[May 16] Alliance Wars Half Way Through

Well, our first new alliance war…we were excited to see what it would be…Our opponent is double our size in points …have 9 players over 3000 (we have 1 at 3040)… their lowest player is equal to our #2. So far, and it’s not over yet…they have killed us all and revived us and killed us again. The second 12 hours? Who knows? Maybe they have nothing and our little 2* heroes will cream them :joy::joy:. We can’t finish off a 5* Titan yet and they are routinely killing well beyond that … come on guys … there has to be a better system​:pray::pray::pray::pray::high_heel:

5 Likes

i feel your pain , exactly the same sort of junk we have

1 Like

I have some sympathy for your plight, although no matching system can eliminate mismatches.

But AW matchmaking used to be based on Titan Score - a metric you imply might be appropriate - and the forum was filled with players complaining about what a poor choice it was. I feel like maybe that would have been a good time to raise your voice and explain why it was appropriate. Too late now. It’s gone.

2 Likes

I would think there is some way to take all things into consideration … just basing it on Titan scores didn’t really show a full picture either as we were beating every alliance war easily and our opponents just gave up … that’s not fun for anyone either. You can’t take one thing I said as an example of the disparity and say “that’s it”. The game has multiple measurements… why not use them all in the calculation?

How, exactly? This stuff is not magic.

1 Like

We (Emperor’s Assassins) have been matched up against a team that it appears we shouldn’t have been by any previous metric. However, the score is still close despite us having a 10 to 15 flag advantage as of right now. I’m willing to give it a chance.

Ok, first let me say that my team and I enjoy the game and my comments were not meant to offend anyone. But you have to admit the disparity is pretty big in this instance. It does appears others were well-aligned. So I have two suggestions you might look into … they may or may not work and I don’t doubt that you and your team are doing your best to come up with ideas too

First, perhaps use all the measurements you have: Titans values, trophies, mean hero values, etc. give greater weight to those that players can manipulate less —- so Titan strikes may carry a greater weight than trophy count for example. It would be reasonably easy to tweek the weighted percentage as you see how it works

Secondly, which would work with the current process, or perhaps any process used, is to tier them just like you do trophies…alliances with 10 war wins would play other alliances with 10 war wins … eventually, it will sort itself out. Obviously, some teams will try to manipulate this by opting out of wars to stay within a tier — which honestly makes zero sense to me, but somehow will to some

Lastly, I realize there is no magic and no perfect answer. Thank you for doing what you can.

1 Like

I should also note that part of the disparity is probably our own fault. I have five level 5 heroes, but I only use two of them as we are a newer alliance and I just have et had time to level them up yet … another of my team members told me he has three 5* which he recently got and also hasn’t had the time to level them up— I’m guessing that if we canvassed everyone who met with this disparity, we might find this to be a systemic problem. The tiering of war wins would eventually sort this out as well?

3 Likes

@Brobb no system is perfect but there is a very easy way to ensure closer battles. The system knows when everyone started, so take the average number of in game time in days and match it up with some team + - 30 days. Thats from when they ACTUALLY started, not in alliance time. Coiners will still be strongest…tempting more to coin( bonus to owners) . if average exceeds 365 then alls fair in war.

Crusher, while I won’t disagree that it will force people to spend more money, it will ultimately kill the game. Game of War and Clash of Clans are two great examples of this. Once your ability to play the game becomes based on how much you’re willing to spend, the days are numbered. Spenders will do what they do, but once all that is left are spenders and they don’t have that advantage anymore, they simply stop playing. You need a balance to keep the game going long term. Even the spenders want people they can beat. :slight_smile:

I think this could really be the core problem for us newer guys, we won all but one AW pre patch most of them really close strength and point wise.
First war post patch enemy alliance got all defense power about 3k with good defensive heroes and a few leveled 5*(3rd ascension) in the mix.
We got only 9 guys above 3k teampower and most of us only got ~10-15 leveled heroes, but many of our members got unleveled 5* and 4* heroes.

AW and It’s matchmaking mechanics! True that It’s not perfect however It’s the closest yet with now SG’s third attempt. Titan scores has diddly-squat to do with measuring alliances strengths for AW, however now It’s much better. My AW match, my alliance Titan score wouldn’t even be a blip on the table graphs compared to the other alliances Titan score. Someone very early on from the alliance commented about this comparison and soon followed, one of our stronger member had left. I wasn’t around at that time but when that changed, I made it clear that this comparison with Alliance, Titan scores doesn’t mean diddly-squat! And once the Wars started, they all quickly learned how irrelevant this comparison really was!

Lost a member before Wars begin, we have several members not yet qualified to participate, so It’s 22 (them) vs 17 (us). Regardless us a blip for titan score in comparison to this other alliance… we from the very beginning been annihilating them!

Even though it’s better matchmaking, It doesn’t mean that our scores at the end of the war will be even remotely close. If you go in with an uneducated and disorganized alliance, you will suffer. You need to have at least one good strategist. To help get alliance up to speed, and prepped for AW battles.

2 Likes

Well, I suppose time will tell. This was the first one … our opponent is currently almost done killing all 26 of us for the third time … we can’t even touch most of them; although we didn’t play dead. We teamed up and used multiple smaller players to systematically kill off some of their players, but we won’t come anywhere near their score … a humiliating loss is sometimes good for the soul. If it continues— well, guess we will try therapy

2 Likes

I’m a player like you, not part of SG’s development team. So I can’t implement the changes you propose.

But I can say that we players have been discussing the best means of war matching here in the forum for a very long time. And while the method you propose has the intuitively appealing characteristic of incorporating many of the elements that might be relevant, it’s notably light on detail.

How would these factors be calculated? How would they be weighted? How would they be scaled? Would a regression be used as a starting point, or some sort of arbitrary formula?

Even if we could answer all these questions sensibly and confidently (we could, but our answers would be arbitrary), the resulting formula would not necessarily provide better matches. It might provide worse matches! We have no way of telling.

1 Like

I’m not convinced that matching alliances based on time in the game would ensure close wars. I suspect that mismatches might be even more common.

3 Likes

Honestly, the devs were pretty lazy with this update. From what I’ve read they have scored all 5 star heroes even a t1 lvl 1 5 star higher than a fully maxed t4 lvl 70 4 star hero. All 4 star heroes regardless of level are weighted higher than 3 star heroes and so forth.

They need to give each hero a grade depending on level and tier if they’re going to calculate war opponents using overall team strength. A t4 lvl 70 4 star is significantly better than a t1 lvl 1 5 star so it should be weighted accordingly. The ranking of top 30 heroes is a great idea but they need to actually calculate heroes base stats and create scores according to level not just saying “oh Elkanen you’re a t4 lvl 80 5 star you get the same grade as Guinevere.” We all know one is significantly better overall.

Anchor has already done a bit of the ground work for them by creating a spreadsheet grading each fully leveled 4 and 5 star hero. I may very well be wrong but from what I’ve gathered a t4 lvl 80 5 star is weighted the same as a t4 lvl 1 5 star. The stats are significantly different and should be weighted as such when creating an overall team grade. Even if they did a simple rating system that focused on base stats instead of individual hero grades - a hero with 700 attack, 950 defense and 1100 health will get X grade and any hero that falls within a +/- 20 point stat range has the same grade would be better than what they have implemented currently.

Unfortunately, we were also paired in a drastically one sided war. The team we are facing has 9 teams over 4000 power. We have 4. Their lowest player is 3418 we have 8 players below that. Most of our players hover in the 3400 range. The team we are facing also kicked their weakest members prior to the start of war and brought them back in today. They have a few members who joined 16 hours ago. I’m optimistic that bugs and kinks will be worked out eventually.

2 Likes

Logic dictates that only the outrageous mismatches will be reported in this thread.

Therefore the success of the new matchmaking method can somewhat be estimated based on how long this thread will be compared to the threads providing mismatches of the old matchmaking method.

So far, things seem to be going in the right direction for the new matchmaking method. But this is a halfway exit poll, so big grains of salt all over.

1 Like

Look at the data and choose the winning team? For me it’s quite obvious. And don’t write about hipotetic benches, secondary heroes or wise strategies. It does matter, but let’s be realistic.
matchup

We never had a match this close before…

We never had a match like this before…