War MM & Cannon Cart changes - the FEEDBACK thread!

[JSpur5]

this is the kind of feedback/info i have been looking for.. do i need to get 5s to drop bows? or ITs? where can i read up on how the MM changes impact .5s?

JD, while there is nothing stating exactly what the algorithm changed, several of us have noticed that putting down all defenses on a base and eliminating the .5s and anything that can be construed as engineered has resulted in much fairer matchup.

I am currently in a war in which neither clan has any engineers or .5s in the lineup and both clans have a fair shot at winning the war. The last 3 war searches ive done using an all maxer base lineup has resulted in an even playing field for both clans. We are currently running 4th11s, 4 th9s, 4 th8s, 2 th7s and a th6 all complete with the full compliment of defenses.

As stated in another post, it's a small sample size(3 wars) but it is what seems to be working for us. I'll be doing another war search tonight upon completion of the current war with the same type of lineup and I'll report back what the search yields.

Hope this helps. You can check me out at #PLRYU20U.

[SheerluckHolmes]

Actually, in our 5th war(I think), after update, we actually still have a chance to win despite being lower both on offense and defense (2 th11 their side with level 1 EA and level 3/4 infernos vs no th11) because they are not organized(hitting mirror and one below).


Of ♥♥♥ thinks we have had a fair match just because of a win or doesn't fix war MM soon, I'll just have to leave a 1 star rating and "Don't download" comment on Google Play.

[Slayer339]

Then you prefer SC put in no absolute checks like equal number of th11s and instead keep working on trying to weight everything properly and trying to avoid engineers taking advantage. Fair enough. I was totally with you a year ago, less with you a few months ago and less than that now.

There absolutely needs checks in place, IF equal number of the highest TH is the best they can do I can live with that and it would be an improvement over current state. I think they could do better by evaluating key components of the highest town hall, such as current lab level (troops & spell) the spell factory level (number of spells) etc...but maybe that's too difficult to program...

[2222]

If it is close was the match really that uneven? If the uneven part comes down to a single star then the war is down to one mistake, one heroic attack, one fat-fingered queen ability, one well placed giant bomb. In short, what we want is a close war. The goal of a good MM is to produce this. Can you think of a better metric to measure whether the MM is making things better or worse?

SUPERCELL, please publish some metrics on clan wars. Percentage of wars decided by the tie-breaker before and after the modification, percentage of wars won by 5 or more stars etc. Hard data will go a long way towards informed discussion.

Yes, for example, I have been in 10v10 wars where I know there was 0 chance for the opponent to win because the best they were going to do was lose to us by 1 star. Other times, I have been in wars where there was a significant obvious mismatch, but the disadvantaged clan was much more skilled at attacking and base designing, making the score close. I don't want just a "close" war. I want a war that is determined as much as possible by the attacking and designing skills of the players. If we face a very unskilled clan, we should beat them by a lot. It shouldn't be close because of some advantage they have in offensive power, for example.

There absolutely needs checks in place, IF equal number of the highest TH is the best they can do I can live with that and it would be an improvement over current state. I think they could do better by evaluating key components of the highest town hall, such as current lab level (troops & spell) the spell factory level (number of spells) etc...but maybe that's too difficult to program...

Yeah, I think they need more than just this one check, but making both clans have the same number of th11s would be a good start (which means if you have 0 the other team also has 0).

[Hellectrik]

I just posted a good suggestion for adding more fairness, even in unfair war match up, in the idea section.

Link

Thanks for adding to the conversation and debate.

[STRIKER44]

If it is close was the match really that uneven? If the uneven part comes down to a single star then the war is down to one mistake, one heroic attack, one fat-fingered queen ability, one well placed giant bomb. In short, what we want is a close war. The goal of a good MM is to produce this. Can you think of a better metric to measure whether the MM is making things better or worse?

SUPERCELL, please publish some metrics on clan wars. Percentage of wars decided by the tie-breaker before and after the modification, percentage of wars won by 5 or more stars etc. Hard data will go a long way towards informed discussion.

Win/loss by one star has no relevance on how good the match making was at all. When we win a war with only 3 th 10 vs their 2 th 11(large heros and max war troops ) and 1 th10 had nothing to do with how even the match was. We would have easily been beat if their 1 and 2 had not left their clan and then came back and hit our 8 to 10 ( was funny to watch)From what I understand your trying to say seeing how we won, it must of been a good match, which was not the case.

[Bigtone2001]

War #6, and biggest mismatch by far. (Getting worse each time)

us
1-11 TH11 with EA
12 TH11 no EA, GW 12. (New th11)
13-22 TH10 with 2 infernos
23 TH10 with 1 inferno
24-26 TH10 no inferno (new th10)
27-30 TH9

No engineered bases, all at different stages of upgrading normally.

them
1-13 TH11 with EA
14 TH11 no EA, GW 20, Max troops
15 TH11 no EA, GW 20, Max troops
16 TH11 no EA, GW 14, Max Bowler, Witch, Valk & Giants
17 TH11 no EA, GW 14, Max Bowler, Healer, Valk, Wiz & Giants
18 TH11 no EA, GW 15, Max Bowler, Valk, Healer & Giant
19 TH11 no EA, GW 13, Max Bowler, Healer, Valk & Giant
20 TH11 no EA, GW 17, Max Bowler, Valk, Hound, Loon
21-26 TH10 with 2 infernos
27 TH11 no EA, no infernos, GW 15, Max Bowler, Valk & Healer
28 TH11 no EA, no infernos, GW20, Max Bowler, Witch, Valk, Healer & Giant
29 TH10 no inferno
30 TH9


C'mon.. Clearly 100% engineered clan, they outnumber us by 10 TH11s all with GW & Max Troops.

In 6 match ups we have had 1 that was even.
I can already see what our 7th war match will be, 30v40...

This algorithm clearly isn't working!!!

War started with their #28 getting 3* on both our 13 & 14, so we figured we were in for a hammering.
We got off to a good start with our 11s getting 2* on their top 11 first go (4 & 12 were 3*) so freed our 2nd attacks for lower 3*.
war result 71 (79.67) v 71 (79.07) to us.

Even with SC's broken MM we scraped the Win.

War #7 (search was under 20 mins) 30v30

Yest again, matched against engineers.
us
1-14 Th11
15-26 TH10
27-30 TH9

them
1-21 TH11
22 TH10
23 TH11 (34/35/16) Max Hound, Loon, Barbs
24-25 TH10
26 TH11(20/20/13) Max Bowlers, Hound, Loon
27-30 TH10


so 9 extra TH11s all with GW and all with varying Max troops. (Bowlers, Witches, Hound, Loon, Healer, Valks, Hogs, Giants etc)

yet again, a non engineered clan hits heavily engineered.


Surely SC seriously can't be classing a new TH as engineered.
If they expect a new TH to drop every defence before they can get an even war they are bonkers.
at th11 that'd be cannon, archer tower, wiz tower, xbow & eagle.
cant remember all the extras at TH10.

Simply give the TH a base rating so it's above a Max of the one below. Does the same thing, just allows people to upgrade as they can.
should be the easiest fix, it's 1 variable, but has the biggest impact.

currently they allow a TH11 to have Max warden and troops, but by not placing defences, give it the same weight as a TH9.

[MarcelLohfink]

The new MM is just awful. Its our 6th war I think and 5 times we got heavily engineered opponent clans. On top it always only takes 15-25 minutes to find a matchup. Why the hell dont let it search for a fair war 50,60,70 min.??!!!? No one cares to wait a little longer to avoid this ♥♥♥♥♥♥♥♥.
Our current war is 15vs15:
we have 2 th11, 6 th10
they have 9 th11, 4 th10
Sry...are you serious?
20 min and your ****** algorithm thinks "yeahhh this looks good"?!

[TheLandofNod]

Yep.... Darian... please change the MMA back to before the update...
(Sarcasm - it's not working the way you intended).

Every war, it gets worse... for my clan its going in the opposite direction to equilibrium.

For our current war, I've managed to persuade some 10s to participate. (They've been sat out due to the horrible matches we get).

First war back and they've been smacked in the face.

Us: 2 x TH11, one max, one relatively new. 4 x 10s, 1 x 9.5, none max. 5 x th9, 2 th8's.
Them: 6 x TH11, 1 of which is an engi 11, 3 x 10s, 1 of which is an engi 10, 3 th9s, 2 th7's.

EAs are matched but they have 10 gw atts to our 4.
They have two more inferno base - Their no.1 is almost maxed th11 and all their next 4 11s are max th10 defence.
Even their 9s are strong defensively.. bottom 3 are junk.

So if they can attack, its only likely our no. 1 will not be tripled.
The only difference I can see is we have a mild hero advantage...

Super duper eh...

[Bigtone2001]

Summary of War #7 on offence.

Total troop levels
Us / Them
king - 1002 / 1138
queen - 1057 / 1171
warden - 245 / 427

bowlers - 63 / 78
witch - 70 / 75
hound - 94 / 109
loon - 191 / 202
healer - 130 / 140

They have 9 more TH11s than us, and as you can see high wardens.
Clearly engineered clan as only 9 out of 23 have EA. (23 out of 30 are TH11 !!!)
Most have low inferno or none also.

We have a clear defence advantage, but this is so overpowered by their attack strength.

Clear proof that offence strength is NOT considered high enough in algorithm.