move on,. and wait for jp to magically invent sometin that mkes fekin sense

[reptile]

lol

[Georgerx7di]

The following ladder was approved by a vote of 24-2 (9Abstains) in the CLA on Nov 6th.

1. The ladder has both decay (aged out challenges) and depth (number of games in a challenge) weightings
Points calculation is (Win%-0.5)*DecayFactor*ChallengeSizeFactor
2. There is NO league data in here
3. In order to get an A rating, you need to have faced 2 A rated clans. If you performance against A clans is so poor, the committee (currently non-existent) can re-class a clan to a B rating.
4. Clans are sorted by class. A clans are sorted by Average A results. Clans in B are sorted first by A challenge only clans, and then by the rest based on their B class ratings.
5. Ratings will be released on the 15th of each month, and become recurring as to create a rhythm.

Not to be a pest, but what was the reasoning for having this A & B arbitrary designations? Does this take into account the records of A clans vs other A clans or is this based on the assumption that all A clans are equal? Is is accurate saying, based on this formula, that a 26-24 victory over thota is less valuable than a 27-23 victory over Generation 1: The clan (no offense to G1)? The RPI thing seemed pretty cool and at least I could figure it out, not sure why we switched to this.

I know that you are busy jp, but I would really like to know the answer to his question if you have time.

[freakns]

26th?! brilliant

[freakns]

The idea of running something along similar lines to the CC scoreboard makes very good sense - and would surely be a great relief for the clan administrators due to its simplicity.

Take, for example, the new ladder - and let's make that the official ranking. Get rid of all historical data and start the whole thing new (tho of course historical data is used to determine a clan's starting position on the new ladder). A clan at #1 has a game value of 2.0 and this is graded all the way down to the bottom clan that has a game value of 1.0. Thus if THOTA (at 2.0) challenged a lowly ranked Clan X (at say 1.20) in a 40 point series and won 26-14 the calculation would be THOTA (26 x 1.2 = 31.2pts) and Clan X (14 x 2.0 = 28pts). THOTA win the challenge 31.2 vs 28 (not 26-14). Note that if the games won had been 25-15 the outcome would be 30pts per team and a draw. Much fairer eh?

Adopting a handicapping system such as this makes it a much more level playing field. It is easy for a clan to work out how many games they must win to win the challenge. It eliminates farming of lower ranked clans and is easy to administer. The movement up and down the ladder would be based on one's handicap (like golf I suppose). THOTA's handicap would be adjusted by the equation of 2.0 x (31.2/28) = 2.23 and Clan X would be 1.2 x (28/31.2) = 1.08

An additional advantage of this system, adopting the mercy rule, is that once a series cannot be lost (say THOTA reached 26 games and led 26-8) then all remaining games may be 'awarded' to the losing team so that the result is officially 26-14 and THOTA's handicap isn't too adversely adjusted. Of course, this is at the discretion of the team that has secured the win already. If they need to win 30-10 for example (to get their handicap to a higher figure that moves them up the ladder) then they may wish to play until they reach that magic number of 30 and then stop the challenge.

I see this as being a far simpler method, with just few & easy stats that can be incorporated on to the chart so that anyone can quickly work out how many games in a challenge must be won to climb the ladder.

In principle it works the same as the CC scoreboard, and also makes every opponent as difficult to beat as any other due to the graded handicaps. Definitely worth considering.

this idea have one big flow, the one that cant be overlooked. it makes sense.

[jpcloet]

We've already tried ELO which is the same as the CC scoring and there were several problems. As to the question above. You can play all the B clans you want, but history has shown, that until you play the "established" clans, your rating means squat. I liked the RPI personally, but it had several problems. Each ladder gets progressively better. I do have a solution for all of this. However, it requires some guidelines which many of the clans won't agree to yet. The suggestion of 2 ladders is out there, but I'd rather have just one.

[freakns]

We've already tried ELO which is the same as the CC scoring and there were several problems. As to the question above. You can play all the B clans you want, but history has shown, that until you play the "established" clans, your rating means squat. I liked the RPI personally, but it had several problems. Each ladder gets progressively better. I do have a solution for all of this. However, it requires some guidelines which many of the clans won't agree to yet. The suggestion of 2 ladders is out there, but I'd rather have just one.

i really dont see a problem with CoF proposition. you establish a starting ladder. then you establish a rules for getting points. and finally, adopting mercy rule will prevent farming. i really dont see a problem with it. as this ladder is maybe the worse so far. and Nemesis is the clan that will prove all of its stupidity, as right now we are ranked 26th even thou we are undefeated. and after clan war with Empire in which we will be severely defeated we will climb up the ladder!!! that makes absolutely no sense to me...

[jpcloet]

I agree the B clans will look weird but, I could leave them all out then until they face an A rated clan and call them all unrated. I'd love to do a pure leap frog table with semi-standardized challenge setups and rules, but a number of clans are hung up on "home games".

[danryan]

Speaking of A and B clans, when will we see an update to the rankings?

Thanks guys.

[jpcloet]

Likely Dec 15th, not many completed challenges of late.

[OlJock99]

So, the last time I checked the rankings (early Nov/late Oct), Nemesis was 3rd. Too high in my opinion, but there we were. Checked it and we're 26th.