War Algorithym Broke??

Reddit posts gave us an idea of how many tier 1 alliances were getting easy wins against tier 2-3 alliances last AW Season. The trend appears to have increased. If the matchmaking process can't be fixed, changing the # of alliances in high tiers may be the answer. For example, change tier 1 from the top .1% to the top .2%, tier 2 from the top .2-5% to the top 3-6%. and tier 3 from the top .6-1% to the top .7-1%. That would double the amount of same tier matchups for tier 1 alliances without negatively impacting lower tiers. Not fixing this problem reduces MCOC revenue.

unknown

This content has been removed.

DTMelodicMetal wrote: »

Reddit posts gave us an idea of how many tier 1 alliances were getting easy wins against tier 2-3 alliances last AW Season. The trend appears to have increased. If the matchmaking process can't be fixed, changing the # of alliances in high tiers may be the answer. For example, change tier 1 from the top .1% to the top .2%, tier 2 from the top .2-5% to the top 3-6%. and tier 3 from the top .6-1% to the top .7-1%. That would double the amount of same tier matchups for tier 1 alliances without negatively impacting lower tiers. Not fixing this problem reduces MCOC revenue.

I'm not sure how that addresses the problem. It sounds like what this would change is that tier 2 alliances currently being matched against tier 1 alliances now, would just be called tier 1 alliances next season so that the match was "within tier." But won't they still be just as mismatched in real terms as before?

DNA3000 wrote: »

I'm not sure how that addresses the problem. It sounds like what this would change is that tier 2 alliances currently being matched against tier 1 alliances now, would just be called tier 1 alliances next season so that the match was "within tier." But won't they still be just as mismatched in real terms as before?

The difference is current tier 2 alliances would receive the tier 1 multiplier. Mismatches would decrease because tier 1 alliances would (ideally) stop matching against tier 3 alliances and current tier 2 alliances would receive increased rank rewards, thus allowing them to compete more equally with current tier 1 alliances over time. I might be wrong, I wouldn't be surprised if other players gave feedback that proved this suggestion to be invalid.

Working exactly the way those alliances want it to work.

Rektor wrote: »

Working exactly the way those alliances want it to work.

@Rektor You're a tier 1 veteran, you think changing the war tier % ranges could do any good?

Zuko_ILC

This has been brought up many times and doesn't look like a fix will ever happen.

DTMelodicMetal wrote: »

Rektor wrote: »

Working exactly the way those alliances want it to work.

@Rektor You're a tier 1 veteran, you think changing the war tier % ranges could do any good?

They need to remove points for defenders remaining or change it so that defenders remaining reduces the score of the losing alliance instead of inflating the winner’s score. That’s the only reason these alliances do this garbage.

There’s half a dozen other changes they need to make to fix this matchmaking nonsense but that’s the main one.

Then maybe they can figure out a way to force master alliances to play each other. Apply a strength of schedule multiplier to scoring and watch how hard everyone searches for top competition instead of this manipulated matchmaking shtshow going on now.

Oilliphéist

The algorithym should also be limited in the spread of war rating. The master alliances all searching at different times now hurts the lower rated alliances and allows them to match a wider spread of alliances instead of matching each other. If the spread was limited and master alliances ended up missing out on a match during the season it may cause them to start searching earlier and more often around the same times, also allows other alliances to pair up against more competitive wars of similar skill levels.

DTMelodicMetal wrote: »

DNA3000 wrote: »

I'm not sure how that addresses the problem. It sounds like what this would change is that tier 2 alliances currently being matched against tier 1 alliances now, would just be called tier 1 alliances next season so that the match was "within tier." But won't they still be just as mismatched in real terms as before?

The difference is current tier 2 alliances would receive the tier 1 multiplier. Mismatches would decrease because tier 1 alliances would (ideally) stop matching against tier 3 alliances and current tier 2 alliances would receive increased rank rewards, thus allowing them to compete more equally with current tier 1 alliances over time. I might be wrong, I wouldn't be surprised if other players gave feedback that proved this suggestion to be invalid.

It is true that more alliances would get the tier 1 multiplier. But that doesn't address my central question which is how does this address or reduce mismatches. If tier 1 alliance A matches against tier 2 alliance B right now and for whatever reason we decide that is a mismatch, how does calling alliance B a tier 1 alliance by fiat alter the fact that that match up would still look like a mismatch.

It sounds like you're continuing to allow mismatches to happen and hoping that by just handing out more rewards to tier 2 alliances (by essentially giving them more season points and more individual war rewards) the increased rewards will make them stronger and reduce the mismatch gap. But that just kicks the problem down the road, because those increased rewards will also advance alliances that are currently lower than that into the same range that mismatches are occurring.

In other words, just by way of example if 3400 alliances are matching against 2800 alliances and we decide that's a mismatch, your suggestion seems to boil down to handing more rewards to 2800 alliances hoping they get strong enough to become 3100 alliances and reduce the mismatch gap. But that is bound to also make 2400 alliances strong enough to become 2800 alliances and fall victim to the same 3400 vs 2800 mismatch. The numbers are just to simplify the example: not to specifically assert what is and is not a mismatch.

Rektor wrote: »

They need to remove points for defenders remaining or change it so that defenders remaining reduces the score of the losing alliance instead of inflating the winner’s score. That’s the only reason these alliances do this garbage.

That would put significantly more pressure on all alliances to fully explore the map at all costs. RIght now a remaining defender propels your opponent upward but doesn't directly move you downward relative to peer alliances. Under your change suggestion remaining defenders are a significantly worse penalty for the alliance that doesn't kill all the defenders. I don't believe that is a good change.

Hamin

Wow.

That is some of the most lopsided point increase/decrease I've ever seen.

My condolences.

LordSmasher

Victory bonus should depend on relative strength of alliances. Beating a higher tier alliance gives more points so killing small fry is less attractive.

Hamin

LordSmasher wrote: »

Victory bonus should depend on relative strength of alliances. Beating a higher tier alliance gives more points so killing small fry is less attractive.

Like alliances have a choice?

LordSmasher

Hamin wrote: »

LordSmasher wrote: »

Victory bonus should depend on relative strength of alliances. Beating a higher tier alliance gives more points so killing small fry is less attractive.

Like alliances have a choice?

There is much discussion about top alliances purposely avoiding each other by coordinating their searches. At the pointy end they do have some control due to the pool sizes.

Hamin

LordSmasher wrote: »

Hamin wrote: »

LordSmasher wrote: »

Victory bonus should depend on relative strength of alliances. Beating a higher tier alliance gives more points so killing small fry is less attractive.

Like alliances have a choice?

There is much discussion about top alliances purposely avoiding each other by coordinating their searches. At the pointy end they do have some control due to the pool sizes.

This is why we can't have nice things.

I wish people would just play the damn game.

LordSmasher

Wouldn't blame the people as the incentive is driving behavior and there is a strong incentive for the top alliances not to face their peers which seems counter to the goal.

They could do the wars in slots.
Aggregate all requests over a period of time, say an hour, then do the matching making and start the wars then. Would even things out.

LordSmasher wrote: »

There is much discussion about top alliances purposely avoiding each other by coordinating their searches. At the pointy end they do have some control due to the pool sizes.

This went on for years before AW Seasons started. If it wasn't against forum rules I'd post a screenshot of Line chats that show how "top" aka piloting alliances used to coordinate war matchmaking. Many players including myself thought AW Seasons would eliminate that kind of manipulation. We were all wrong.

DNA3000 wrote: »

It is true that more alliances would get the tier 1 multiplier. But that doesn't address my central question which is how does this address or reduce mismatches. If tier 1 alliance A matches against tier 2 alliance B right now and for whatever reason we decide that is a mismatch, how does calling alliance B a tier 1 alliance by fiat alter the fact that that match up would still look like a mismatch.

It sounds like you're continuing to allow mismatches to happen and hoping that by just handing out more rewards to tier 2 alliances (by essentially giving them more season points and more individual war rewards) the increased rewards will make them stronger and reduce the mismatch gap. But that just kicks the problem down the road, because those increased rewards will also advance alliances that are currently lower than that into the same range that mismatches are occurring.

In other words, just by way of example if 3400 alliances are matching against 2800 alliances and we decide that's a mismatch, your suggestion seems to boil down to handing more rewards to 2800 alliances hoping they get strong enough to become 3100 alliances and reduce the mismatch gap. But that is bound to also make 2400 alliances strong enough to become 2800 alliances and fall victim to the same 3400 vs 2800 mismatch. The numbers are just to simplify the example: not to specifically assert what is and is not a mismatch.

You may be right, there're a ton of variables to consider when looking at how to fix tier 1 alliances manipulating matchmaking to earn master rewards from steamrolling lower tier alliances. Someone recommended something I didn't initially agree with but now do. Take matchmaking out of players' hands and restructure AW to run somewhat like AQ, start war at a set time every week and assign alliances opponents based on war multiplier.

My alliance has been matched against several opponents who've had drastically different time zones than our own. Many of these alliances didn't move until war had been going for 4-6+ hours, which means matchmaking being done by Kabam would work even if alliances slept through the first few hours of war. Furthemore, AQ opens at the same time for all alliances and everyone handles that fine. The matchmaking process being left to players may be too broken/corrupt to be fixed at this point.

Kerneas

My alliance has been picked impossible matches for so long, that we were forced to fall from tier 6 to tier 14. We are ca. 8,5 Mil. alliance and we always get 11-13 Mil. opponents.

Just a small difference you may say, but it is not true. Their defense is very challenging for us, while ours is just piece of cake, becuase their champs are so much stronger. For us, it's like Act 5 gameplay with 4* 5/50 champs. For them it must be like playing Act 4 with 5* 4/55 ones.

mum_m2

I'm welcoming the challenge but 96 and 2 is the largest i've ever seen. Not complaining about it, just saying. Edit: I feel that in order to be the best you gotta beat the best. Gotta find a way to win these types of wars or never be in the higher ranks.

DNA3000 wrote: »

Rektor wrote: »

They need to remove points for defenders remaining or change it so that defenders remaining reduces the score of the losing alliance instead of inflating the winner’s score. That’s the only reason these alliances do this garbage.

That would put significantly more pressure on all alliances to fully explore the map at all costs. RIght now a remaining defender propels your opponent upward but doesn't directly move you downward relative to peer alliances. Under your change suggestion remaining defenders are a significantly worse penalty for the alliance that doesn't kill all the defenders. I don't believe that is a good change.

Then give master alliances the tier multiplier of the weaker opponent in each war.

Wonder how fast tanking your war rating and matchmaking manipulation would end if people were afraid of accidentally matching a x6.2 multiplier opponent instead of x7.

I can assure you that you wouldn’t have 3500 war rating alliances sitting there with a fraudulent 3100 rating while hunting 2800 rating opponents.

mum_m2

Rektor wrote: »

DNA3000 wrote: »

Rektor wrote: »

They need to remove points for defenders remaining or change it so that defenders remaining reduces the score of the losing alliance instead of inflating the winner’s score. That’s the only reason these alliances do this garbage.

That would put significantly more pressure on all alliances to fully explore the map at all costs. RIght now a remaining defender propels your opponent upward but doesn't directly move you downward relative to peer alliances. Under your change suggestion remaining defenders are a significantly worse penalty for the alliance that doesn't kill all the defenders. I don't believe that is a good change.

Then give master alliances the tier multiplier of the weaker opponent in each war.

Wonder how fast tanking your war rating and matchmaking manipulation would end if people were afraid of accidentally matching a x6.2 multiplier opponent instead of x7.

I can assure you that you wouldn’t have 3500 war rating alliances sitting there with a fraudulent 3100 rating while hunting 2800 rating opponents.

maybe, I think if you average it out you'll be more prone to battling the best though. that will help the smaller tier gain points, and make it a ton more competitive. if my alliance faces another with a higher tier, those points should be more meaningful for us, and less meaningful for them.

mum_m2 wrote: »

Rektor wrote: »

DNA3000 wrote: »

Rektor wrote: »

They need to remove points for defenders remaining or change it so that defenders remaining reduces the score of the losing alliance instead of inflating the winner’s score. That’s the only reason these alliances do this garbage.

That would put significantly more pressure on all alliances to fully explore the map at all costs. RIght now a remaining defender propels your opponent upward but doesn't directly move you downward relative to peer alliances. Under your change suggestion remaining defenders are a significantly worse penalty for the alliance that doesn't kill all the defenders. I don't believe that is a good change.

Then give master alliances the tier multiplier of the weaker opponent in each war.

Wonder how fast tanking your war rating and matchmaking manipulation would end if people were afraid of accidentally matching a x6.2 multiplier opponent instead of x7.

I can assure you that you wouldn’t have 3500 war rating alliances sitting there with a fraudulent 3100 rating while hunting 2800 rating opponents.

maybe, I think if you average it out you'll be more prone to battling the best though. that will help the smaller tier gain points, and make it a ton more competitive. if my alliance faces another with a higher tier, those points should be more meaningful for us, and less meaningful for them.

There’s a logical solution to be found. Current system is an embarrassment to competition.

Rektor wrote: »

DNA3000 wrote: »

Rektor wrote: »

They need to remove points for defenders remaining or change it so that defenders remaining reduces the score of the losing alliance instead of inflating the winner’s score. That’s the only reason these alliances do this garbage.

That would put significantly more pressure on all alliances to fully explore the map at all costs. RIght now a remaining defender propels your opponent upward but doesn't directly move you downward relative to peer alliances. Under your change suggestion remaining defenders are a significantly worse penalty for the alliance that doesn't kill all the defenders. I don't believe that is a good change.

Then give master alliances the tier multiplier of the weaker opponent in each war.

Even better might be for everyone to get their opponent's multiplier. In other words, if a tier 2 faces a tier 1, then the tier 2 alliances gets the x7 multiplier and the tier 1 gets the 6.2 multiplier.

Right now, the "meaning" of the multiplier is sort of a reward for winning a lot. You win, so you increase your rating, so you get a higher multiplier, so you get more season points. The bigger haul of season points is the "reward" for winning more often, and that reward is delivered with the higher multiplier. Instead, we could change the meaning of the multiplier to be "this is how hard it is to beat this alliance, and thus how valuable every point scored against them is." The multiplier would then not be how much points you get, but rather how much points your opponents get for beating you.

There is still an advantage to fighting lower tier opponents, because the victory bonus is larger than the multiplier differential between war tiers (the 50k victory bonus is at least a 33% bonus on a "perfect" score of 149,100 points before remaining defenders are counted). But at least any alliance fighting equal opponents will have an edge on alliances fighting lower opponents if they can win often enough. It reduces, but doesn't eliminate the incentive to punch down.

The victory bonus itself should probably also scale with competition. It should be something like 40k beating even competition, 60k beating superior competition, and 20k beating inferior competition. That would also disincentivize fighting lower opponents.

Rasilover

Same here

Rasilover