Conquer Club

Donelladan wrote:October ranking.


~~~~~Unranked : LOTZ ranking 621 weight 122~~~~

1st post updated.

The way it looks now is what I meant
Now it shows the active Clans and we can see how far they are away from being added to the main section
Ascension and Death by Dice could also be included in the unranked section since they are Active Clans

Ascension and DBD only had one clan war completed at the 1st october. I'll add them on the next ranking.

Hi Don, great to see the ranking is up running again

I have no doubt that TOP will be dropping a lot of points and some spots, when war against OSA and Fiasco is finished.

So my question wont matter much in the long run.

But I cant keep wondering a bit, why TOP keep loosing a few points every month, even though we do not have any war or e.g. CL fight finishing ?

It is not a big deal, just a small missing

You answered it yourself. If you're not playing any wars your weight slowly decays.

dkmaster wrote:.

But I cant keep wondering a bit, why TOP keep loosing a few points every month, even though we do not have any war or e.g. CL fight finishing ?

To calculate someone ranking, for each war I first calculate the "performance" which give me a number of positive ( or negative points) for a given war.

Then I sum up the results of all the war multiplied by a weight coefficient ( bsed on how many games in the war) and a decay factor
So someone ranking looks like :
Ranking = Sum ( decay factor *weights * performance) / Sum ( decay factor * weights)

The decay factor is 1 if the war finished today, 0 if the war finished 2 years ago. and 0.5 if the war finished 1 year ago, (it's linear).
So each day that goes by, decay factor will decrease the influence of all wars.

If you had bad results 2 years ago it could also means your score increase even though you aren't playing.
Ofc the decay factor is affecting every clan war; so for high ranked clans it always means your score decrease with time if there is no new results. ( it go towards 1000 for everyone).

Thaks for a good explanation


I have been looking for that for years now


I always thought that gained points are to keep, like at cc scoreboard and ranks.

But sounds like it is kinda like in tennis. There points are gone every year. and you have to get new victories/points every year to keep your rank.

As I understand it here. Both positive and negative points for every war fought, is halfed after 1 year, and fully gone after 2.

Is that correct ?

Correct.

dkmaster wrote:As I understand it here. Both positive and negative points for every war fought, is halfed after 1 year, and fully gone after 2.

Is that correct ?

To be clear, it is linear and the decay is calculated daily.
It is such that .5 at exactly 1 year and 0 after 2 years, but also .75 after 6 months, .25 after 18 months etc.

Donelladan wrote:

dkmaster wrote:.

But I cant keep wondering a bit, why TOP keep loosing a few points every month, even though we do not have any war or e.g. CL fight finishing ?

To calculate someone ranking, for each war I first calculate the "performance" which give me a number of positive ( or negative points) for a given war.

Then I sum up the results of all the war multiplied by a weight coefficient ( bsed on how many games in the war) and a decay factor
So someone ranking looks like :
Ranking = Sum ( decay factor *weights * performance) / Sum ( decay factor * weights)

The decay factor is 1 if the war finished today, 0 if the war finished 2 years ago. and 0.5 if the war finished 1 year ago, (it's linear).
So each day that goes by, decay factor will decrease the influence of all wars. (1)

If you had bad results 2 years ago it could also means your score increase even though you aren't playing.
Ofc the decay factor is affecting every clan war; so for high ranked clans it always means your score decrease with time if there is no new results. ( it go towards 1000 for everyone). (2)

Excellent explanation, though have couple comments about the underlined.

(1) - Not exactly, the weight decreases as time passes, but a particular war's influence on a clans total ranking does not necessarily decrease as time passes.

(2) - In the case of No New Results, score does not decrease for all, and does not particularly decrease for a high ranking clan. Nor does it go towards 1000.

To address the original question from DKmaster,
As Donelladan explained: even if a clan has no new results, the calculation of said clan's total score actually changes daily.
Furthermore, the proportional influence on total score by a particular war result is not constant as time changes. The influence can increase or decrease(or sometimes be the same).
In general, if no new results: as time passes, your newer results matter more, and your older results matter less. So if score slowly decreases, this is probably b/c newer results were not as good as older results.
[spoiler=Example]x,y,z are scores associated with 3 separate results and are all results for a clan. Assume all same amount of games.
a,b,c are decay factors variable of time.
Then, Score = (ax+by+cz)/(a+b+c)
let a - b= .1 and b - c= .1 (this is like saying each result happened exactly 73 days apart)
So when a=1(the day war x was completed) => (1x+.9y+.8z)/2.7 = Score
and a=.9(73 days after war x was completed) => (.9x+.8y+.7z)/2.4 = Score
when a=1, z's influence/weight is (.8/2.7) = .296 , y's influence = .333, x's influence = .370
When a=.9, z's influence/weight is (.7/2.4) = .292 , y's influence = .333, x's influence = .375
z's influence decreases, x's increases, y's is constant.
This can be shown to be true for all such that c>0.[/spoiler]

JPlo is right on everything. Especially on the 2nd point, big mistake from me. Ranking will actually go towards the last score registered. ( doesnt really matter since there should always be a new results, and last score would only have a strong influence if weight is really low, at which point you aren't on the scoreboard anymore, hence my comment about clans with low weight having a meaningless ranking previously).

but a particular war's influence on a clans total ranking does not necessarily decrease as time passes.

If there is no new results. But for an active clan, influence of past wars will always be decreasing as long as they play war or league.
At the moment we do have clans that are still "active" but haven't played since a while, but in general it shouldn't be the case, and those clan will go below the 150 weight soon if they didn't already.

So you are right, but it is not supposed to be the case, active clan should always have new results.

When you get a chance, it would be great to see the November rankings - pretty sure TOFU is back at #1.

Good for us, the more points you have when we beat you, the higher we'll go

Unlike the october ranking, I plan to do november before the end of the month. Should be there quite soon.

Yup and when I outlift Dan Green everyone will be impressed with my strength.

Nah, it’s a great opportunity for you, but it’s the finals it’s not about rank, it’s about the cup and in that sense it’s a great opportunity for us too.

Thank you for updating this, nice work!

rockfist wrote:Yup and when I outlift Dan Green everyone will be impressed with my strength.

Nah, it’s a great opportunity for you, but it’s the finals it’s not about rank, it’s about the cup and in that sense it’s a great opportunity for us too.

Sure. but at the same time the clan winning the cup usually reach the 1st rank in the F/D400 ranking. It might not be the case if LHDD wins though, since we are quite low, and we didnt get to face top so our semi final victory will give us much less point than yours.

Donelladan wrote:

rockfist wrote:Yup and when I outlift Dan Green everyone will be impressed with my strength.

Nah, it’s a great opportunity for you, but it’s the finals it’s not about rank, it’s about the cup and in that sense it’s a great opportunity for us too.

Sure. but at the same time the clan winning the cup usually reach the 1st rank in the F/D400 ranking. It might not be the case if LHDD wins though, since we are quite low, and we didnt get to face top so our semi final victory will give us much less point than yours.

OSA-TOP and LHDD-OSA still needs to be added. OSA will win quite some points for defeating TOP I guess, and that makes your convincing victory over OSA worth more points. I dont know if you will bridge the current gap, but if you win it will be close I think.

If LHDD were to beat us they would rise to #1, they are #4 now....both TOP and S&M have losses coming in the rankings so they will drop points...maybe not to below LHDD but much closer, if LHDD won against us they would gain points and we would also drop, but unless it happens its all magic unicorns in the sky.

Clan wars included :

show

RGV and DBD are really close to be included.

I was really surprised to see TOP above S&M, but that's likely because their war vs OSA isnt processed yet. I also want to note that this is the first time of this new ranking that S&M is not in first place anymore. They held first place in F400/D400 at least since their most recent CC victory, but I dont know when exactly they got it. Probably since they beat us in the semis of that year.

All the same, feels good to see TOFU in the #1 spot again. Now it's up to us to prove again that we belong there.

DBD's rating seems unusually high. How did that happen?

iAmCaffeine wrote:DBD's rating seems unusually high. How did that happen?

They started at 1000 points and won their wars 36-27(FSC) and 27-14(TS). Not sure if that should give them 135 points but those are big wins. They will go down next month though after lost vs ATL.

Is this the lowest ID has ever been?

iAmCaffeine wrote:Is this the lowest ID has ever been?

Probably, previously their lowest was 922 rating tho the rankings vary slightly so without running the F400 impossible to tell.

What is the weight and how is it calculated?

Weight is number of games played that count towards the ranking (within the last two years). Its added, although I'm not sure if games within the last year carry more weight than games in the prior year.