Aliance Wars | Matchmaking Problems.... in Kiril's name fix this already!

This message is for the DEV TEAM!

How hard can it be to make a matchmaking system work??? You just made changes and it still doesn’t work.

  1. in matchmaking rules, the first one is easiest to solve, and that is the number of members from an alliance to be equal tot the other alliance… how hard is this??? is just a simple if x = y then WAR else check again… not hard right? (look at the matchmaking we have as an example…)

  2. Alliance score should not to be taken into account or heroes that are displayed as defense teams… this is something that can be mess with and a lot of teams already do this. They lower the score and change the teams in the day of the matchmaking so they trick your amazing matchmaking system :money_mouth_face:.

  3. What should be taken into account is:
    a. user’s level that cannot be changed
    b. the number of players in an alliance
    c. the power of the first 10 heroes he has in his hero list so the display hero trick won’t work anymore.

Make a score between the number of players, the sum of the players lvl and most important the power of the first 10 heroes… an that is matchmaking…

You are overthinking something easy.

Don’t tell me you cannot do it… it is a Client -> Server system you have access to all the information from our accounts so do it already…

Fix the wars or take them out, the system should be fair not rigged.

Keep in mind we won the last 9 wars out of 10 but just because the players from the other team didn’t use all their flags. All the matchmaking was against us.

Hope someone is reading this and has an response.

Have a great day!

2 Likes

Do you know how many of your adversary opted out? Last war our alliance of 30 was set against an alliance of 27, as 3 of us opted out. We were same number in battle screen

last time we were 28 and the team we were matched against were 29 the score difference was huge and also the power. Not to mention that team had 10 players over lvl 35 with heroes of 1* or 2* in their defense team, to lower the score.

The true power of the teams were all over 3400TP…

Now lets take this matchmaking.

If you look in the game and check out both alliances you will see is not only the number of players or the score, is the power. more then half of the alliance that we have been matched against are over 35 in level with more than 250 days of game play. This will result in a higher number of heroes. This must me taken into account.

I’m not crazy the matchmaking is bugged hard and it is the way the algorithm works.

If you are trying to design a system that produces closely matched wars, then yours is a terrible suggestion that would likely produce more mismatches than the current (flawed) system. If you are just proposing an arbitrary set of matching criteria because you happen to like them, then I congratulate you on your effort, but many better approaches are available. They have already been discussed in the forum at great length.

That depends entirely what you mean by “work”. It’s easy to make a matchmaking system that randomly pairs alliances. Does that “work”? No? So you probably mean that to “work”, a matchmaking system should produce close battles.

But not every battle should be close, right? Sometimes one team will be better organised than the other, or luckier, or have more team members sober. So even a matchmaking system that “works” will not always produce close battles.

Does the current system “work”? What proportion of wars are close? Oh, you don’t know? Me neither. So what makes you think that the current system is not working? Oh - your own experience over the last ten wars, of course:

Ten wars is a very small sample size, frankly, but it’s not worthless. If you have had ten consecutive mismatches wherein you were slaughtered by opposing alliances that vastly overpowered you, then maybe that does provide a little bit of evidence that something isn’t working.

I beg your pardon?

[Bangs head on desk. Bangs head on desk. Bangs head on desk.]

So you are saying that AW matchmaking is so flawed, that you were so consistently mismatched to your disadvantage, that you won nine of your last ten wars “because the players from the other team didn’t use all their flags”. What an amazing way to totally undermine your own case.


Developing matchmaking systems for alliance wars that will consistently generate close battles is an insanely difficult problem, as we see whenever players comes to the forum with their ‘simple,’, ‘obvious’ solutions, all of which have glaring failings and would generate worse mismatches. (Close wars need not necessarily be the primary matching objective, of course, but the weight of opinion seems to support that as a goal.)

I don’t like the current system and advocated retaining a previous status quo or developing some sort of AW Elo-type system as an alternative. But as players we have no way of assessing whether matchmaking is generating close wars or not - our own experience is too tiny a sample, and our own alliance might have issues that make it a matching outlier. SG can review war results. They seem to have been doing this over time, and to have been tweaking the matchmaking process. I assume based on staff comments that they have been making iterative improvements.

I would not have chosen the current matchmaking system, as I explained ad nauseam before it was introduced. But the hatred directed towards it seems unwarranted, especially when justified by a player who observes that their alliance is so consistently overmatched that it has won nine out of its last ten wars.

2 Likes

Bad update! Our aliance have this problem too!

Those are proposals of a system that the user cannot modify in the alliance advantage.

All the things they use for matchmaking can be very easily modified by alliances players. like putting teams of 1* and 2* stars heroes in defense teams to lower the score.

Is not something that i like was just a proposal.

Further more… the fact that we use all our flags as an alliance rule it’s easier for us, we are all from the same country and this is an advantage that help us win wars. but the matchmaking was flawed.

if 2 players from the other alliance would have attacked we would have lost most of the wars. the difference was under 100 points all the time.

Keep in mind we were 28,29,30 and only won by at most 100 points… but the other team had more than 5 players that did not attack.

Last war we lost… you know why? this beautiful matchmaking system matched us with a team that had 10 players that had in their defense team only 2* heroes. All the players were over LVL 35.

Is the matchmaking system flawed… for sure. If the user can modify stuff that affect the matchmaking it is flawed.

Right?

Sure, but so what? There are many things a user cannot modify: just because a matchmaking system uses metrics that cannot be modified does not mean that it will generate close matches. (I assume your are trying to achieve close matches.) We have no reason to suppose that your proposal would be better than the status quo, and I can think of many reasons that it might be worse.

You have just described very close matches. If you want them to be closer than that then I think your expectations are totally unrealistic.

That might just be what we call a ‘clever strategy’ - it sounds like your opponents successfully gamed the system and you didn’t. I wouldn’t worry about it: SG have already demonstrated that when that sort of approach becomes widespread, they change the system to negate its effectiveness.

Loool… i’m sorry that is not a “strategy” is cheating. if this is how you think i know for sure you probably do the same :wink:

We have not used that strategy, but if you think it’s “cheating” then you don’t understand what “cheating” is.

the matches were close because the other team were sure of a win… the fact we use all our flags all the time is a surprise for “these clever strategy” teams

The matches were so close that just by using all your flags, and your opponent having a handful of inactive players, you won by about 100 points. That is very, very close.

Again, if you expect anything closer than that then you’re kidding yourself.

lets see if a Dev will answer, i doubt it but maybe we are in luck. And see what they think of the clever strategy and other stuff.

By the way great talk @Brobb. We are all allowed to have opinions and these opinions are based on game events and is nice to see other opinions :+1:

1 Like

I’d be surprised if we hear from a dev, but you never know - sometimes they pop in. Go with my best wishes.

1 Like

Seems like smrth went wrong after last updates. Second war in a row we are getting much more strong opponent. We didn,t have a chance last week, their defence teems were much stronger. Now we got a similar opponent, besides one of our strongest players took of the war mark. Maybe several opponent players did the same, We’ll see tomorrow)

The matchmaking not working my alliance still getting much more strong opponent. The opponent (each player) 3000 up power 12 player, my team 3000 power only 4 or 5 player -_-

Same thing last war, our average defense team power was almost 200 points less with same number of players. All our teams had the strongest heroes in stacks. Thats noncense, no chance to win.

Sorry man. Your post displays an almost comical lack of familiarity with this issue, both of actual statements from the devs and the overall community sentiment about matchmaking. Almost nothing in your post that you state as fact is correct. Almost none of the suggestions you make are new, many have already been implemented, the rest would not help.

The current matchmaking DOES take this into account. It simply doesn’t limit it to an exact match. Alliances within 1 member can be matched.

Alliance score is not taken into account. Congratulations. This has been true for many months.

?? really?? what does level have to do with anything? The correlation between strength of team and player level is so loose as to be useless. This would also encourage people to try to level more slowly so as not to outpace their team strength. Why would anyone want this?

Again, it is. Yay.

They have done you one better here. The strength of the top 30(!) heroes is taken into account. The displayed heroes in a defense is not (and has never been) considered. There is no trick. The reason to game the system and put weak defenders into your defense team is to drop cups and have easier raids. It is entirely unrelated to wars and does not affect match-ups.

You seem to be basing your post on a match-up that from your screenshots looks very close and totally winnable. Your alliance generally is well organized and wins most wars. I can’t see any reason for you to be complaining. It’s weird. As @Brobb states, the fact that you consistently win is such solid evidence that the rest of your post is poorly thought out that it probably doesn’t deserve much more of a response.

I’m sorry that this post is somewhat confrontational. Please do not take it personally. Familiarize yourself with how matchmaking has actually been designed by reviewing the build change logs and various developer responses before your next complaint and I promise you will find a better reception.

Best of luck in your war.

3 Likes

Sounds like your win streak has lead to your current match among many other factors.

Surely you do not have to change anything?

players team 1 = 7
players team 2 = 27

trophies team 1 = 12918
trophies team 2 = 33848

One team can face another that has 2.62 times more trophies, and that has 4 times more members than this one?

If you think that is fair, or this can not be fixed, or you are part of one of the teams that always benefits from this situation, or is that you have a serious problem of interpretation of justice.

Dev, please do something.

How deep are the benches of team 2? What were team 2’s last five war outcomes? How well organised is team 2?

What’s that? You don’t know? So what are you basing your assertion of imbalance on?

Just the things that you can see most easily. Well cool then.

Edit: I can’t wait to see screenshots from the actual battlefield of this war, for obvious reasons.

Cookie Settings