Serious matching problem with alliance wars

I think the attraction of these types of ‘fixes’ is that it will reduce the number of matches that LOOK like mismatches during war prep. People continue to complain about the matchmaking before a war happens by comparing things like:

  • Alliance scores
  • Titan scores
  • Number of defense teams above score X compared to other teams

Because people feel they ‘know’ the war will be an upset before fighting it, they lean towards ‘solutions’ that will affect these visible mismatches.

There IS a matchmaking problem. This is evidenced by the overwhelming community response, the valid complaints raised by people who lose wars by 1500+ points despite full participation. and probably most importantly the active SG dev response.

The problem they should be focused on (and I believe they are based on their comments) is how to match up alliances who will preform similarly in wars so as to increase the number fo fun fights and reduce the number of blow-outs. Let’s hope we get there.

Funny, if I wasn’t on these forums, I would think the work they are doing has made HUGE strides towards fairness. My alliance won our first 10-12 wars in huge blowouts saving one war in which we were hugely blown out. Since then, we have had a number of wars where we traded leads back and forth. Our wars these days require full participation, solid resource planning, and coordinated attack strategy to win. These things are actually pretty darn fun when they work! They are working for us. I’m counting my blessings. Hope everyone gets to this point eventually.

2 Likes

Having attended an Ivy League school, most of our football games seemed as though they were Div I Vs. Div II despite having 2 Div I teams on the field :wink:

Look though, Rev - I think we all agree we would like good matchups and that blow-outs are no fun. TO continue the sports analogy though, remember that blowouts happen and it’s not always because of bad match-ups. How often does a basketball team blow out another, then play the same team within a few days and get blown out by the same players? It happens. In this game it will happen even more due to the random factor of tile distribution having significant effect.

1 Like

So to be clear, your matching proposition is just that, to be matched:

  1. Alliances would have the same number of players above level 12, and

  2. Alliances would have approximately equal aggregate Top 5 powers.

Given that the AW matching system already requires approximately equal numbers of players, the only substantial change you are suggesting is a switch from Top 30 to Top 5.

This seems comprehensively wrong. For months, players argued in this forum that requiring war pairings to have equal numbers of players and equal Top 30 scores would result in pretty equal teams. SG listened, changed to exactly the system proposed, and hey presto! Mismatches were eliminated!

Wait - what? They weren’t? Because just because alliances have equal Top 30 scores, doesn’t mean they are evenly matched, right? Similarly, just because alliances have equal Top 5 scores, doesn’t mean they are evenly matched.

Your proposal would continue to ignore the significance of the intra alliance power distribution, and would gleefully ignore bench strength, something the current system at least accounts for. I like that idea - it sounds fun and it would favour deep benches - but it would probably make mismatches even more common than they are now, which is what I thought you were trying to avoid.

I have never in my life watched a college football game, regardless of the divisions, team colours or mascots involved.

OK so at risk of getting too speculative, I believe the motivation for switching to top 5 over top 30 is that people are seeing issues where there are teams on the battlefield that are so much higher than they are used to raiding that they feel they have ‘no chance’. This can happen due to these 5 heroes being wildly stronger than those players benches but averaging out over 30 obscures this. Of course this simple reasoning ignores the fact that stronger benches on the opposing team would be necessary to balance out the apparent mismatches in the top teams.

And more to the point, all of your counterpoints remain valid.

Most importantly, IMO, SG has already taken this to the next level, coming up with a class of solutions (that they are currently testing) that base matches off of top 30 heroes but with additional weight given to top 5. So…what are we talking about here?

1 Like

The “approximately” is a problem. The 30 is a problem as well. You mostly have to fight the top 5 as defense teams (not to mention the stupid arrows). If the defenses are near equal, then so should the games. Someone else commented on tile selection and it’s sometimes a problem as well. No way should a 2800 team lose to a 2500 team and not kill a single hero. We are on our 6th or 7th war as underdogs. Most of them we got blown out. Not for lack of skill or effort. Two of those we got lucky the other team didn’t put forth the effort. I enjoyed those wins but it almost doesn’t feel like a real win. I don’t want to blow teams out either. We had that once before. Maybe they have something that works with the new update. Whatever works. Take care.

Ok just a newbie here who has been casually watching these boards for a couple weeks with a suggestion for you to parce.

It seems everyone is focused more on alliance scores and heroes for matchups. I don’t see much discussion about an alliance’s prior performance having any influence on upcoming wars.

While not perfect what about using alliance score with a graduated multiplier based on war wins and losses. You win you face an alliance 5% stronger next war. Win again another 5% and win again it goes up 10% etc until you loose then resets. You loose again and it goes down 5% etc. Percentage can be changed to tweek system. Bcuz it’s graduated I don’t see a big advantage to sandbagging.

New alliance’s may have to fight a minimum of 5 or 10 titans to be eligible for wars. Limits could be placed on not being able to fight too weak of an alliance.

Those who fight stronger alliance’s and win should get better loot and possibly a ranking system like trophies for raids. I don’t think anyone here is happy with the participation trophies we’re getting now!!

Would this not level things a little?

Given the mismatches we’re facing now isn’t it worth a try? It will take some time to balance and not every matchup will be perfect but participation and morale are dropping fast in my alliance and that is NO FUN!!!

Lets hope. As I mentioned above, our wars have recently been fun, close match-ups. They are fun when they work!

1 Like

Hi :slight_smile: I’d highly suggest reading through the multiple threads on this subject. A War Score/ladder/ELO based matchmaking system has been discussed many times. The SG devs have even weighed in. I am a champion of this type of solution.

FWIW The devs have countered that the high rate of member turn-over in alliances renders ELO type alliance scoring problematic.

Probably better to leave those discussions to the threads that already cover them.

1 Like

I see, thanks for response! I would hate suggest a solution that would be so problematic compared to the current non-problematic system. :wink:

1 Like

Well I agree with the approach, as I said, but also appreciate the SG is working hard to try to address the underlying problem. Hope we get to a better place. The current situation does stink.

This, I think, is the difficult bit. Not only do we not have any reason to suppose it is true (in fact, we have multiple reasons to suppose it is untrue - benches and power distribution) but we also do not have any evidence to support it (though it should be easy to collect - presumably under the current system alliances with equal aggregate Top 5s face each other from time to time).

That’s the thing: it’s easy to invent ways to make good wars. (‘Equal Alliance Scores will make good wars!’ ‘Equal Titan Scores will make good wars!’ ‘Equal Top 5 Scores will make good wars!’) Just saying it doesn’t make it true, and in this case we have good reasons to think it is false.

That doesn’t make it a bad system, by the way. It just means it would result in many mismatches.

1 Like

Only thing I disagree with @Brobb is I think that systems that result in too many mismatches are inherently bad systems. Otherwise this is the issue in a nutshell. The easy ways have been tried. They all suck. The remaining solutions are more complex. Hopefully one of them will be better.

Suggestions are helpful. Assertions without basis are not.

1 Like

I can’t argue with that, although I don’t know how to judge when there are “too many mismatches”. SG can see all the war results, so maybe I should wait for them to tell me that things are now okay? Or should I wait for players to stop complaining? (That might take a while.) Should I just accept things are okay when my own alliance gets good matches?

I don’t have a good answer.

I personally think they have the matchmaking pretty darn good.

We just fought an alliance, their overall defense strengnt average was 3390. No one under 3200tp.
Our alliance has an average of 3310tp, 7 under 3200, 1 of which is 2900, 2 at 2600

They had an alliance score of 111000, us 109000.
Both had 28 members.

Final score was us 5100 to 2500. They gave up half way.

Both our 2600tp players got 200point+!! We strategize that they would use their weak teams to clean up.

In the end, we coordinated 1 all kill attack, and got a 2nd one later.

This happens war after war.

A bunch of 2900tp attack teams can take down a bunch of 3500tp. All they need to do is work together. Not attack out of frustrations. Use the right heros to attack the right tanks, etc…

People complaining that they only get 2500points in a war. Well thats 14points per attack. Match ups is not the issue. Players ability is. Strategy is. Raiding ability is. Your alliances might lack the proper leadership to make players better.

Hahaha - yes let just wait for complaint rate to reduce.

No - only SG will be able to make this determination. Agreed. All we can do is agree that changes to matchmaking that will increase mismatches go in the wrong direction.

1 Like

Agree that Alliance War skill is a huge factor in wins/losses.

Disagree that matchmaking is pretty darn good. SG recognizes this as a problem worth releasing a number of fixes.

In the end I’d like to see a system that matches teams against teams that are likely to give them a good fight by measuring strength of current heroes and prior war experience. Then see them give us a way to gauge our progress through the ranks as we proceed. Wouldn’t you like to have more interesting matches. And for your wins to allow you to progress to a more worthy opponent? I know I would.

2 Likes

This is my point. You had a good matchup. Defensive averages were near equal. Our averages have been way off. Battling teams with 13-15 3k+ defense teams and the lowest being 2200 or higher. Our best 3 are barely in the 3k’s. We take noobies so our low’s are the lowest. We keep a full roster and hope to be all 3k+ someday but we are what we are. No matter how you slice it, 3k defenses (and arrows) obliterate anything 2k and below. You might keep a close first round but if they play the second half… RIP.

I really do think that the disparity will remain in any sort of match up system until they start requiring people to set up six defensive teams with each successive team replacing their prior defensive team once it has been defeated.

Once all Six defensive teams have been defeated, then the person’s icon becomes a “white flag” and they can no longer be attacked.

1 Like

I’m not a strategy expert, but I certainly think that would add a fun element to the war. I love the wars, but it isn’t much fun to keep attacking 3400 teams with 2200 teams once you’ve used your big guns.

Would it also make sense then that the attack teams should also be pre-set?

We do have another update soon that may fix what some are seeing as a problem. I think AW matching is spot on.