Clan War System 2013

The condition that only 10 coalition wars count stagnates the drive for more coalition wars. This slows down the entire season after the active coalitions make their 10, as there is no reason to try to get them anymore. A good number of times I've asked for coalition wars and leaders would decline saying they wanted their last matches against another coalition. Or they declined because they only have limited chances. Which is fine, I understand, thus the system is the problem. This needs to be changed so we encourage more coalition wars. More throughout the season.




Another problem with the current season system is competence. I don't mind competence, it's basically coalition-elo, but using it to rank a season isn't very nice. Stalins Martians could have went 10-0 this season and we still wouldn't of gotten even 2nd place via the current system. In the future when(if) atWar continues to grow, there may be many more coalitions participating and thus contributing their competence to the mix, and we'll have the rise of many coalitions with 1.5 or higher competence. These coalitions will be stifled from the competition since they cannot win the season, unless they purely played against coalitions with 1.5 or higher competence. Even then, a new coalition with experienced players can just get #1 easy because of their low competence.

Coalition wars are the pinnacle of competition on atWar, with multiple people involved and bragging rights on the line, they should be encouraged. One of the ways to encourage growth of the competitive side of atWar, is to encourage coalition wars. There are other threads about changing UN games, messing with duels, sp rates, tournaments, plenty of competitive stuff out there but there isn't much being done for coalitions. I seek to change that.




There were plenty of ideas thrown around in the threads at the top, let's compile them into something useful. My coalition has played 39 coalition wars since this seasons start(at time of posting). This is more than most coalition's entire CW history from their beginning, barring 7(including us). With a 28:11 w/l, a 2.5 ratio, and 72% of CWs played won for this season. Lets use these types of stats to form a new cw-friendly plan.

Winning coalition wars to increase a percentage out of your own total should be the goal, rather than a few wins out of 10 total. This gets rid of the competence concept and gives a drive for more coalition wars as you go. Here's some stats.

[Ordered according to this seasons ranking, stats are from start of season, 88 days ago, till present.]

coalition name - [T]otal games, [W]ins, [L]osses - percentage of games won of total
01. .187. - 9T 9W 0L - 100%
02. Elite Comrades - 9T 7W 2L - 77.78%
03. Campire Fellowship - 19T 10W 9L - 52.63%
04. Titans Creed - 12T 8W 4L - 66.67%
05. Stalins Martians - 39T 28W 11L - 71.79%
06. L.B. - 8T 5W 3L - 62.5%
07. Syndicate - 10T 6W 4L - 60.0%
08. The Avengers - 17T 8W 9L - 47.06%
09. The League - 21T 9W 12L - 42.86%
10. The Ancients - 4T 3W 1L - 75.0%
11. Mortak Kombat - 19T 7W 12L - 36.84%
12. evoL - 7T 3W 4L - 42.86%
13. Victorious Secret - 6T 3W 3L - 50.0%
14. Art of War - 9T 2W 7L - 22.22%


Now if we take these stats and re-order them by the win percentage...

100% - .187.
77.78% -Elite Comrades
75.0% - The Ancients
71.79% - Stalins Martians
66.67% -Titans Creed

This is the result. Now the problem is to stop a clan from getting 100% by just winning a single coalition war. Instead of the original maximum set that stifles coalition wars, lets set a minimum. This makes it so there is an ample sample for the percentages to even out to what they really should be like, and encourages coalition wars since there isn't a max you can have, just have as many as you want. Coalitions have 3 whole months, 10 games isn't nearly enough, you can get that in a week.

The minimum number of coalition wars to be ranked on the season list should be 25. The sample is large enough to create an accurate ranking and you can increase your percentage by having more coalition wars after your 25. Yes 25 CWs is pushing it, however the point is to encourage as many CWs as possible, so a high goal is a good choice. An ambitious choice.

Support, but I suggest 15 minimum and 25 maximum.

Prispeval Vetsd, 27.08.2013 at 16:11

Support, but I suggest 15 minimum and 25 maximum.

I'm completely against a maximum of any kind.

However I am pushing it saying 25, I was originally going to put 20 there but the point I'm making is to encourage as many as possible.

If the idea goes through the administrators can lower it to 15 or 20, their choice.

Prispeval Desu, 27.08.2013 at 16:17

Prispeval Vetsd, 27.08.2013 at 16:11

Support, but I suggest 15 minimum and 25 maximum.

I'm completely against a maximum of any kind.

However I am pushing it saying 25, I was originally going to put 20 there but the point I'm making is to encourage as many as possible.

If the idea goes through the administrators can lower it to 15 or 20, their choice.

Sounds fine to me, I just think 25 is too many and may force clans to make bad matchups for an automatic loss

Prispeval Vetsd, 27.08.2013 at 16:18

Prispeval Desu, 27.08.2013 at 16:17

Prispeval Vetsd, 27.08.2013 at 16:11

Support, but I suggest 15 minimum and 25 maximum.

I'm completely against a maximum of any kind.

However I am pushing it saying 25, I was originally going to put 20 there but the point I'm making is to encourage as many as possible.

If the idea goes through the administrators can lower it to 15 or 20, their choice.

Sounds fine to me, I just think 25 is too many and may force clans to make bad matchups for an automatic loss

I'm all for forcing cw's. Though that's me, others will disagree lol.

Throw in Pulse's suggestions for more SP and we're on the road to more encouragement too.

25 is a good number. Most clans won't attain it in a season as of now and the near future. However, goals must be established and this is a wise one. Simple system, but more effective and much more accurate under technical terms. Thus, giving all clans a shot at winning season.

Support.

Full Support

Also if these changes were to be implemented into the game, I think there should be a better prizes instead of just trophy alone, like double sp for a week etc.

http://atwar-game.com/forum/topic.php?topic_id=9175

I am not an expert on this anyway so the idea might sucks. It is probably better that the idea comes later after the first season of this new changes.

Hello Ivan and Amok,

.

I. Support.

I fully support this. I once tried to convince Amok otherwise but he literally said, ' I'm very reluctant to change the system. Ten CWs is enough already.'

I'd like to see if he's changed his mind.

As for me as soon as my coalition played all of out 10 clanwars my activity has definitely taken a fall. I no longer see any motivation to vigorously train my members and to keep pushing my coalition. One of these new systems must be implemented. The competence system is horrid. Quite terrible. Only a determined and loyal coalition like Stalin's Martians would continue to compete even though striving for another medal is hopeless. Without them the CW system would be dead.

Once again this has my full support. Thanks!
Cheers!

I support Desu

This was already posted, plz close this topic.

I support all except the 25 min-that needs to be lowered in spring, fall, and winter, because many players will be more inactive.

Desu, there's a problem in your system: if CLNs are motivated to go for wins, they won't CW a CLN they know they won't win against, thus picking weaker CLNs and then high rank CLNs wouldn't get any CWs at all. With a minimum of 25 CLNs it wouldn't be much different, I don't think people would risk a loss to get to compete.

I'm against the maximum of 10, we should be able to have unlimited CWs in a season. THEN add a nifty piece of code that makes matches against the same CLN and same players have diminishing returns (i.e. Stalins CWs some clan, where they use player A, player B and player C. if player C leaves and joins CLN Y, Stalins win against CLN Y, if player C plays, is slightly diminished; same thing with a CLN cwing the same CLN more than a certain number of times).

Then add huge SP and protocoin chance bonus, like 6 times for a CLN you didn't fight yet (and decreasing as you have more matches against that CLN).


I believe this would motivate people to play CWs for the SP involved; play different clns; and use different players.

^Well look at every active clan. Every clan that would actually compete would be able to cw with even ranked teams. With even ranked teams, I doubt anybody would back out of a cw. And since the players who would actually be playing in the CW are ranks 7+ who don't play UN games and don't allyfag, they know backing out is a LOT worse than playing and losing. Clans hold the most competitve drive out of all of atwar. I'm a noob rank 6 and I play 3V3 cw style games against high ranks. I've 1V1 ranks 8s and 9s. And I'd love to cw against a higher rank. Ill either win, or learn from my mistakes and become a better player.

CLAN POINTS AWARDED THROUGH CWs

I think the way to go is to create a separate type of points only obtainable through clanwars. Instead of SP have CP (Clan Points). And then offer a few cool things (upgrades, customizing options, something) unobtainable with SP. Then have an uncapped CW season.

Tbh 25 is very very ambitious, 1 a week should be the minimum so 12 Cw's to be counted in the standings at end of season. As Pulse mentioned there are still flaws with this system, however I think it is a stronger and fairer system then the current one. With a few details ironed out I will give my support!

Prispeval Desu, 27.08.2013 at 16:03

- Season limit changed from 10 maximum, to 25 coalition wars minimum
- To create a good sample, you need the full 25 CWs to be ranked
- You can still have more coalition wars to increase ranking/percentage after the 25 are done

This will result in only 1-2 Coalitions in the final ranking. Most Coalitions are too casual play 25 CWs per season. Look at how many reached even 10 games - last season it was only 2 Coalitions.

Citiraj:

- Win percentage out of total games per season used as marker, not competence.

This will just force coalitions to pick easy targets and avoid coalitions with established winning reputation. Current system rewards challenge, for both sides - since a more powerful coalition always results in more points. In the proposed system fighting top coalitions will be too risky - and why bother when defeating newbies will get you the same result? Top coalitions will really struggle to get the required 25 games.

Citiraj:

You have a full 4 days, I can only hope changes are made.

We're certainly not making any changes in a rush now, when there are so many unintended consequences to consider.

Support. I thought this system (the thing you're suggesting to replace competence) was already how it worked, obviously not, but I think it should be that way.

Prispeval Ivan, 28.08.2013 at 02:25

Prispeval Desu, 27.08.2013 at 16:03

- Season limit changed from 10 maximum, to 25 coalition wars minimum
- To create a good sample, you need the full 25 CWs to be ranked
- You can still have more coalition wars to increase ranking/percentage after the 25 are done

This will result in only 1-2 Coalitions in the final ranking. Most Coalitions are too casual play 25 CWs per season. Look at how many reached even 10 games - last season it was only 2 Coalitions.

Citiraj:

- Win percentage out of total games per season used as marker, not competence.

This will just force coalitions to pick easy targets and avoid coalitions with established winning reputation. Current system rewards challenge, for both sides - since a more powerful coalition always results in more points. In the proposed system fighting top coalitions will be too risky - and why bother when defeating newbies will get you the same result? Top coalitions will really struggle to get the required 25 games.

Citiraj:

You have a full 4 days, I can only hope changes are made.

We're certainly not making any changes in a rush now, when there are so many unintended consequences to consider.

I fully support cw games to be atleast 15.At least 15. 10 is not barely enough.Listen please,when you have 10 cw's to play total,after you play your first 5 ,you kind of start planning your cw's because you have so few left.Imagine how much great games we are missing.If we would raise the numbers of cw's we would see cw's every day,people would have motivation to train,to improove etc.Its a way to make the competitive scene bigger and better.Also like madara said after most clans play their 10 they loose interest pretty much.
And to clans that cant make 10 games,seriously Ivan why do you think "casual" clans have or should have anything to do in the competitive scene?Clans that cant make 10 cw's in a season dont deserve to be in the listings anyway.
I hope i dont sound too cocky,but people who play competitive are the backbone of this game so i think a satisfactory change in cw system should be priority.

Prispeval Ivan, 28.08.2013 at 02:25

This will result in only 1-2 Coalitions in the final ranking. Most Coalitions are too casual play 25 CWs per season. Look at how many reached even 10 games - last season it was only 2 Coalitions.

The number of coalitions have changed, more players are playing, it may be summer but there has been a massive boost in the number of coalition wars and players in coalitions. Last seasons ranking are also under the current system which makes every game 10% of your entire season. It's slow because of the system.

Prispeval Ivan, 28.08.2013 at 02:25

This will just force coalitions to pick easy targets and avoid coalitions with established winning reputation. Current system rewards challenge, for both sides - since a more powerful coalition always results in more points. In the proposed system fighting top coalitions will be too risky - and why bother when defeating newbies will get you the same result? Top coalitions will really struggle to get the required 25 games.

Weak coalitions fighting weak coalitions is how it currently is. And forcing a high number like 25(or a similiar number) won't let all the games be like that.

"since a more powerful coalition always results in more points." A powerful coalition newly made will get more points by facing established powerful coalitions. An established top ranked coalition cannot win the season. Weak coalitions only need a few lucky games to push a powerful coalition off the ranking, since every game matters so much. To make an accurate ranking you need to make each game matter less.

"In the proposed system fighting top coalitions will be too risky." Do you understand how risky it is now? In my system a CW is only worth 4% of your season, and each game will matter less the more you have, creating a more accurate percentage attached to your coalition. In the current system low coalitions usually refuse because they can't just waste one of their 10 games. I will keep on mentioning the 10%/game in the current system as many times as needed.

This idea allows all coalitions a chance at the top 1/2/3, weaker coalitions and strong coalitions. I don't mind rewarding challenge, competence can stay, but it doesn't help to rank each season by it. The current season system only rewards challenge for a limited time. Once your competence rises, there almost isn't a reason to get any better, no reason to train your fellow coalition members, nothing to push for everyday. Every game is 10% of your season. This isn't a large enough sample.

Top coalitions won't struggle to get 25 games, we have a whole three months to do it. If games actually continued to matter after the original 10, we could of gotten 50+ coalition wars very casually, 70+ if we pushed for it. Or, you know, if coalitions actually accepted and wanted to play, or if their leaders had not quit AW because their 10 games were up.

Prispeval Ivan, 28.08.2013 at 02:25

We're certainly not making any changes in a rush now, when there are so many unintended consequences to consider.

The consequences have been looked at and explained. The current system is more detrimental than any system proposed in the threads at the top, or my idea in this thread. This has been talked about through and through.

I, and many others, have basically quit AW once the CW season slows down as coalitions reach their 10 games. This should be proof enough that something needs to be changed.

Do you at least recognize that competence and max game limit is a problem stifling competition?

Prispeval notserral, 27.08.2013 at 22:59

Desu, there's a problem in your system: if CLNs are motivated to go for wins, they won't CW a CLN they know they won't win against, thus picking weaker CLNs and then high rank CLNs wouldn't get any CWs at all. With a minimum of 25 CLNs it wouldn't be much different, I don't think people would risk a loss to get to compete.

They are motivated. Again, forcing 25 games makes it so the sample is spread out, people can't face the same coalitions over and over again, they'll eventually stop.

"they won't CW a CLN they know they won't win against, thus picking weaker CLNs and then high rank CLNs wouldn't get any CWs at all."

How different is it now? You know people naturally don't want to face a stronger, better person for fear of a loss or a result where they don't gain anything. How is this different in any competition, game, medium, fight, argument, dispute, or anything that one person or persons face off against another person or persons? The only thing you have to say is competence.

However, competence that carries over from the last season has already been proven to be a terrible way to rank the current season. It would be completely fine if you reset competence every season, that would be about the same as my idea. This makes it so things are fair again. However the current system uses competence that has existed since your coalition has started, and uses this to rank you per season. It is unfair.

Prispeval notserral, 27.08.2013 at 22:59

I'm against the maximum of 10, we should be able to have unlimited CWs in a season. THEN add a nifty piece of code that makes matches against the same CLN and same players have diminishing returns (i.e. Stalins CWs some clan, where they use player A, player B and player C. if player C leaves and joins CLN Y, Stalins win against CLN Y, if player C plays, is slightly diminished; same thing with a CLN cwing the same CLN more than a certain number of times).

Then add huge SP and protocoin chance bonus, like 6 times for a CLN you didn't fight yet (and decreasing as you have more matches against that CLN).

Sounds interesting, I hope you can expand upon this type of idea and propose it to the administrators. It sounds like it'd work in any system as well. Also adding SP/protocoins is a good idea, anything to encourage coalition wars.

@khal.eesi, what you have said is also my point of view. Even changing to 15 games a season would help.


@Everyone, do you recognize that the maximum game limit and ranking by competence are bringing down competition, and slowing the rate of coalition wars? If so, do you also recognize that any change or adjustment, even rising the game limit to 15 instead of 10, would help the season system? Again, if so, where are your ideas? Almost everything is better than the current system.

An example of why the competence system is flawed - look at our battles with Mortal Kombat

Now Mortal Kombat has some good players high ranked, when they beat us they get 100 or 80cp, when we beat them, which is hard because they are good players we get 27, so we have to beat them 3 times to match their 1 win even though rank wise they are stronger than us...this makes no sense. I understand if we play a 'noobie' coalition then we should not be rewarded as much, but the current system really doesn't encourage consistently good coalitions to keep playing (just ask Dalmati)



also the same with Campfire Fellowship, a clan that includes Tophats and Acquiese, hardly 'noobie' players.

Hmm, Competence+CP is basically Elo for Coalitions, and I don't see anybody complaining about Elo. A few tweaks might be required, but we don't want to change the ranking system to pure percentage, which doesn't take into account comparable strength of the opponents.

Here are few ideas:

- Unlimited number of games per season. Coalitions can win by a sheer number of CW battles.
- Counting the last 10 games of the season, as opposed to the first 10. Coalitions can amend the screw-ups in the beginning of the season by "replacing" those early CWs.
- Reducing competence penalties.

Prispeval Ivan, 28.08.2013 at 06:08

Hmm, Competence+CP is basically Elo for Coalitions, and I don't see anybody complaining about Elo. A few tweaks might be required, but we don't want to change the ranking system to pure percentage, which doesn't take into account comparable strength of the opponents.

Here are few ideas:

- Unlimited number of games per season. Coalitions can win by a sheer number of CW battles.
- Counting the last 10 games of the season, as opposed to the first 10. Coalitions can amend the screw-ups in the beginning of the season by "replacing" those early CWs.
- Reducing competence penalties.

The reason people aren't complaining about ELO is because it isn't divided into seasons. If you had a "top ELO per season" and based it upon ELO that carries over(like competence), Tophats/Chess would never, EVER win a season. Then you'd see complaining. Good comparison, no?

Taking comparable strength of the opponents is fine, I understand you want to reward challenging higher coalitions, but this doesn't let there be a level playing field. A new coalition with strong players from previous dead coalitions will have every advantage going for them.

Yeah, I would like an unlimited number of games.

Counting the last 10 games of the season is almost the same deal, you can just win 10 then stop and hang onto your place. It doesn't encourage more(for other coalitions that want to increase their ranking as well, with one coalition stopped playing that's a number of potential CWs never to be played).

Reduced competence penalties is a step in the right direction.

Well, these ideas could be implemented quickly before the season starts and would have a huge impact on it:

* SP in a CW is boosted 5x.
* Losers in a CW keep their full SP.
* Chance of protocoins in a CW boosted 3x.
* Unlimited CWs per season.

Then we work on changes for the next one.

Good arguments, Desu.

I agree the situation of CWs and coalitions has changed a lot since we introduced the current system back in early 2012. At that time a coalition was barely able to reach 10 games at all in the same period. And the 10 games limit was set to prevent farming but nowadays it limits the opportunity to CW all the coalitions you want. Another important change is the improved balance of CWs weaker clns have learnt to avoid stronger clns so farming is not a big aspect at the moment.

Your suggestion to implement a static number of 25 games minimum requirement is not effective. This will only work for coalitions that are really active and arent denied to CW by others, like yours. And additionally it would only work in the future if CW activity rises in general otherwise the system would stagnate.

In my opinion a dynamic minimum requirement would work better here. For example: The average amount of games played (+-x ??) per coalition in the running season. With your numbers 13.5 would be average. That might motivate the less active clns to play more to get into the ranking. But in the first place it would motivate the active clans to play more games and kick the clans with 100% win ratio but minimum games out of the ranking.

I would like to see a sum of ( winratio * x ) + ( earned CP / games won ) or some other incorporation of competence. That way noob farm CLNs wouldnt end up on top every season. And new coalitions are not completely discouraged. Factors might do the trick here so the balance of winratio and competence isnt off in the end.

Edit:
If its just going to be a small change to the current system I suggest the number of games accounted could be equal to the number of participating coalitions. So that would make it 20 games right now. If there are less than 10 clns participating the number of accounted games doesnt drop below 10.

I prefer Desus suggestion. It is not perfect but it offers less limitation and more motivation for average coalitions.

Prispeval VRIL, 28.08.2013 at 07:50

Good arguments, Desu.

...

With my numbers yeah. However I only went to 14 and not the rest, all 20 coalitions this season it'd be 208/20 = 10.4 average. I set 25 since it was a high goal, if my idea was implemented I wanted to aim for that, however a minimum of 15 or 20 is fine too. I wanted something everyone can aim for. But of course my idea was static as you said, and a dynamic goal may be better, and your #of coalitions = games idea sounds alright. However when participation drops or rises it'd have to be changed to another way to calculate the dynamic goal. 10 Isn't enough and shooting higher than 30 is over even my goal.

Anyway 20 minimum sounds like the best number.

"sum of winratio + ( earned CP / games won ) or some other incorporation of competence." My goal was to eliminate any relation between the coalition stats and the way a coalition is ranked for that season. Leave the stats to rank overall on leaderboard in the coalition page. However as it is a step into correcting the broken system, it'd be far better than nothing to compromise onto something like this than leaving the system as is.

I stand by my initial plan in the top, I believe it's as perfect as we can get it. The CW minimum limit part is subjective, I want it 25, and 10/15 are too low of a sample for an accurate ranking so I guess 20 is the best compromise.

Hopefully my idea holds up.

a) Scale down 'competence'.

b) Have the season be determined by "10 best cw scores".


Even more fun would be if there were a meaningful way to combine "10 best cw scores" with "total cw points scored." Weighted heavily towards the 10 best cw scores?

For example: 10 best cw scores + (total cw points won/50)

Ivan, I know you are saying that we shouldn't be hasty in making changes, but the current system sucks pretty bad, so its not like you are risking upsetting the player base or ruining a season. I'm down to gamble. Nothing wrong with a little trial and error as long as steps are being made in the right direction.

No because scores would still be connected with previously attained competence, so again a consistent coalition would be at a disadvantage because their 10 best scores would not match the scores set by a 'weaker' coalition. I thought earlier about 10 best results, but you still have the problem of 1 win for a 'weaker' coalition being much more valuable than to a stronger one, therefore the problem is still the same. Competence is the stumbling block on all these ideas. Desu's original suggestion is the best solution so far although 25 games is ambitious.

How could a weaker clan ever beat SM?


Well, I did suggest scaling down competence. If you scale it back enough, I think it could still work.