Alliance wars still unfair


#167

Since they are matching by the 30 most powerful heroes of each player, you may want to consider how many 3-70 5* you have hanging around that are counting for your strength and are not being used in AW.


#168

So without mats they can’t be ascended, without ascension, no mats. Nice system.


#169

Unfortunately and fortunately you can’t speed through this game. Well, you can to a certain degree. You can invest money and/or a ton of time and planning. There are several ways to get ascension mats and I don’t need to rehash that here. I’m certain you’ve heard it before.

Yes, it is possible to have 25 level 3-70 5* heroes waiting for final ascension and many players have exactly that. At level 3-70 many of those heroes are inadequate to the task when it comes to AW use, especially at a high level of play, but I don’t think those are the people complaining about mismatches.

I have two accounts and both are in top 10 alliances. The matches we get are all fairly close because at that level nearly all of the players have fairly stacked benches. It is the lower level matches that are out of whack and I think that is due in part to the number of 3-70 heroes they have on their bench, but are not strong enough to be of much use in AW.


#170

Yes Otto, I was merely remarking on how adding war chests makes the war mismatchup system even more eggregious


#171

Ah, I see. A strong get stronger type of thing. Well, I have to say that What is needed is to really sit down and think of ways to compete on a more earnest basis, but I’d wait to see what the rewards actually are in this case before complaining too much.

Meanwhile, there are things that can be done to make your Alliance fight more efficiently. It is becoming more of an issue for people to contribute and participate in AW. I think that you’ll see more of a segregation in the future. You’ll see more alliances comprised only of those who do not want to participate in AW.

If you have not already done so, I strongly recommend reading Yo Momma’s war guide. It is posted here in the forums if you do a search. There are many things that can be done to help your Alliance fight well above any comprised score you can give them.


#172

Thanks, I’ve checked out the different guides like Copperskys, Anchors, etc. Plus I get lots of tips by lurking in PS chat. Just wanted to add my say in on the topic not going to dwell on it. I did suggest that maybe using the alliance scores with -1000 point for each loss and +500 for each win would develop a war level that would even things out.
We will see, maybe the individual opt out will change the dynamic, maybe not.


#173

This makes me sad. As you say, Otto, 5* at 3/70 are usually pale reflections of their ascended glory. I leave 5* at 1/1 until I have five of the color-specific mats (which have always been my choke point). The only exceptions n my hand: Thorne, an early 5* that I didn’t realize was garbage, and Natalya, who got bounced from her rings by Gravemaker and then Red Hood.

Until I have the mats to complete a 5*, I work on more 4*. Especially with the new crop of S2 heroes, there are always more good candidates.


#174

Yeah, I agree that most 5* don’t really reach their glory until final ascension, but there are some 3-70 5* heroes that I do use in AW. I tend to have one maxed and one at 3-70 of the ones that can be used in AW at max 3rd tier so I can mix and match to make duplicate teams that work, but that’s not what we are talking about here. The problem is that many people have Max third tier Lianna or early mistakes like your Thorne out there that really can’t be used in AW effectively.


#175

I agree that 3/70 5* healers are fine. Who else are you thinking about, @Otto0000?


#176

Heroes that I use in AW at 3-70 are heroes that I mix and match with teams that work well for me. I use Panther, Natalya, Athena, Perseus, and a few others mixed in with some of my teams during AW.


#177

but its not nearly a fair matchup not by far, my alliance is going up against an alliance with a score of 49232, and our score is 42502, and their top 30 heroes are a hell of alot stronger then my alliance top 30 heroes, not a fair matchup by far, this needs to be fixed immediately.


#178

we are not going to be winning any wars if this keeps up


#179

If you lose this war—and I hope you will try hard to win—then the matching algorithm will find you an easier foe next time.


#180

Hi- this is my first post. I am Vica at Thunder Monkeys alliance. I have a mathematics degree and would like to make some suggestions to the game developers that should improve significantly the alliance war matches.

I suggest applying the Mann–Whitney U Test formula to an alliance’s hero roster. Then matching should work on a basis of equal number of teams and matching U numbers. Additional metrics could be used such as average team power, average variance and standard deviation.

By example, my current war opponents- our U tests are completely not matching- what should be a score of 55 (even assuming 10vs11 would be a fair match) our U scores at 42 vs 68- meaning we are heavily out matched against this alliance. Our average power is just 2150 vs theirs at 3150.

If developers would like to consider a more fair and statistically balanced war matching- please let me know :slight_smile:

Thanks,
Vica


#181

I get that “if you lose this time we let you win next…” is perhaps very easy to get and therefore implement. But actually, no one feels motivated to get battle knowing you will lose also- no one wants to be given victory on a plate. If, statistically, the matching would be balanced and fair- the customer experience would improve- meaning, when you win, you won on your merit due to skill, strategy and teamwork- this is what people really want.


#182

seems like we all have to take deep breath and wait some time…

Continuing the discussion from Version 15.1 Release Notes:


#183

Sure- from my side I will anyway monitor the stats and see how it plays out.


#184

There are several problems with this approach:

  1. M-W U compares two distributions and doesn’t provide a descriptive statistic of a single set. Thus SGG would have to compute all the pairwise Us for all possible,pairings and then find matchups where U was smallest, while matching all alliances. This sounds computationally expensive.
  2. U looks at pair-wise matchings but the game mechanics match 30 heroes against the same 5 top (or near-top) heroes.
  3. Sole reliance on U matching fails to take account of actual performance of teams in the past
  4. U relies on having a valid summary statistic for the power of a hero. I’m not yet convinced we have that.

#185

I’ve never seen reason to complain about war matching until this one. Mind you, our alliance was ranked #345 at the time we got our in game notice, and the opponent was ranked #34.

I don’t know where the matching went wrong, but many on the team expressed concern immediately, but we gave it a chance to see how the war actually went. Unfortunately, it’s been a total slaughter, see as anyone would think looking on paper.


#186

If this was the result of the new TP numbers, I vote to roll back the change