Alliance wars - matching manipulation

alliance_wars

#284

We have a family alliance and rotate some in/out, but were not exchanging 17 players all on the same day to sister (training) alliance and having 9 of them also swap to the other on the day prior.

Overall this only the second time I think this happened, so not too big a deal. Of course it could just be me seeing that like everyone who swears the RNG is out to get them and in that case I’ll put my tinfoil hat on.

The staticians always say to look at the data, so I could track war scores on both teams, monitor members every war, and log it over a period of a couple months. Practice my Sherlock Holmes detective work.

I’m curious to see if how much someone could drop a war score by horribly losing 5 in a row.


#285

It’s not clear that margin of loss matters. Unfortunately, we know little enough about how past wars factor into the war score that it’s hard to gauge if such a strategy of AW-dropping would be genuinely effective or not.


#286

And also what their purpose would be? Cup dropping allows for easier raids and quicker chest filling but to “war drop” would take a sustained effort over a few weeks. Plus the coordination it would take since you are talking about several players, not just a single cup dropper.

I just can’t see how the rewards would ever be worth the effort, and are there that many players willing to throw their war scores for those small rewards?


#287

Would it make a difference how well you been doing against titans and last few wars? It seems to me every time we take down a 9* titan ( we usually do 8 then get a nine after we’ve beat a few… we usually kill the nine but don’t have enough to stay for a second 9*so we drop back to 8… you guys know the drill) we get a war match up where we get blown away by 1000 points. AW have been pretty good the last dozen match ups with a few mismatch ( never in our favour) lately. Just wondering if I need the “tinfoil hat” or coincidental.


#288

If the devs are going to discuss how game mechanics actually work, it’s extremely unhelpful to do so where 98% of the community can’t see it.


#289

[quote=“princess1, post:101, topic:90372”]
Plus the coordination it would take since you are talking about several players, not just a single cup dropper.

Wouldn’t be too hard, just a line message to change over main/alts. If they are, I should see a change over again of most players after 25 to 30 days all fairly simultaneously


#290

For me an ‘exploit’ would have to offer a huge gain and I just don’t see that with this.

To fill a war chest, even if you do win all 5 on the bounce, will take 2 and a half weeks.

Even if they switch back to the ‘main’ alliance in between they’re still missing at least 5 titans in that time for the 48 hours they’re locked to the new alliance for matchmaking and the war itself… I’d take five titan chests over a single war chest any time.

Such a huge amount of effort for very dubious rewards seems more like stupidity than an exploit to me.


#291

Moderator’s note: I’ve merged three closely parallel threads; apologies for some jumbling.


#292

What happened to the responses from my thread earlier today? I get the merge but I was hoping to reply. Dont see them here :frowning:.

The solution to this is really simple:

  1. Apply war score to players not alliances. If you change alliances your chunk of the score follows you. This should be easy to implement.

  2. Simply add a war cooldown for changing alliances. If you stay you have a 5d war CD (2 wars). This is super easy to implement.

  3. Remove the past performance factor. I dont like this as much as it could lead to some crummy match making streaks but it would work.

The argument that “alliances merge” or that this isn’t beneficial is naive. It’s a clear advantage or it wouldn’t be so rampant. It’s exactly like sandbagging in golf.

“Just beat them”. Ok yea of course, that’s the plan. Were 1-1 vs these guys. Got our 3rd match in a row coming up vs this strategy. It’s just dumb we are forced to go in at a disadvantage having won our last war and they may not have any history points.

Cup dropping doesnt effect a whole alliance and chests that take weeks to open. Apples to oranges.


#293

Everything is here; your thread came in second, then a third thread. Why Discourse doesn’t sort by time is a mystery to me.

Your original post is here: Alliance wars - matching manipulation

Please do check the existing threads before starting a new one!


#294

Thanks.

Believe it or not I did I just didnt have great search parameters apparantly. Keep up the good work Keri!


#295

You can narrow the search to a particular user by using @‘username’. Very helpful in finding something when you can remember who said it but can’t remember exactly which thread.


#296

Have glanced through the whole thread, think I understand the gist of it, but apologies if what follows is a repeat.

Firstly I cannot understand why any alliance/person would bother to go through whatever ‘process’ beyond playing the game to just win a War. War loot doesn’t appear to be much different if you win or lose, or where you come in the rankings. The only advantage can be to get the chest quicker. Ok I am not vastly experienced but that seems to be the case.

Secondly a view from the bottom as such…Our alliance is low level 150K war score, 50% of the matches are pretty even, 50% a mismatch, but not always the war score!

Worst case score mismatch was opposition +28K, so in 150K that is quite a difference! Strangely that was one we actually Won? although the other team did appear to give up after a while.

What has been the case for a while is a mismatch of the top 4-6 players Going by LEVEL and Defense rating). Currently we are facing an alliance with 6 3500+ teams with 4 players lvl35 and above, one at 41. Our top 5 are levels 27-32 with defenses 3000-3400. The rest of out team is possibly slightly higher but at 2500 and below that does not really matter.

Okay so how does this affect things, simple answer the other team will most likely wipe/clear our team so gets bonus points, we cannot. Even with co-ordinated (mono) tank busting etc there is not the strength in depth to follow up on the big teams.

So the easy answer to ‘manipulate’ is to have 4-6 Strong players (24-36 attacks). I have continually mentioned a players level as that is probably the best measure of how experienced they are, and what team depth they would have. I can’t imagine anyone would get to level 40 for instance without really doing anything.

Although it appears to be a complaint its not, when I first joined this alliance it was the other way around 2-3 Very high players level 50+ 4K teams, the rest of us filler basically lvl 25 and below. They would wipe all the top 6-9 opposing teams and multiple attacks left for the rest of us to clean up. The opposition could not kill the top players so no bonus.

It has almost got to the point or predicting a win or loss before the match even starts :grin:.

Apart from that I enjoy alliance wars, win or lose and that’s all the matters. I have probably also sent several people to sleep with my first post.


#297

I’m agree with you and am pretty well convinced that players cannot manipulate the matching algorithm by switching players between alliances, and that the main reason for losses is the disparity in use of WE. However, since the player switching right before and after wars is both continuing and readily apparent, it got me thinking about why an alliance would do so, and I thought of some advantages.

A well-organized family of alliances will likely deal harshly with players who don’t either use all WE or opt out, so they’ll know who will and will not be available for a war. By swapping out unavailable players for available players from a sister alliance (before matching occurs), they can ensure they use up all WE and likely get the win. In addition, they can construct a roster of players all in or near the same time zone to use the board clearing strategy multiple times near the end of the war.

This is a legal exploit and not cheating, so allows more wins and a resulting higher war score. However, that also means matches against tougher opponents for the winner and easier for the loser, so there doesn’t appear to be any long-term distortion of the kind of teams that more- and less-organized alliances will face. It will ensure that more-organized alliances are generally higher on the list than lesser ones, but those are the breaks. :wink:


#298

This has been discussed a lot elsewhere, but player level is a very poor proxy for strength—remembering that the matching code can see everything: full hero roster and troop roster, as well as track record in war.

Take three players:

  1. Level 40, completely free-to-play. Has five maxed 5* (some good, some less so), twenty maxed 4* or 3/70 5*, plus some scattered 3*
  2. Level 40, big spender. Has 12 maxed 5* (all top quality), 18 maxed 4* or 3/70 5*
  3. Level 25, moderate spender. Has seven maxed 5*, mostly good, and 15 maxed 4* or 3/70 5*.

To my eyes, players 1 and 3 are more similar for matching than 1 and 2.

What really matters is how many strong heroes you have, backed by whoch troops. That’s exactly what the current war matching algorithm looks at. We as players can’t see all that information, so we use player levels as proxies, but if you can actually know exactly how deep someone’s bench is, isn’t that anfar better meteic?


#299

Thanks for the replies to my post of mild frustration, mostly due to my lack of ability :disappointed_relieved:.

Both valid points, an organised alliance will have a much better chance. Midweek wars is always toughie starting early morning (UK) with most of us occupied at work, so attacking early is not really an option.

The depth of team is obviously what wins the wars, and the main unknown element.

@Kerridoc I’m level 31 and don’t have a full ascended 4* team yet, would love to be any of 3 scenarios above :wink:.

Overall I think the Devs have done a good job on alliance wars. Also the game in general, we would all love more Mats and top heros, but if it came easily there would be nothing to work towards.


#300

I’m reading through all the comments and finally come across someone that might have the answer I am looking for with regards to why war scores are adjusted based on past performance. (hallelujah).
I’m puzzled as to why past performance should matter if the depth of our hero rosters are the same?
@alyssaann1014 mentioned you might have more on this?


#301

Just imagine two different alliances with exactly the same heroes and troops:

Alliance 1) Everybody is a skilled player and uses every flag in war. They know how to pick teams to counter defenses, and are great at working boards.

Alliance 2) People are casual, and not great at picking heroes for teams or in working boards. Some use all 6 flags, some use 3, and some use just 1 flag.

If we pair these two alliances purely based on heroes and troops, will it be a fair fight?


#302

I keep reading these posts, and am amazed that people still believe that cups, titans, or alliance scores have anything to do with the war score and matching process.

I am in an alliance where every member fights every single war…and while we do see teams with members that have joined in the past few days, after actually talking to a couple of these players…most come from a group of alliances, and they regularly move around within that group. So far we have been fairly evenly matched…but after winning 2 or 3 we do tend to get a very strong opponent and usually lose…which is what I would expect.

You aren’t going to win every war…but it seems more and more that when people lose, they immediately want to blame the matching process or some sort of cheating going on. As @Garanwyn pointed out…RNG or not…skill, participation, and experience can be huge in whether the war goes your way or not.

It makes no sense to me (the alliance-hopping)…your war chest participation goes to zero %, but if that’s how they choose to play then so be it.


#303

Thank you for your response @Garanwyn.

So… Should there be “levels” of war perhaps along the lines of rare, epic and legendary battles in the current event quest?

I’m new to this game but I would expect when people tick to participate in war they actually use their flags. If they don’t is that not more of an alliance leadership responsibility to communicate with members and take action accordingly? This may link to a thread I’ve read about leadership being able to select who will participate in war.

I digress… Instead, the solution is a mathematical formula to “adjust the war score based on past performance”? The fact that the war score is “adjusted” implies it will be to the benefit or detriment to one alliance.

Again… How can that be fair? Is there really a need for the “adjustment”?