Alliance Wars Matchmaking (Discussion & Developer Response)


#2014

12? I see that each extra ally adds 6 hits, but how do you get to 12? If the member counts are 30 to 29, then there are 180 vs 174 hits.


#2015

!

I disagree. No, you cannot play effectively when there is a large discrepancy between teams. Unless you have multiple 3000+ teams to attack a single 3500+ team, it is not going down. Now think about trying that with 1500 power teams. There is tile luck at play, but even if you get the luckiest tiles and colour stack against a particular hero, 1500 power teams are not going to do anything significant against 2000 power teams.

From the screenshots you can see the opponent has a higher War Score (hard to say how significant).


#2016


Apparently I can only upload one picture per post.


#2017

Not exactely an easy match up.

Please check how pairings are done this could save you and will save me some writing.

Of course you can reset them it just costs you more flags than them.

Level says not much about bench depht because relaxed players who have played for a long time my have a high level but a weak bench. Seen in my own alliance.

You had a loosing streak of 3 wars ( who doesent with the exeption of the really top teams ) and before that you went 4-4. Nothing too unusual there.
Winning against inactive teams might be boreing but cant be avoided and is the opponents mistake.

Please include your alliances and your opponents war score next time.

So overall a challenging pairing but still within possibilities if you look at all the non calculable factors that are involved.

Hope you have better luck next time.


#2018

That’s a fairly large difference in the war scores, but I wonder if SGG should increase the weight on the top 5 heroes (presumably representative of the defense team)?


#2019

Check again now because I did add them and you can add some comments.

The issue isn’t the loosing streak, it’s the fact we had no chance to win. Sure, we can’t tell bench strength, but I can guess. When one single player is taking out our top 3 players on their first 3 attacks, I’m going to say that is pretty significant when the best that we could perhaps do is take out one or two of their top players after multiple attacks.


#2020

Yep the war scor difference shows a challenging match.
Depends on how mutch they wight the top 5ves right now. :slight_smile:


#2021

In our wars I usually take out the 3 strongest opponents teams with the 3 first flags ( on a normal to good day :smile: ) .
This is because I am far stronger TP and benchwise than all other of my alliance members.

Usually the next couple of teams TPwise belong to our opponents .

The war still is often very challenging ( had a loosing streak to ), ( because of a bad day or sometimes a bug, or good play by opponents )

How where the war scores of the 2 other wares you lost ?

During our last war I was very impressed to see one of our newer members ( Bane is his only hero over 2* ) using 5 flags and scoring points with every attack.

So wars a rarely hopless from the beginning but can turn out a lot better or worse than is to be expected after war scores.

Before the war beginns I am only interested in war scores.

When I see who is on the battlefield I analyse opponents teams for weaknesses and give directions for attacks after opening the war with my 3 flags.

So in the end.

You had a vey challenging task.

You lost as was to be expected but not sure from the beginning.

You and your alliance hopefully enjoyed and learned from the free training.

Hopefully your next matchup will be better for you. Is to be expected.


#2022

Honestly, there is nothing to learn from this loss. I don’t recall the other war scores, but they were usually slanted the other way. I think there was one where we were slightly higher and it was a close one. But like I said, every opponent we’ve seen has multiple defensive teams with higher scores than our top. So tell me what use is it to use up 3 or 4 high level teams against 1 opponent, leaving 2 or 3 other teams that we no longer can beat because we used those resources on that one opponent?


#2023

“Honestly, there is nothing to learn from this loss.”

Now this is very sad. Wars are the place to test all your heroes and find new ways to use them. If you dont learn something in every war you miss a great opportunity. ( result of war is irelevant for that )

Hmmm….look I dont know how you prepare and fight wars.

On my side I usually split up my 6 fully ascended 4* heroes depending on opponents team composition and round them out with 3 stars.

Luckyly on our level a lot of opponent players still make basic mistakes ( weak center, neighbouring heroes of the same color and so on ) which improves my chances a lot.

Luckyly again most opponents dont implement simplest war strategies ( mono color tanks ) and are weak on team work which allows us to show better results than expected.

Most of the time we have better participation. ( Took a good handfull of wars to get all participants in line. :smile: )

It doesent take a very strong team to take out one tank and after that the next team can be weaker and still take out the tankless team.

So, often there are hidden possibilities that can lead your alliance to victory even against bad odds.

Of course there are the days where you loose even if the odds were in your favor. :unamused:

So finally I want to say :

Keep going.
Always try new things.
Enjoy the things that work out and dont give up if they dont.
There are better days ahed, just around the corner.

With best regards.

Elayanith


#2024

Sorry, I find the response a bit insulting simply because when you have 1400 level teams going up against a depleted 2300 team and not make take any points, there isn’t much takeaway other than low level teams can’t do much.

That’s great you have the luxury of 6 fully 4 star ascended heroes. Now tell me what you can do meeting the same opponent with half leveled up 3 and 4 star teams Or 2 star teams? That is more the situation. You are commenting based on the strength of your team, I am commenting on the weakness of our lowest level players that have little to work on.

We’ve had participation as we can. I assume your team is primarily located in one time zone? While ours is, we still have members on the other side of the world so we don’t always touch base, but we do keep chatter alive as we can.

And again, I point out that aside from a few members, We all participated whereas the opponent left attacks on the table.

My point in this is that we were overmatched by any metric, but you are saying that we did something wrong and it is our fault for losing.


#2025

I am saying you had a very challenging match up and it was not your lucky day so you lost.

You have 9 Teams with TP>3000 which gives you already some 4 stars at least.

As I said I dont know how you fight your wars and I never wanted to say your participation wasnt good ( it was good ) or that you did something wrong.
My point of view was supposed to show the possibilities I use in wars.

As you could see our weak players very often play clean up or fight against other weak players (if available),
or even attack a stronger team to take out the tank if their heroes match.

I can see that you are very involved with your alliance and I am defenitely did not mean you did anything wrong. Just tried to show you how I do it because I learn a lot from other players in this forum and I defenitely dont think my way is the best or only way….it is just my way.

Best regards

Elayanith


#2027

6 attacks are attacks of an extra person. In addition, the price of each team of a smaller alliance is higher, which is equivalent to another 6 additional attacks.


#2028

Hmmm, interesting discussions and seemingly misinformation here.

First off: is the new matchmaking system truly an ELO-based system? Do we move up and down a so-far invisible ladder based on our “war score”? If so, then players have to understand how an ELO system works.

A true ELO system (and I suspect this is at best modified ELO-like) ONLY takes into account your previous results and your opponent’s ELO score (as a measure of their strength). You beat a team with a similar ELO score and you will go up roughly equal to what the other team goes down, which will only be a small change. You beat a weaker team (much lower ELO) and your score will only go up a bit (much less than when equal) and theirs will only drop a bit. You beat a team with a much higher ELO and you will go up a LOT with a corresponding big drop to the other team. I don’t know if this is true of AW. Of course since the goal is to avoid wars with large war score difference, this may be moot.

ELO relies on previous performance data however, and that is the general shortcoming here. Currently for matchmaking alliances previous data includes data prior to the new system and that skews things badly in some cases. You are in an alliance that had an excellent winning percentage pre-change. Now you are facing overwhelming opponents. That is the system self-correcting, You have been the recipient of an unfair matchup system, and now you are paying for it. Your war score is inflated. Unfortunately you are also (probably unwittingly) the reason why the change had to happen.

All I can say to help is that after a few more wars the pre-change data will have very little influence on your war score and you will be positioned more favourably. Then you can face more equal opponents and, if you are more skilled or more lucky, you start winning again.

The other option would have been to have everyone start at 0 at the change: no previous pre-change war data used.

As for the 50/50, that is not a goal of ELO or rankings. The goal is to match balanced opponents and let the better team win. Overall (taking the entirety of AW into account) the results will be 50/50. There will be teams that do better and teams that do worse. It likely ends up a bell curve. You need to work on being on the downslope side rather than the upslope side of the curve. Most teams will be somewhere around 50%, some will be higher and some will be lower.

As for the alliance I’m in, we won some and lost some both pre- and post-change. Serendipidously we were rarely badly over or under matched. Maybe once or twice in 6 months.


#2029

Nope.

Yep; just postulate a ladder based on the war score, and you’ll move up and down it. But it will be a war score leaderboard, not an ELO ladder: You’ll move up/down not just by winning/losing, but also move up/down by getting/losing eligible and opted-in members, move up by improving each player’s top 30 heroes and best troop of each colour (whether by pulling, levelling, or ascending), and move down by feeding away top 30 heroes …

Some wish for an ELO system. As long as the alliance membership and opt-in/out is as fluid as it is, I’m not sure that’s a good idea.

Even without an ELO system, it ought to be possible to implement a ladder based on the results rather than the score. Might be more of a pool ladder though.

I would welcome a leaderboard or ladder either way. It’d make this whole stuff more tangible.


#2030

Yeah, hence my “at best modified ELO-like”. ELO takes account of changes in player strength by players winning matches against higher ELO-ranked players. This system tries to take account of changes in alliance strength prior to the wars and adjust opponents to match. In ELO sports such as tennis or chess you WANT to face stronger opponents. Even doing well while losing can actually increase your ELO. Here you don’t. There is no chance of increasing your score when you lose no matter how gallantly you do.

And a leaderboard would be good. Then you could see the alliances around you (your potential opponents).


#2031

Either way there is a correction going on in the system for some alliances that is causing a lot of resentment because they don’t understand why.


#2032

Even with true ELO-System there would be mismatches because of the factors mentioned above and other reasons like possible point droping. Almost every system can be abused but as in Chess if SG would keep an eye out for strange ELO-point fluctuations it might work.

The main advantage would be that it would be almost only a question of interalliance discipline to get stable paring results which for the dedicated alliances would be good.
Of course the less dedicated/disciplined alliances were still prone mismatches but that’s the way reality works.


#2033

I agree our wars have been a lot closer and now we win a few. The only issue I have is recruitment of higher players just before war starts during the preparation time you can tell when you look at how long they have been in alliance then they leave right after the win. Should stop people being able to join during preparation time this would fix that issue.


#2034

Those that join during preparation phase are not able to participate in wars, I believe.

Those that left, then join again, during preparation phase may appear on the battlefield (as ex-member, perhaps?), but still won’t be able to attack.

Do you have screenshots showing differently? (Player info and the list of attacks made?)