Alliance Wars Matchmaking (Discussion & Developer Response) MASTER

@LooSeefur - yes, they use each participants 30 strongest heroes to calculate the alliance score used for war matching. More weight is assigned to the top 5 strongest of each team.

The regular alliance score has nothing to do with matching anymore. It’s all a backend calculation based on strongest heroes. Since an unleveled 5* has a higher hero count than a fully/partially leveled hero of another * rating, the unleveled 5*s are often used in calculations.

3 Likes

If you care about AW, one small suggestion along the theme of the recent chats is don’t take 5* to 3/70 unless you are planning on 4/80. My 2/60 Thorne is only 52 power lower than my 4/70 Hu Tao…and it is well established that a 3/70 5* despite being higher power is less effective than a 4/70 4*.

1 Like

This is why they are changing hero power scores, to alleviate some of that mess.

It’s senseless to not be able to level heroes because of a poor matchmaking metric.

1 Like

Yeah, that’s really, really stupid. I have to level guys I don’t really want to in order to have any chance to get a decent matchup? This whole this is just completely non-sensical. We are ranked 618th. We just got smoked by an alliance sitting just outside the top 100. I cannot concieve of any matchmaking system that could possibly believe that would be a good fight.

1 Like

They are? First I’ve heard of this. People have asked for that change, but people have asked for a lot of changes.

1 Like
4 Likes

We just always anticipate our opponent being at least 12k stronger than us. What gets me is the matchmaking is so lopsided. Why isn’t it happening to alliances over the pond? We have not once had a war where we were the stronger team and we’ve been in 60 wars so far.

It can’t be random when its your team thats always getting the shaft

2 Likes

It is not random, no. I hope no one suggested it is.

Whatever factors make the matching algorithm overestimate your alliance, are to some degree persistent.

Those factors may be unlevelled 5* and 4* heroes … or low participation (without opt-out) … or other factors the algorithm either doesn’t account for or gets wrong. But there’s no reason to believe the same factors won’t be present for the next matching.

1 Like

Are you looking at alliance scores? First, 12k is a very small difference. Second, those tell you how active the alliance is on raiding and titans. It’s not how SGG matches for way—they instead measure the top 30 heroes’ power.

Im not saying the matches are great, but at least make sure you are maximizing your chances of success. Get everyone to use all their hits. Sort out roles: one-shots, tank-breaker, clean-up. Have a war leader or two help sort people. Coordinate defenses. There’s lot of more detailed advice on winning wars on these forums. It helps to be adept at raiding, and there’s good advice on how to do that kicking around, too.

1 Like

I’m not sure I’d say 12k is a small difference. Like I said earlier, we’re in the mid-120k and rank about 610 overall. We fight Russian alliances in the mid to upper 130k range that are flirting with the top 100. It takes a lot of heros (and a lot of money) to maintain that ranking up there. Ours hasn’t really changed in probably 6 or 8 months. Taking those alliances down is extremely difficult. Could we beat an alliance that much more powerful? Maybe, but we’ve got an alliance full of people across the world that have a lot to do. We work, have kids, activities, vacations. Planning a complex war narrative that 30 busy individuals can all come together on is extremely difficult, at best. This is why the game desperately needs a system based on past war performance. There are surely other alliances out there like ours, they probably score pretty much the same range of war point values on average. Why can’t we fight those alliances? I think that’s all anyone is asking.

We were just matched with a 3 day old alliance that had a score of 55k to our 101k. They were so much stronger than us they beat us by 300 points with about 60 percent participation. Hopefully the hero power change improves matchmaking.

Hello,

I’m in an alliance that has a few hard-hitting members, but the power levels are mostly in the 2K range. Lately, we’ve been fighting teams with power levels mostly in the 3K range during wars. We’re using about the same amount of energy, but due to the huge power gap, we’re getting beaten by 500+ points! Like Tyler above, I hope that the hero power change will change things, but I think there needs to be a compromise between looking at the top 30 heroes of each member, weighting the top 5 most, and paying attention to the alliance scores (we’ve been down 10K+ on scores iirc).

Anyways, this was a thought to improve war matchmaking:

  1. Take the powers of the top 30 heroes of each member participating in wars. Weigh both the top 5 and bottom 5 of the top 30 more than the rest. This addresses not just the power on top, but also a strong taper in the back (if it exists), which often happens in developing hero rosters.
  2. As before, factor in number of members participating. Inactive members, as is current policy, will be automatically excluded from wars. (Specify with numbers for days/wars/etc pls)
  3. Winning alliances will face other winning alliances with similar matchmaking scores. Losing alliances will face other losing alliances with similar matchmaking scores. The cycle continues.

What are your thoughts on this process? I realize that there will be a maintenance break over this weekend, but there’s a real war drought going on, and it’s a bit disheartening, to say the least.

1 Like

We just faced an alliance 50k below us in score that overpowered us, alliance score means nothing. I think they need to start the AW ladder using the new power ratings as a starting point.

I agree. That was a quiet change but much appreciated.

@mhalttu

May I make one final suggestion.

Would it be possible to also put a total flags used out of possible flags.

It can go onto the screen where you see the individual war hits.

Something like Blue font for you and red font for the enemy:
150/180 War Hits 130/174
*Example

3 Likes

Based on what I read, if indeed the alliance was new, it wouldn’t really have a significant titan score attached to it, which means that most of the score would likely be dedicated to trophies. I have no idea what the strongest heroes were for the opposing alliance, but if it was a team power mismatch despite top 30, I already gave a detailed suggestion in the latter half of my response beyond just trophy score and acknowledged the point you’re making about power adjustments.

I understand that alliance score might be misleading depending on circumstance, but I believe a good amount of people to be trying the hardest they can by not manipulating the scores to their favor. Additionally, the hero powers (and likely strategizing power) would be reflected by the amount of trophies hard workers attain. I do think using heroes to aid matchmaking is a good option, but just looking at the 5 strongest of 30, even with the power adjustment, doesn’t necessarily represent each player’s teams well.

It was mismatched like this at 3440 I have the 5th or 6th strongest team in the alliance, I would have been 25th or 26th in theirs. The matching metric weights unleveled heroes too high. We are consistently matched with stronger alliances and this is the reason why. Now they are changing the hero power in a different way which I am skeptical about because it will leave a gap between the heroes at the 3 70 and 4 70 level. They are doing this to preserve numbers.

Ok so how are they doing match up by alliance score and ? Else over all team power

Top 30 heroes with extra weight given to the 5 strongest.

1 Like

Tkx tyler so this only effects top 5 players with big hitters