Current rules for opponent selection in AW?

I also prioritized the fast attackers to 3/60 for mopping up teams. So I have a few teams of three 3/60 fast attackers and two 3/60 healers on the flanks. It works pretty well as it’s over 3k team power.


I want to add my assertion that AW matchmaking is a joke. First, alliance score only really attenpts to measure your best team in action. AW is about something completely different — depth of your hero roster. My Alliance is getting utterly destroyed in the second round of every war, because we’re a relatively young group of players without a bunch of powerful heroes. We’ve been up against vastly more powerful teams, especially in the second half. There’s just no way we can compete. It’s not fun.

The only saving grace from all this is that AW rewards are terrible whether you win or lose, so we don’t mind too much. But seriously, something has to change.

1 Like

The matchmaking is a joke, we just had to learn this again this round. What bother me is that the alliance war is arbitrary in a certain way. You win, you lose … who cares? The titan fights are present with 4 numbers in your player card as well as the alliance score. The alliance war doesn’t show up anywhere and if you don’t take a screenshot it is just gone…

You are explaining that your alliance is poorly suited for wars because you lack a deep bench of heroes. The solution is to develop a deep bench of heroes. That’s the point of alliance wars.

1 Like

My alt is in an alliance that graced top 100 months and months ago and is currently fighting 5* titans. Most of our wars have been relatively close. This is with not evreyone participating on either side and me sending unlevelled 2*s and trainer heros into battle.
I’m having a blast with it. I totally expect us to get curb stomped at some point as even our last opponent has much stronger teams

I think you missed the point. Obviously my alliance need to flesh out its roster. The point is we are being matchmade against teams who have already built up deep rosters of powerful heroes. We lost the most recent war by almost 800 points. This thread is discussing how they match teams up, and how their algorithm is sorely lacking.

Their algorithm is fine. Alliances are matched based on their alliance score, which is how they are ranked on the leaderboard. When one alliance has much deeper benches than the other, it will win easily. The solution is for alliances with weaker benches to hurry up and improve their benches.

Or don’t bother. There is no cosmic rule that says you must be good at wars. If you want to direct your efforts elsewhere (levelling your top teams, for example, to make your members better at raiding or Titans) then go ahead - but that’s your decision and you’ll keep losing wars.

Want easier matches against weaker alliances? They’re for alliances that are not as good as yours, with lower alliance scores. It would be wrong for you to prey on them, just because your members haven’t fixed their benches.

1 Like

Their algoritm sucks big time. Imagine this we are 30, they are 21, the same alliance score. This was by far the most unbalanced war till now. Sorry it was not a war it was extermination.

Then you need to get better. Complaining that you just got beaten up, when you know exactly why it happened and how to fix it, undermines the case for re-examining the matching algorithm.

The whole point of alliance wars is to provide an incentive for players to level their bench of heroes, to give value to the disliked 5s, the neglected 4s, and even the oft-ignored 3*s. If matching is changed so that it is based on, for example, aggregate power of the top 30 heroes for all players, then that incentive immediately disappears.

Worse: I think it would introduce a new incentive in diametric opposition. I think the optimal alliance under this new approach would have a barbell shape - lots of fully ascended and levelled heroes, lots of low level trash to clean up, and as little as possible in the middle. This would very effectively undo what SG seem to have been trying to achieve.

1 Like

I am ok with what war means in this game. We definetly working hard to make a better bench, but the matchmaking is hilarious. I repeat 30vs 21. Do you know what that means when the aliance score is the same. Only 5 guys from our team are better than the poorest from them. You think that is a fair matching?

1 Like

No, their algorithm does indeed suck because Alliance Score is a meaningless metric for war performance… and frankly Trophies measure nothing relevant at all except for those that want bragging rights for whatever reason.

This isn’t coming from a “oh woe is me we lost” vantage point, quite the opposite in that I haven’t lost a war yet on any of my accounts that I care about. That still doesn’t change the fact that it’s broken in the top ranks, it’s badly broken in the middle ranks, and it’s laughable in the low ranks. I have the perspective of seeing wars in the top 5, in the top 100, and another which is way way way down the ranking list at 10k score with the test alliance: this isn’t idle hyperbole from a limited perspective.

The top ranks will get fixed in time, maybe, as every alliance will trend towards having members with 30+ maxed heroes, but that still neglects the rank and file where the vast majority of the players are, and frankly we deserve better than this.


If it’s 30 vs 21 that means you get 54 more attacks than they do.

54 more attacks than they do.

Who do you think that is unfair to?

1 Like

Forget it. Definetly your imagination doesn’t work so good. You are acting like a Russian propaganda channel on this forum.


If you’re winning all your wars then either:

  1. Your players have well-levelled benches of strong heroes. Good! You should be winning. Or…

  2. Your alliance score is lower than it should be, giving you matches that are too easy. This is either because you are bad at killing Titans, or you are bad at raiding (or just don’t care).

If this situation bothers you then there are obvious solutions. You could eat some of your bench heroes (seems silly, but hey). Or you could improve your alliance score to where you are getting competitive matches, either by getting better at Titan killing or by making sure your members are all putting a decent effort into raiding.

Sorry, but that doesn’t make any sense. You are saying it is okay to have grossly mismatched alliances pitted against each other in Wars? That is obviously not the intention of the game developers, and would be an awful design if so. They attempted to make Wars competitive by matching teams with similar alliance score (Otherwise, why not just make it completely random?) The point here is that is a terrible way to measure alliance strength for this new gameplay mechanic.

They could measure teams’ relative strength by adding up each player’s strongest 30 heroes, and matchmake based on that. Or they could track a new War Score per alliance, similar to the titan score. Or they could track war results and match teams with similar win-loss records. There are several alternatives that would work better. Hopefully by pointing out the problems and offering solutions we can help make the game more enjoyable. Claiming everything is fine and players just need to stop sucking is not accurate or helpful.

So I told my experience in this war. I have a team with 2800 my best, next one 2000, after that only 3* heroes not leveled. My target have 3500. I wipe out only 2 heroes because there are 2 maxed 4* healers on every team from the enemy. So like this was the experience for another 20 of my teammates. Yes, definetly a easy war for us. 54 attacks extra… Can they give us 100 and is the same. We are 30 children playing football with a champions league team. And you said that we are favorites.

Summing top 30 heroes is a bad idea. I’ve not come to this conclusion lightly: for a while I thought it was the best solution we had. But it absolutely disincentivises levelling third and fourth team heroes, which is one of the best things about the alliance wars. (In fact, I think it might incentivise not levelling those heroes).

Tracking war results using an Elo sort of system (like raiding) is the best idea in theory, one I initially advocated. But it was immediately (correctly) pointed out to me that war results would not provide enough data points for this system to work, plus changing alliance composition would really mess it up.

Alliance score is not an ideal way to generate matches, we agree, but it does provide a built in compensation mechanism for alliances who don’t have great top teams, but have strong second tier heroes. While they won’t thrive at killing Titans or raiding, they ought to clean up at alliance wars. And as noted, this matching mechanism actively encourages players to level their benches, one of the great pluses of this new game aspect.

You have a terrible bench. You should lose at alliance wars. When you fix your bench you will start winning.

Yes I know, I start playing this game a month ago, many teammates are like me. But why our oponent is match with a young alliance like us. I told you: bad matchmaking system. Alliance score works well when the alliances have the same numbers of players. Maybe 3 players +/-. But when we are talking about +/-10 or more you will find wars like we have. Old, experimented alliances with 20 good players are matched up with new alliances with 30 beginners. But is OK the matchmaking system woks fine in your opinion. Maybe up in the leaderboards wars are fair but here where the most players are is a mess. This is by far the worst war implementation in a rpg game I’ve played till now. How to match alliances with different numbers of players(30% diference in our case)?

If a full alliance gets matched against a really short handed one and still can’t win, then their benches must be terrible. You have already told us that your benches are terrible. Fix your benches.