Alliance Wars Matchmaking (Discussion & Developer Response) MASTER

Can individual players be tracked with their own ELO score?

And ignore my previous comment about being lambs to the slaughter because they were 7k higher… we actually won by 300.

We were a whopping 250 points lower in our last war (based on average defense team power) - we lost (as mentioned) by less than 500 so it was a moral victory of sorts, but still, we’re tired or losing by 500+ every war.

1 Like

Still lost by 800, and they had 3 less players and 2 more non participants (3-1) including their leader whom was faaaaaar superior to anyone we have.
Posted too early, their other players showed and ran it up to a 1400 point disappointment again.

Different war same story:


Their top 3 higher than our highest (3702). Their high 4001.
35-3900 them 6 us 2
34-3500 them 12 us 9
32-3500 them 5 us 9
Under 3209 them 6 us 9
Avg team piwer: 3375 to our 3215.
At every position their team had more power from #1-29.
Same as every other top 30 war:
Opponent goes up 500 points early. We rally and get close in the middle. They finish strong and win.
At one point we out attacked them 144 to 95 and had the lead but it didn’t last. We finished with 149 attacks them with 126. If they had equaled our attack number they would have blown us out of the water.

I’m sure the devs wont count this as a blowout loss and say it was a good matchup. But again every top 30 match has been this way.
We are always at a disadvantage in terms of top power and overall power. There is no indication our benches are stronger and make up for the difference, certainly not enough to overcome their better defensive team strength.
Our alliance score is always 15k lower than opponents.
We do not expect to win every war.
We shouldn’t have to expect to lose every war either but that is now where we are at.

1 Like

That is why I said you base it on the opponents you fight.

As I said, calculate the score for each player. If a bunch of strong players leave an alliance to form a new one they will carry the score from their old alliance over to the new one. So the new alliance will have a score that kind of reflects their strength.

Think i know how teams are always so much more powerful than ours, easy to game the 30 man roster. Keep top 25 4* and 5* and 3s to fill 5 teams , then have unleveled 1*s in the other spots, or nothing. Lowers overall dramatically and screws honest alliances over.

1 Like

3800-2100. I have multiple people ready to quit. Yeesh.

@AlexMex. Just disagree strongly here.

Individual alliance war score is not a direct indication of that individual’s contribution to the alliance’s war performance. They may be correlated but different war strategies will result in vastly differing scores. Sometimes taking fewer points as an individual is the right move for the alliance. Hero strength (which they are currently trying and we are complaining about) is a much better indication.

I very much prefer a system that awards alliances with moves up a ladder based on team performance, not a sum of member performances. I honestly don’t care that new alliances will have to find their way. I doubt that small groups of high level players make up a significant portion of the 1000’s of alliances. But I do trust that SG is thinking about this problem and trying to solve it.

@mhaltu did say they were considering a combination of hero strength and alliance war performance which I think is more complicated than necessary but also think would work very well. I hope they do it when they have time to get it right.

I think we’ll have to agree to disagree. My alliance has had a number of close matches but that’s really besides the point. SG has the data for all alliance wars. Our sample sizes of a handful of wars simply isn’t nearly enough to make sweeping judgments about what does and doesn’t work.

I do think we should absolutely complain if we don’t like the outcomes. Our perspective as players is valuable and valid and should be shared. I hope and believe that SG is paying attention to us. But we should also understand that there are limitations to our vision and have at least some faith that they are bringing their perspective to the table when trying to solve this.

I also honestly believe that losing by 1000 is about the limit to what could be attributed solely to a bad match-up. Participation rate, individual war/raid skill, team strategy & coordination, and even luck of the draw with boards all can turn what looks like a good match int a blowout and vice versa (we have either won or come close against a number fo alliances that looked significantly stronger than us).

Do we know yet what match up method will be used in this weekend’s war?

1 Like

I’m still trying to digest your method and don’t pretend to understand it. Where I get confused how it would work is in my own performance. The part I enjoy most about the war is directing the battlefield. I carefully watch who is coming online. I know generally their team strengths, I like to throw some kill shots at new players so they enjoy the rewards as well. Sometimes I’m at the bottom of the leaderboard because I’ve given away my kill shots. How could that be calculated as part of my own performance? I get a LOT of thank yous from our players as they are directed to appropriate targets. Maybe I could get a medal for that :smile:

3 Likes

Why I think matchmaking based on the strength of heroes is not a good idea:

  1. Matchmaking based on the strength of heroes can not take into account that there are players who simply choose not to participate in a war. (One of my accounts is in an alliance where our strongest player does not like the war and he chooses not to fight. He’s a great guy and we do not want to boot him but right now, his heroes are counted towards the matchmaking but they do not participate in the war.)

  2. If you want do matchmaking based on the strength of heroes, you have to incorporate in the matchmaking algorithm that certain heroes are simply stronger then others. But also keep in mind that SGG keeps changing heroes, so every change in heroes would affect the matchmaking.

  3. If you do matchmaking based on the strength of heroes, you also have to take the strength of troop in account. I’m not sure if that is done now. You must admit that (for example) level 20 4* troops are not comparable to level 7 3* troops.

  4. There are players out there that are simply better then others. That might be because they do not play the game as long, because they do not understand the game as good as others or because they simply ‘miss the combo’s’ that can be made during battle. If players are below average they will always get opponents that beat them. I think it would be better for everyone if they get a little ‘softer’ opponents. That way they might win some wars and they will stay interested.

1 Like

The killshot gives a player a lot of points, but I do not think those points should count in you war-score. In my opinion this war-score should not be a direct copy from the points on the war leaderboard.

This is what I meant by:

I have had a lot of wars where I deliberately take on stronger opponents. I know I will probably not kill his whole team, but if I take out two or three heroes, an alliance member who has few ‘good’ heroes left can then finish the opponent.

Latest mach making mechanism was just perfect, heavy fighting till the end. Not sure which setup was used, but please do not change anything anymore! :slight_smile:

One good match doesn’t mean it’s working, even a broken clock is right twice a day

3 Likes

Thank you fro the clarifications

I did not say that we have to take AW score as is. In my alliance we put in place a team strategy that is definitively more efficient than individual one. Stronger teammates always attack the stronger members of other alliance. Sometimes it’s a complete success (I consider 5 win as complete success) sometimes not. But we open the path to weaker teammates. On last wars we got at least 20 teammates above 100 points . Before applying this strategy our score was always between 3500 and 4200 and fighting team with bigger alliance score were often synonym of defeat. Now we are always above 4500, we are taking down other alliances one time (with bigger alliance score, a highger sum of levels / powers, more people than us, we were only 27/28 and this is affecting hardly our chances/score) and now AW are really funny.
After attacking a 4100 team and only killing one or two heroes teammate receives a low AW score and the one who finish get a good AW score. That’s a good improvement for the team as weaker attacks can now rewards 30/50 points instead of classical 4/9 points. But in the individual score perspective that should be exactly at reverse. Also this indicator has not to be published. It would be used only to calculate alliance score.

If SG can’t make a good balance they should remove it from the game. We have only members that play the game for 2 months, we are all below lvl 30 almost all below 27 and our opponent has 15 lvl30+ members. They have much more time and much more heroes for second, third, fourrh teams. We don’t have a chance again.

1 Like

We got an alliance we fought and lost to in the last 20 minutes 4 or 5 wars ago (by about 150 points). FINALLY a good match!

Not us by a long shot. They are over 30k alliance power stronger. They have 26 teams over 3000 power, we have 16. Its over before it starts :disappointed_relieved: It would be fun for a change if the bad match ups went in our favour.