Alliance Wars Matchmaking (Discussion & Developer Response) MASTER

Good catch; I misread that part.

We have started a war…enemy have an extra player. And the points that player worth are almoust simular… Becouse they have an extra player,they have six more flags??? Please explain??



It happens from time to time. We’ve had it twice against us and once for us.

We won all 3 so whilst it’s an annoying thing the match isn’t over yet. Heads down and give it your best shot.

1 Like

Yep, it’s just bad luck. The matching process can’t pair everyone equally if the numbers don’t allow it. Whilst it makes it harder, it’s still possible to win

Part of the war matchmaking algorithm is based on war score. As some have mentioned on the forums, matchmaking prioritises war score over the same number of participants (which at last check still allows a difference of up to three).

As the others mention, it is still possible to win. Only had it happen once to the alliance I’m in, but it is possible to win.

I went through a lengthy break down of the variables when it happened to us if that’s of any interest.

@ThePhilosopher

War score, yes. But of the individual players.

I think there’s one piece still missing to make these matchups fairer. If they don’t use the Alliance Strength (as used for Alliance Rankings) or use a Players Strength (same), then somehow another variable could be used to penalize a Players War Score.

Players can still “tank” (lose) to lower their War Score. A team of such players could do just that. I believe we just ran into just such a team. Team and Player strengths unbelievably high. Should have been a wipe, but somehow they managed to win only by a comfortable margin. What I noticed is that some of their players, who should have done much better, performed only weak and losing attacks, hence lowering their war score.

Simple solution, if not in already in place, would be to penalize a Players War Score based their team’s wins vs. losses. High win ratio of the team increases an individual’s war score regardless of their personal performance.

Equally matched teams, which the current algorithm is presumably meant to produce, should give teams a roughly 50/50 chance to win. I’d be interested to know if all teams are close to that overall goal.

Not sure, currently, how someone could deliberately lower their war score except for losing war on purpose. Or stripping their top 5 of all emblems to lower their roster strength which forms the basis of an individual’s base war score. Still unsure whether a war score can go lower than the base value of the individual’s roster.

I don’t think any ‘solution’ will satisfy all players. I’ve seen a ton of proposed solutions but all have their flaws.

That’s how it works. Not sure what else it could do. War score doesn’t track with individual performance in a war. It tracks the win/loss of participated war regardless of performance as far as my tracking goes.

2 Likes

I suspect we are just matched with an alliance who manipulates their war score by losing some wars on purpose.
We are warring 9 against 9. Their war score was 31300 and ours 31306. These numbers were valid before start of the war. Their war score went up by 3k right after them using first 9 flags, our score remains unchanged after using first 4 flags. The 3k war score gain shall reflect current war performance. I don’t think it has something together with emblems or leveling heroes as they are barely online and 3k is just a huge gain. They are an old alliance and every of their members is within this alliance from 360 to 1380 days.
I read this as players of opposing team who used first 9 flags had terrible to zero performance in previous war. Is this correct or am I missing something?

How many people are in the alliance? You say that 9 are participating in the war. It is possible that those who aren’t participating are opting back in, so the alliance war score will go back up; however, it doesn’t change the field. As for losing wars on purpose to lower their war score, I doubt it, it’s pretty counter productive to purposely lose wars so that you can win, when matchups over time will do that for you. If they had a losing war record, it’s possible they just lost them.

As far as I know, individual war performance has no bearing on matchups. The alliance war score is the sum of all the members’ individual war score. The war score is based on top 30 heroes and top 5 troops, plus the war win-loss record (not battle win-loss record).

2 Likes

They are 12 and yes, this could explain this huge gain. They have 9 on the field and 3 remaining. Out of that 2 of them were not online for a day and 1 was online 3 hours ago. This means the one who was online could opt-in for the next war. Ok, this could be it.

Sorry if this has been addressed before, I couldn’t find it. I understand (mostly) the concept of war match putting two alliances together that have similar war scores. My concern is how an alliance (mine) can continually be matched against bigger teams. I run a training alliance where we have level 17 through 84, and all of our players participate in war. We have a team at 1976 team power, one at 3037, and one at 3384 (out of 16 teams). Our opponents lowest two teams had 4309 and 4392 team power. We have one team over 5000, they had 2. Our three training teams don’t stand a snowball’s chance in a confrontation like this. Our last four wars have been very similar. Can someone clue me in if this is part of the matchmaking process. (Our three trainees can’t possibly have a strong war rating, or, I could be wrong)

Do you win 50% of your wars? The system is really designed more to try to keep you at about a 50% win rate than it is to match teams evenly based on power. The previous war history has a big impact. For full alliances it usually is about ±2000 pts per war based on if you win or lose(unless you hit the war cap which is the max score where your score won’t go up with a win).

You’ve essentially answered your own question. War matchmaking is mostly preoccupied with matching war scores. TP of the defence has very little to do with it. I write ‘very little’ because the choice for defence might not be representative of the player’s bench depth which is a factor in their war score contribution.

As ffphier states, the algorithm is essentially such that it keeps us at a 50/50 win/loss. Alliances would need to have tactics and indepth strategies to overcome this particular aspect.

1 Like

Thank you for your response. I did not know about the win rate control.

Thank you for your response, I appreciate anyone that can teach me something.

Just to add a bit more info, as it hasn’t been brought up for a long time. As far as I’m aware this is accurate:

War score is based on:

  • Top 30 heroes
  • Top 5 heroes (weighted)
  • Top troops
  • Total number of participating players
  • Win/loss record

There are some common misconceptions due to how matchmaking works:

  • Alliance score does not contribute to matchmaking
  • Titan score does not contribute to matchmaking
  • Raid trophies do not contribute to matchmaking
  • Matchmaking prioritises war score matching
  • Matchmaking allows up to a difference of three players
  • Points are health based, not TP based
1 Like

This war looks like a terrible matching… :grimacing:


4k difference in the war score and almost 40k in total score. :joy:

How many on the field?

Seeing your alliance was at 29 members I’ll guess someone opted in after matchmaking. Last I checked my figures someone opting in/out can make a difference of 2500 (I think). Though a player’s war score is dependent on their bench.

As much as this is an observation, I would try not to let it dampen your spirits. Give 'em your best shot! :slightly_smiling_face:

Does your next war opponent change depending on by how much you defeated he previous alliance at all? It just seems like if we beat a team by 1k, our next opponent tends to be tough.