Clan Rankings - RPI Version [Archived]

[ahunda]

I would say No ahunda, because in clan challenges its all against one clan, the clan league is with a variety of clans.

2 separate things, again, that is why they should be separate.

All right. I guess, we simply disagree on this one then.

As far as I can see, the league has been the greatest single event in terms of clan competition at the site so far, and I can´t see any reason, why it should be excluded from the ladder.

I understood the concern about counting those 20-game-mini-sets on the same level as official challenges. But not the general rejection.

thank you for implementing this jp

does this also mean you are separating the 2 as well? challenges and league?

If I understood jpcloet right, he suggested to have one ladder, that exclusively counts official challenges. Minimum games yet to be voted on.

And then another comprehensive ladder, that includes both challenges & the league. And, if this is ever going to happen, clan tournaments too.

[hwhrhett]

ive been trying to keep up with this, but i think im missing something...

is the only reason that people dont want it to count towards the ladder because they think 20 games is too small?

[Bones2484]

ive been trying to keep up with this, but i think im missing something...

is the only reason that people dont want it to count towards the ladder because they think 20 games is too small?

And they don't want a loss on their precious records!

Seriously. The vote was 16-3. THOTA was present during the vote. TSM was present during the vote. Bruce was present during the vote.

I don't understand why, when we have a CLEAR majority, that the "top" clans are once again forcing decisions on this? Again I repeat: this is laughable. There were no complaints until a few clans saw a few marks in their loss columns this past week. Where were all the complaints the past half year when all this info was in the Clean League thread? Why did no one speak up against this in the CLA forum (past saying "I'm on the fence about this")?

[Scott-Land]

I for one will NOT be playing in season 2 if clan league is counted towards clan challenges.

Did you read my suggestion, Blitz ?

I understand the concerns about counting 10-20 game sets as challenges for the ladder. But could you live with taking the 180 league games as one big total ? That are far more than your suggested 40 game minimum for official challenges, and it should really rule out the luck factor. After all, at the end of the season most clans ended up, where one would expect it.

I repeat: Each participating clan played 180 competitive games in the league. Do you want to throw these results out of the window ? Or could we find a way to incorporate them into the ladder ?

And to clarify my suggestion: The league results would count towards the ladder, but not as a challenge.

The current table in the first posts shows Grade, Win % and W-L-T. I would add another column for league results and remove them from Win % and W-L-T. Thus the league would be in a category of its own, having no impact on the challenge record of clans, but still count towards the overall ranking (Grade).

I'll have to compare the calculations 2 ways.

1. Using the actual results
2. Using a fictional "Average Clan" of all the clans involved.

I´m not really good in these things, but if you calculated the fictional "average clan", wouldn´t you simply need to take the overall wins / losses for one calculation per clan ?

In theory, I don't think the results would change much, although, it will affect W-L-T by only 1 now instead of 7.

As stated, I wouldn´t count it as a challenge, but put it in a separate category of its own. Taking it out of the W-L-T record for official challenges altogether.

And one difference would be, that all 180 games would count, not only the 80 of your division.

Do you figure out the fictitious clan at the beginning or the end? I think D1 would be ok at the beginning, D2 has a number of clans with not enough history (less than 1.00 reliability factors) and likely should be at the end.

I guess, I´d do it at the end. So challenges, that took place during the season, are considered already.

1. RPI Clan Ladder - Challenges only-minimum 40 games
2. RPI Clan Ladder - League or Comprehensive-minimum 20 games

Couldn´t we assign some kind of quotient to challenges, so that their impact on the ladder is directly related to the size / scale of the challenge ? A 60 game challenge counting as much as three 20 game challenges ?

Small & new clans would have the option to try out smaller challenges, whilst established big clans would be free to set their individual minimum game numbers for challenges.

Great suggestion Ahunda-- I think TSM would be satisfied with that. Even potentially having the league count as a percentage of the overall clan ladder along with what you're suggesting. It will eliminate a lot of dissatisfaction from a lot of clans.

[khazalid]

all hail the henner!

[hwhrhett]

i think that adding the 180 games as 1 big challenge will hurt the smaller clans too much... they already have a bunch of losses being recorded, why steal their few wins from them. it will create too big of a gap between the lower clans and the upper clans no?

are we considering rescinding the fact that the league will count for the ladder? i mean, im all up for changing it for season 2 if people are unhappy, but i dont like the idea that we will be rescinding what we already all agreed upon before based on 2 or 3 unhappy clans, regardless of where they are on the ladder...

and on that thought, if it was 2 or 3 clans at the bottom of the ladder complaining about the system, would it be taken as seriously?

[ahunda]

i think that adding the 180 games as 1 big challenge will hurt the smaller clans too much... they already have a bunch of losses being recorded, why steal their few wins from them. it will create too big of a gap between the lower clans and the upper clans no?

I´m not entirely sure, which smaller clans you mean. But with the current approach, this would be Wins-Ties-Losses recorded for the ladder for the clans, that finished last/bottom in the season:

Last Warriors 0-0-4
DVLL 0-0-4
Regulators 1-1-2
Mythology 1-0-3
LoD (now AoD) 2-0-2

My suggestion was to not count the league in the ladders Wins-Tie-Losses at all. So they´d all get 0-0-0 in that scenario. And for most of them, that would actually look better, because you would "steal" more losses than wins here.

But as I understand it, the ladder is not merely based on Wins-Ties-Losses anyway. I am still not sure, how the ranking works in detail, but jpcloet posted this in the current THOTA - DBC challenge thread:

38-22 or higher

What THOTA needs to win by to increase their ladder gap over LOW

So I assume, that if THOTA would win by a lesser margin, they would still get a win booked in the ladder statistics, but actually lose points towards LOW, because they did worse than their current standing would require. And I also assume, that DBC can actually gain points for the ladder, even if they lose the challenge.

You can compare it to the CC scoring system, where a cook will gain points, when he wins 1 out of 6 or 7 games against a colonel. The gap in the ladder will regulate itself, because the higher ranked clans need not only win against lower ranked clans, but do so with a certain margin. If they don´t manage that, the gap begins to close.

As for the rest: I get your point. I was not part of the original discussion, but only got involved here & now and put forward a suggestion, that I like better than the current approach.

Because I actually agree, that counting 20 Doubles on the same level as 40-60 game challenges with Doubles, Trips & Quads is not right. And because I think, there should still be a way to incorporate the league into the ladder. And because I don´t see, why only 80 out of 180 league games should count.

I agree, it might not be great to discuss this now, after the season is finished and agreements were made months ago. On the other hand: Why not try to evolve & improve the system ?

[hwhrhett]

Because I actually agree, that counting 20 Doubles on the same level as 40-60 game challenges with Doubles, Trips & Quads is not right. And because I think, there should still be a way to incorporate the league into the ladder. And because I don´t see, why only 80 out of 180 league games should count.

I agree, it might not be great to discuss this now, after the season is finished and agreements were made months ago. On the other hand: Why not try to evolve & improve the system ?

i am totally for improving the system to everyone's liking for the future...

but as far as a 20 game challenge not counting as much as a 60 game challenge, it was my understanding that it was already like this.... or i may be thinking of some older version of the ladder... did we shitcan that?

[waseemalim]

what a lively discussion. I believe LoW voted in favor of having the League games included.

The objections that people have raised against it probably stems from a lack of understanding as to how this ladder works. It doesnt count individual challenges to rate clans. It accumulates all the games that two clans have played against each other to get the result. Which means every game counts. To explain the difference:

Clan A beats Clan B 40-20
Clan B beats Clan A 12-8

If you are counting challenges it would be (1-1), which we report. But this 1-1 result has absolutely no bearing on the points. The points are based off cumulative games -- in this case, 48-32.

If someone has any sense of statistics, they will agree that more data points is better than less. And as such, I have been in support of including the clan league results. Because they add a lot more meaning to the numbers. Especially given that the ladder works off of second and third degree results.

On a side note, I personally am not a fan of having the clan war results column (I mean the 9-1-0 column). Because I think it is worthless. That column exists in a vaccum. It doesnt help anyone understand relative strength of clans.

I cant see why it is so difficult to grasp. In the end, I guess some clans are just unwilling to grasp it.

[Bones2484]

On a side note, I personally am not a fan of having the clan war results column (I mean the 9-1-0 column). Because I think it is worthless. That column exists in a vaccum. It doesnt help anyone understand relative strength of clans.

I cant see why it is so difficult to grasp. In the end, I guess some clans are just unwilling to grasp it.

This is also true.

But if people really want to see this column, is it possible to just split it into two columns? One column for challenge results and one column for league results? This way the stats still feed into the overall RPI score, but the records are separated for those that truly care about it?

[hwhrhett]

On a side note, I personally am not a fan of having the clan war results column (I mean the 9-1-0 column). Because I think it is worthless. That column exists in a vaccum. It doesnt help anyone understand relative strength of clans.

I cant see why it is so difficult to grasp. In the end, I guess some clans are just unwilling to grasp it.

This is also true.

But if people really want to see this column, is it possible to just split it into two columns? One column for challenge results and one column for league results? This way the stats still feed into the overall RPI score, but the records are separated for those that truly care about it?

well, the win/loss record is kept here by marval:

viewtopic.php?f=441&t=64025

so, if people want to see win/loss records they can look there. no?

[GrimReaper.]

just curious... when will this be updated?

[jpcloet]

Generally done once per month.

[GrimReaper.]

Ok we just need to see the updated version before our next challenge

[jpcloet]

Updated.

[GrimReaper.]

thx JP

[Fireside Poet]

Wow. We had a nice win over Sky Force and we didn't move... did we?

[jpcloet]

Wow. We had a nice win over Sky Force and we didn't move... did we?

Nope, I haven't done a scenario check but SF's ranking was slightly lessened by the league results I think, which indirectly lessens your victory.

I don't see an update to the ladder for a while, but we may need to revisit the weightings for a slight alteration in the future.

[Chuuuuck]

I know I am joining this discussion late but I feel since it is still an active discussion with opinions on both sides then I might throw in something that isn't really being discussed as much.

First of all, I don't agree that the league results should be added to the ladder because 20 games is simply too small and is very susceptible to luck. But more importantly I don't think they should be added because of the heavy concentration of doubles games in the league. If you look at most big clan challenges there isn't a focus on doubles games. This is for a simple reason, most everyone that has put some thought into this knows that doubles games can be based FAR more on luck than skill. A good team over a mediocre team might only win 55% of the time (or less) on the low side whereas they might only win 70% of the time on the high side (and IMO, this is on the very high side). So if you take that over time then there is no doubt that teams that don't deserve wins will get them in the clan league due to pure and simple luck and nothing else. And IMO, the ladder should take out as much luck as absolutely possible and try to post records that truely represent who is the best and in what order. Thoughts?

[andy_is_awesome]

I know I am joining this discussion late but I feel since it is still an active discussion with opinions on both sides then I might throw in something that isn't really being discussed as much.

First of all, I don't agree that the league results should be added to the ladder because 20 games is simply too small and is very susceptible to luck. But more importantly I don't think they should be added because of the heavy concentration of doubles games in the league. If you look at most big clan challenges there isn't a focus on doubles games. This is for a simple reason, most everyone that has put some thought into this knows that doubles games can be based FAR more on luck than skill. A good team over a mediocre team might only win 55% of the time (or less) on the low side whereas they might only win 70% of the time on the high side (and IMO, this is on the very high side). So if you take that over time then there is no doubt that teams that don't deserve wins will get them in the clan league due to pure and simple luck and nothing else. And IMO, the ladder should take out as much luck as absolutely possible and try to post records that truely represent who is the best and in what order. Thoughts?

Looking at clan league season 1 (dubs), I think the best teams finished on top and the worst teams finished on bottom.
...so I don't know about mediocre teams taking a crack at the top dogs, just 'cause there's more dubs.

There certainly is more luck involved when there are less games. I don't deny that. But I don't think it's enough to be significantly concerned.
In my opinion, 20 games is plenty to apply toward the later.

[jpcloet]

I know I am joining this discussion late but I feel since it is still an active discussion with opinions on both sides then I might throw in something that isn't really being discussed as much.

Still active, it must be in the other thread in the CLA, but the league only represents a small portion of the ranking and really only solidifies the experience OOWP etc.

First of all, I don't agree that the league results should be added to the ladder because 20 games is simply too small and is very susceptible to luck. But more importantly I don't think they should be added because of the heavy concentration of doubles games in the league. If you look at most big clan challenges there isn't a focus on doubles games.

I would argue now that 1v1 games in clan challenges don't make sense either, but they are also in the ladder. The league data is good data to have, just wish the challenges could be slightly bigger, but time and scheduling prohibits that.

This is for a simple reason, most everyone that has put some thought into this knows that doubles games can be based FAR more on luck than skill. A good team over a mediocre team might only win 55% of the time (or less) on the low side whereas they might only win 70% of the time on the high side (and IMO, this is on the very high side). So if you take that over time then there is no doubt that teams that don't deserve wins will get them in the clan league due to pure and simple luck and nothing else. And IMO, the ladder should take out as much luck as absolutely possible and try to post records that truely represent who is the best and in what order. Thoughts?

I think luck plays a part of any game. If you really thought luck was the primary factor of doubles games, then no one would play them. Part of the reason of expansion to trips and quads.

[reptile]

who did thota lose to?

[jpcloet]

who did thota lose to?

They lost to ID in the clan league, 12-8.

[khazalid]

i thought we lost to g1 too, but maybe that was just over 10 games

[jpcloet]

i thought we lost to g1 too, but maybe that was just over 10 games

That is correct, it was only a cross-over 10 game match.