Printable View
Below is the recently published Video by Clash with Ash where he interviews Darian (SC Community Manager) about the new upcoming changes to the war matchmaking system :-
https://www.youtube.com/watch?v=Qei_m6cL5mM
Lots to digest! Note : It is NOT a brand new algorithm that they've developed from scratch (as the community kind of believed anyway), but they did add additional variables (more checkboxes!)
:)
Edit : Also a suggestion (22.00) that they are working on a 'possible' solution to Clouds!
If they're going to segregate engineers to only face other engineers in war then to be fair, and not punish them for playing this way since they say they don't want to restrict player options, then they should make the same algorithm change to multiplayer. No maxer should have the option to attack an engineered base in multi if they're unwilling to face them in war.
Quote:
Originally Posted by Shanahan
'Unwilling to face them in war' implies that the Non Engineer is afraid of a fair fight? I think you will find the desire for a lop-sided, unwinnable war for their opponent is the sole reason Engineers indulge in their nefarious* practises.
*The use of the word nefarious is a personal view and in no way is meant to imply that Engineers do not love their grandmothers or kittens. They are in fact fine and upstanding people who simply abhor the violence associated with a fair and even scrap. :smirk:
OMG, hahaha. Enough said!Quote:
Originally Posted by Holps
Why is it a punishment for engineers to match those with similar builds in war? I don't understand what you mean there.Quote:
Originally Posted by Shanahan
I can't see that the MM change will restrict player options in any way, just that it aims to match like to like in war in terms of real offence and defence strength. That's what it was always trying to do, and engineers are people who have been building in ways that aim to affect that matchmaking, specifically to pull a weaker opponent than they otherwise would have got, one who isn't in fact a good match for their overall offence and defence profile but whom the matchmaker was matching previously. It's no punishment to have that change such that you'll match more realistically, but rather the whole point of a matchmaker.
Multiplayer is an entirely separate play mode to war - the base that has intentionally not built or upgraded things that are there to be upgraded is simply facing the consequences of that war-advantage-oriented build choice for the other play mode, and is free to change development direction at any time.
Hope your mouth isn't too dry after spitting your dummy out.Quote:
Originally Posted by Shanahan
As a non engineer, I am more than happy to never face one again either in or out of war.
If we really want to 'even' things up then perhaps engineers should be blocked from upgrading defences forever, afterall no non engineered players can turn back the clock to become an engineer.
thanks to skrags
but darian.. what the ...
a video interview?
how about:
change logs
documentation
anything written is 100000 > a video
a video is not searchable, index-able, can disappear at the whim of the creater in a second, subject to interpretation, etc..
i can think of all kinds of things wrong with this.
Sorry my man but you won't get much traction with this post. You chose to engineer so that you can get easier match ups. Granted, not every engineer chose to go that route for that reason. But you also have to accept the consequences of this route. I often see engineers say "take responsibility for your actions." That also applies here as well. So I'll be stealing all your loot to my hearts content and try not to use 30% or more of my army lolQuote:
Originally Posted by Shanahan
How many times do we have to disprove this before it becomes a violation of rule 1 and rule 21?Quote:
Originally Posted by Holps
The truth should never be a violation.Quote:
Originally Posted by littledoctor
Thank you Holps.
We just want a fair fight. We don't care how you fight amongst your like minded friends.
Well, to cite rule 1, moderator decisions are final in such matters... so it's a problem of the rule violating kind when mods decide it's that and treat it as that... and it's probably a bad idea for any of the rest of us to go too deep into or meddle with what it isn't our role to decide.Quote:
Originally Posted by littledoctor
To be honest, I think you only think you've proven it, rather than having actually proven it... the usual argumentative standard is that you're able to convince a a fair-minded reasonable person that it's the better position to take, and I've not seen that.
Lol, you cannot win this argument when there are engineers that have flat out said that they enjoy winning war before it even starts. Of course not every single engineer has this perspective but if you think none do then you are just naive.Quote:
Originally Posted by littledoctor
Recheck the quote that I claim is false: "the desire for a lop-sided, unwinnable war for their opponent is the sole reason".Quote:
Originally Posted by Tomville
I don't have to show that nobody does it for that reason. Obviously some do. In order to disprove the statement, all I have to do is find one person who has any additional reason other than the desire for a lop-sided, unwinnable war.
And I have found one such person: me. Not that I'm unique; there are many who do it for other reasons. The GOBS ONLY(tm) clan is famous for doing it for the lulz. Lots of people on these forums have said that they have started engineering, some reluctantly, in order to compete with engineers. Are they all lying, and they're really doing it to produce lopsided unwinnable wars?
This has been discussed for ages. It is obvious by now that the quote is not true.
You have just disproven it yourself, by saying that not every single engineer has this perspective.Quote:
Originally Posted by bouges
I need a Denton graphic that deals with unwinnable wars or wars that are won before they start.Quote:
Originally Posted by bouges
What is the actual percent of these that you believe are really happening?
Specifically, unwinnable? As it is impossible to have a war won before you start.
Darian, thanks for the interview you gave.
I believe many will want to hear now when things will start happening, i know you cant give an exact timeline and i don't ask for it, but more or less an estimation (days? weeks? months? next update whenever that is?)
When will we have "clear" release notes? (XXX tower HP increased by zz%) instead of: XXX tower HP increased
And that kind of things, no more "hidden and undocumented" mechanics changes of units (recently: loons, witches)
MM changes?
mid-high end game updates?
"new features" you mentioned?
rough estimation will also be appreciated.
Tnx.
Quote:
Originally Posted by littledoctor
Mmmmm...OK. I was actually indulging in a little light hearted hyperbole in response to the OP's point. I am happy to concede that some people have indeed started engineering in response to engineering. The odd player may just engineer because they want to Clash but are pacifists so only want one cannon in their base:smirk:. Personally I wouldn't indulge in the dark arts if I was the last Clasher in the room with an Eagle...but I'm stubborn.
As for rule 1 let's not be silly. Nothing I posted is remotely on the level of abuse or personal attacks. Engineers who choose the Clan War forum to come hang out in are very well aware that a majority of people here do not like what they do even if they do have some sympathy with why they started it. I have said many, many times that I feel Engineers are a bigger threat to Clash than Modders ever were. That's my view. It will never change. It has nothing to do with rule 1.
Can you explain why you feel that Engy is greater threat than modding?Quote:
Originally Posted by Holps
When I read it initially I didn't feel it was light hearted hyperbole at all. If you want light hearted hyperbole you should have used the word "filthy".Quote:
Originally Posted by Holps
Good to see, I am interested how this works
I have a almost maxed base TH11 but I also have a rushed TH11 with max dragons and loons, the rushed base is very fun and I see why a lot of people do it, in war, it is really fun being the lowest or 2nd lowest option and moving up 10-12 spots and smashing a base
If it REALLY effects match ups too badly, I probably will stop warring with my rushed base
I am really interested to see how the new MM works if there is only 1-3 engineered bases and 1-3 .5s and see if the MM picks this up.
I feel the new MM is only going to put full engineers against full engineers
I really have nothing to back that up, but I have a feeling its going to happen
i had a different takeaway from the interview.Quote:
Originally Posted by sheedy17
Minimal changes.
I'll give it a shot.Quote:
Originally Posted by Ripple1972
I truly can't say I represent all casuals, but I am indeed very casual. I have played this game for years (yes, many Christmas trees) and have never seriously considered leaving the game until recently. My clan and I really enjoy war. We may not be the best, but we enjoy a good (fair) fight. We will not continue to war in an environment in which we have to be perfect in order to tie on a regular basis. So, if the game continues down this path, you lose us, and it's only my opinion, but I would guess you would lose many more. I do not see the game making it very long with strictly an engineering environment. I truly think most are here for base design and attack strategy.
I truly believe if they do indeed curb this practice it will be wonderful for the long term health of the game. I do believe that if the engineers are forced to fight one another, they will tire of the common ties. If they lose their advantage, they are likely to change their style of play as well. This is where it gets even better. We will begin to see good and fair matches much faster. Now don't get me wrong, the engineers will be formidable in a very short period of time. Many are very good attackers and base designers. I look forward to the challenge. You are welcome to crush my base from same level anytime.
This is similar to how I feel too. I'm a bit more than casual but I can certainly feel for the more casual clans/players. Esp the ones who have no idea how the mm works but just want to have even wars, have a legit shot at winning, and just have fun. Continuing down this road extreme engineering will only push a lot of those casual players out. Not to mention it'll make it difficult to bring in new players and force them down the engineering pipe so that they can compete. Not everyone wants to be that hardcore war clan that wins at all costs. And losing the casual players will truly hurt long-term for the game.Quote:
Originally Posted by Playforwar
Anyone got a cliff's notes version? Not watching a 40 minute interview.Quote:
Originally Posted by jobob145
Hey guys! It's a long video and I really think it's worth listening to in its entirety. But just in case you're short on time, here are timestamps for some of the big topics we discuss:0-9:00 -Quote:
Originally Posted by TankSinatra
Getting hired at Supercell9:03 -
Matchmaking and engineered bases18:17 -
Builder Base20:10 - Future updates / game direction25:50 -
FairPlay and modding 34:45 - Future of CoC as esport
I took away -
He mentions keeping search time low.
He mentions aggregate clan matching. ( ash asked specifically )
He uses "total weights" a few times.
Not built from scratch, just more tweaking.
Therefore, imo,
It will still be broken.
OK. :)Quote:
Originally Posted by Holps
"Inflammatory" is the part of rule 1 that caught my eye. But I guess that strict enforcement of that would essentially close the forums...Quote:
Originally Posted by Holps
Also, he mentions "spirit of play" which is the opposite of engineering, unless imbalanced matches is within the spirit of the play.
he did say he wished there was more transparent patch notes.
Lot of positive things in there however. builder base section was good.
The unwillingness to explain the math in general terms, such as a line fit, or what general algorithm is being used, makes me think it is probably going to be deficient.
Thanks, 9:00 to 18:00 I can do. Nobody else at the office yet.Quote:
Originally Posted by ryryshouse
Interesting, I was sure matching was done base by base (though with enough leeway to not be obvious). That's what I believe they've said before, and that's the way it seems to work when you roster engineer - anchor plus trash tends to match anchor plus trash. But maybe anchor plus trash is just a common lineup, often unintentional by casual clans.
Yeah think you're right it will still be broken. But if they're putting a bit of lead in the nerfbat vs the masses of lopsided accounts, it'll be an improvement at least.
ETA Got the lane-by-lane thing from the March 16 update notes - "Additionally, Clan Wars matchmaking has been fitted with new routines aimed at better balancing entire teams, rather than just individual lanes. It will place greater emphasis on matching the top of the war map, and less emphasis on matching the bottom. Total team defensive strength will also be analyzed to help filter out war matchups that are imbalanced on a map-wide scale." I interpreted as base by base, with additional filters after.
I totally agree. I feel that engineering is driving more casual players away from war, which is also driving them away from the game in general.Quote:
Originally Posted by Playforwar
And it's not just the extreme engineers, we are currently in a war with only 1 engineered base, but that one base in a 10 man war means that we have to be perfect as they have 2 extra TH 11 attacks and 2 extra TH 10 attacks. That's huge and if we are perfect I still have to get a high percentage 2 star on a max base where mine is not.
I assume that a people spend a good chunk of their gems on rushing heroes for war, SC may have noticed by now that less people are gemming them. Why would you spend gems rushing back to something you are really not enjoying anymore.
I for one no longer gem my heroes. I'll wait out the week. More and more of my clan mates are doing the same. We are down to 10 man wars because people are not gemming heroes anymore.
Could that be the motivation for these changes after being ignored for so long? Doesn't really matter, but I would love to know how many players they lost to this issue.
Don't think I have seen any threads with people complaining about matching GOBS only clans.Quote:
Originally Posted by littledoctor
It sounded to me like they are doing a sort of two list system. Seems like they will be totaling offense and defense respectively. Not matching on a list, but still on totals with offense having more significance then it has had in the past. And they will maybe be matching clans with similar totals proportions of offense to defense.
Happy to Ripple but to be honest this post by Playforwar is a very good comment regards this issue. https://forum.supercell.com/showthre...1#post10742148Quote:
Originally Posted by Ripple1972
For me the threat is that a vast majority of Clashers war. Not all, but I think we can all agree those that simply farm or push are not a huge bulk of players? Of that war playing base only a tiny % could genuinely be considered as hard core war professionals and the rest are average attackers who love warring for the drama and the fails, the wins and the losses. What Engineering is doing is increasingly presenting these players with wars where they simply cannot win. This is not anecdotal because I have personally suffered these dead wars many times now and it literally kills the war vibe in your clan for 2 days. Often clans won't even bother filling CC's or try to win when the match is absurd enough.
I agree that this has led to a group of Engineers who have gone down the 'Can't beat 'em, join em route' but the more people that do this the more casual war clans are alienated and beaten down in terms of their enjoyment of wars.
I am now on a level where most of the clans I war with are heavy with TH11's and Engineers dont factor as much but the new players joining Clash are the future of the game and the only possible way we will still be discussing this game with the 10 year anniversary cake on the verge is if they enjoy wars as they learn..and choose to stay. This is where Engineers carry the greatest threat to Clash. If those new players, who will normally only have 1 base, are rapidly turned off Clash and Clan Wars then slowly the game will lose it's future. They will not care why the clan they face have a defenceless TH11 they will just be angry that their top base is being smashed by the enemies number 25 and simply go and play Candy Crush.
That is why I truly believe that Engineers are a threat to Clash.
Traveling with limited servic and can't watch the video. I feel so in the dark!
Yes, he sounded much more anti-engineering (in a way) than the previous statement did. Although in the interview he was still saying people should be allowed to play how they wanted, there was much more of a tone that balancing wars was the most important matter.Quote:
Originally Posted by ryryshouse
Obviously we'll never know, and SC may not know themselves why players quit. But SC are giving the impression that it's enough to have gotten high-level attention in the company. They released that statement about the MM in game; now this interview focused largely on war matters - and was quite clear that BhB is on the back burner for a bit. The modding part was slightly in the PR-speak "we're always trying to improve" (and to be fair they HAVE knocked modding back a LOT compared to 18 months ago), whereas the matching part, to me, felt more like a "we're going to fix this".Quote:
Originally Posted by Antonioca
I'm still a bit dubious on the little of the "how" that they described in their statement. But the best part of the statement and interview is not so much the details, rather the overall fact that the issue does seem to have become genuinely (and fairly urgently) important to the team.
They clearly don't intend to say anything very specific. One possibility that occurred to me listening to the interview was about the 2 totals. We've never known what they do with the 2 separate offence and defence scores, but it's smelt like they merge them into a single score per base and then work from there. Darian may have been hinting that it may still do this for the lane-checking, but any clan total comparison would now be completely separate Off-Off and Def-Def comparisons.Quote:
Originally Posted by SuperStorm103
This is what I said in a similar thread in general:
I think that was an excellent interview, congratulations to Ash and Darian, but specially to Ash, I think he made a lot of relevant questions and he handled the interview very well. I really hope we'll see more of them in the future, it really lessens the gap between players and developers, in my opinion.
I know that a lot of engineers like to say that engineering isn't against the ToS and that Supercell isn't against it, specially since after Darian said that they don't want to invalidate any play-style, but the answer to the first question about this topic enlightened us more about how Supercell views engineering: gaming the system and not "in the spirit" (of the game, I assume).
9:15
Ash: In a nutshell, what changes will be happening that are going to kind of fix or at least put a band-aid on engineered bases?
Darian: As far as the actual algorithm and the number crunching and the actual weights themselves, we won't be able to go into - and that's for proprietary reasons. We want to make sure that players try to play, I guess, "in the spirit" as much as possible and I think if we were able to provide those numbers, that would create more gaming of the system - and that's largely what we feel engineering is: trying to find out where the strengths and weakness of the MM system is and utilize them to their advantage. I played a lot of miniatures war gaming [...] where trying to find out what the maximum amount of strengths are, in a meta game that allows you to win at any cost and I think engineering is kind of a similar... I don't want to say... it's not a problem, but it's a symptom of a system that allows something like that to happen.
Although I'm not very optimistic about what will happen when the update goes live, I think it's clear that Supercell is against the advantages provided by engineering and that they're trying to remove them, which would consequently "remove" engineering.
Again, congratulations for the interview and thank you. This is a big step in improving the communication between Supercell and the players.
I am very surprised such a strong choice of words was used on the engi subject. I'm not surprised that even after such a statement they say they wanna keep all styles valid. SC is trying to have its cake and eat it too. I don't expect a positive change in the MM if this is the path they choose. Hard to tell with such conflicting statements tho.
Meant to multi quote.Quote:
Originally Posted by Shanahan
Mate, it doesn't matter if they make you base invisible to maxers, you won't be defending when your base gets pillaged by an engineer. Why does it matter who takes your loot? It's clearly not important anyways as you offer it to the community for free.
Few thoughts :-
"Engineering is a symptom of a system that allows something like that to happen" - I'm glad Darian pointed this out for everybody.
"New matchmaking will look at "total offence and total defence weights" where before it was very defence heavy.
This MM will make it "much more evenly weighted across the board" and will look to find clans with "similar weighted bases"
Not sure what to make of this if I'm honest, plus while yes back in March last year it was very defence heavy, there was still a general consensus that X.5 had been nerfed somewhat back in October/December.
Was matchmaking "very defence heavy" back then?
Overall however, it does seem that the Devs are prioritising this right now, which can only be a good thing!
Yes, I was a bit surprised by the "defence heavy" comment, and I'm not convinced he hasn't got that mixed up. Our current war we could well be down 70k-80k on the top 8 bases alone - and it's not even a bad match (they don't have 'free' wardens and TH11 troops everywhere for once). It's hard to see how that happens if our better heroes and presumably therefore offence scores haven't counted for much!Quote:
Originally Posted by Skrags
no consensus in our clan about .5 being nerfed. still going strong with the right lineup i'd say.Quote:
Originally Posted by Skrags
also, if that's what he said "New matchmaking will look at "total offence and total defence weights" where before it was very defence heavy.
This MM will make it "much more evenly weighted across the board" and will look to find clans with "similar weighted bases"
not sure what he is getting at or how that changes anything.
for an overly simplified explanation if the weight is going to be 50/50 split ...
11 (offense) + 11 (defense) = 22 "balanced"
11 (offense) + 4 (defense) = 15 "engineered"
with 2 attacks vs 1 defense, i'll take being a "15" any day vs a "22"
if a 15 = a 15, and there are lots of different ways to add up to a 15 (1 score for 1 base still, right?)
i'll take my chances with the current status.
keep on trucking as normal with some "tweaks".
Sounds like well manicured engineering will be unaffected while the defenseless bases will be a bright red flag marking a clan as engineered.
!!!Quote:
Originally Posted by OnyxDS
yet another reason information like this shouldn't be put out in a "video"
very easy to slip up saying something.
way harder to do that in writing after proofreading, etc.. not to mention its much more "accountable" in writing.
but who am i kidding hah, its SC after all.
Quote:
Originally Posted by TankSinatra
I was pretty sure they reverted it back, which is an entire roster , defensive method.
this is why roster engineering works the way it does, and every match we have is easily attributable to this type of algorithm.
Any time the opp has a few more TH11 at the top, they have a few more sub TH8 at the bottom. ( we run nothing under th9 )
clearly not lane matching, as if you create lanes out of say 15% of the bases, 2 lanes are way imbalanced. ( in our war, the 2nd grouping, and the last grouping(s) )
Also, if you re-read Humbob's line fitting posts, this method also support outliers at various points in the roster, which is the way engineering is effective.. 9.11 here, 9.5 there, defenseless or 2 down there.. etc. ( you also have to understand linefitting )
the "fix" could be reordering the dots, and changing the shape of the line, which is what most people advocate without understanding what they are advocating in favor of.
What caught me was how Darien said defence holds significant sway in MM at the moment. How does that jive with the statement balanced bases will do best. Perhaps the community is way off on what is balanced.
After further thought on the matter, darian also mentioned that you won't get a clan full of max 9's matching a clan full of 11's. So the upgraded MMA will be able to differentiate between them.
From his statements, since they can't match on a per village level. He said matches would take to long, they are doing the next best thing and matching offense totals to offense totals and defense totals to defense totals. Then matching a clan with similar totals. This still allows for some manipulation, as it's another system of averages. But, as long as they get sensible weights for offense units, they MMA should do okay.
This is just me, but I think they have already tweaked some defense weights. I'm noticing in the few wars are participate in, 9's are much more scrambled. Some have xbows some don't. My SuperStorm account is now setting over several bases with xbows, even though I have not built mine yet. All defenses I have are maxed (except my wiz towers and about 25% of my walls are maxed. But I'm sitting over players that have come from th8 with maxed defenses + all new th9 defenses (including xbows).
So back to match making, once offense and defense are totaled, MMA will attempt to match your clan with a clan that has similar totals and similar proportion of offense to defense. So clans running multiple defenseless accounts will have a large difference in offense to defense. They will most likely have longer wait times and will be matching other clans that have a large disparity.
Again, I think a big factor on how well the new MMA will work will depend on how well they do weighting things. It has potential though.
To me the closer a clan is to balanced, the more unpredictable the outcome will be. If an offense heavy clan matches an offense heavy clan on a regular basis, they could be destined for more draws. However if you run balanced, and match a balanced clan, the war could go either way. Which is how it should be.Quote:
Originally Posted by Clark11308
I have been saying this for about a month now. That based on the statements I have read from SC and witnessing mine and others actual in-game experiences that the community and SC have potentially very different meanings of the words lopsided, balanced, and engineered.Quote:
Originally Posted by Clark11308
My experience is that casual engineering is pervasive throughout the clash community. Only that tiny % are engineering so much that they produce unwinnable wars; the rest get a mild advantage that's often offset by their lack of skill (because they're casual).Quote:
Originally Posted by Holps
Nailing down these numbers so we don't have to rely on our subjective experiences would be great. I think only supercell has the access to do that.
When a clan of casual engineers matches a clan of casual non-engineers, it might seem at first like the war is unwinnable ("look at all their TH11s! They have a TH11 opposite my TH8!" and not noticing that the "TH11" has TH7 defenses and no AQ, and their most powerful troop is dragon level 4) but that casual engineering clan still loses when it comes up against a skilled clan, whether it engineers or not.
There is obviously work to do in the matchmaker; even though TH11 mass drags is a poor strategy, it was still enough in the case of the recent poster whose clan's #1 was a new TH9.
That's what an anecdote is.Quote:
Originally Posted by Holps
We never have wars where we're tempted to not fill CCs. Even when it looks like we are outmatched up top, we remember all the times when clans that outmatched us handed us the victory by failing to clear the middle of the map.
This is why it is so, so important that the matchmaker not match clans with TH11 against clans that don't have TH11; or clans whose top base is TH10 against clans that don't have any TH10; and so on down the line. This one step would basically eliminate engineered matchups for noob clans. They'd still lose, because they're noobs, but they wouldn't be discouraged ahead of time.Quote:
Originally Posted by Holps
We can debate how big of a problem it is when a clan with 10 TH11s and 10 TH10.5s matches a clan with 10 TH11s and 10 TH10s, but this kind of matchup isn't going to affect new players.
So much to unpack here!Quote:
Darian: As far as the actual algorithm and the number crunching and the actual weights themselves, we won't be able to go into - and that's for proprietary reasons. We want to make sure that players try to play, I guess, "in the spirit" as much as possible and I think if we were able to provide those numbers, that would create more gaming of the system - and that's largely what we feel engineering is: trying to find out where the strengths and weakness of the MM system is and utilize them to their advantage. I played a lot of miniatures war gaming [...] where trying to find out what the maximum amount of strengths are, in a meta game that allows you to win at any cost and I think engineering is kind of a similar... I don't want to say... it's not a problem, but it's a symptom of a system that allows something like that to happen.
He must have played some of the same miniatures wargames that I did, where the algorithm was fully public, players exploited it, and the people making the game didn't have the math or the intelligence or the inclination to understand why their algorithms were faulty, and shut down discussion of any improvement. If the algorithm were not fully public, then it would be harder to discover the flaws and the people making the game might not be faced with demands to improve something that they don't know how to improve... I am not sure that things are working out quite that way in this case, because even without knowing the exact details, people have been able to come up with sufficient flaws to raise hullabaloo. The solution is not to hide the flaws, the solution is to fix them. I think we would get to the desirable end point much faster with a public algorithm and weights, which the community can identify specific flaws in, which the company can then fix.
A good analogy is security by obscurity vs security by openness. In the computer security community, algorithms that are secure are ones that the community has tried and failed to break. Keeping your algorithm a secret doesn't mean that its flaws don't get exploited, it just means that the flaws take longer before the public discovers them and forces a fix. Skilled attackers discover them soon enough either way; the difference is how long they get to exploit them before they're patched.
Would be interesting if they just matched wardens up eh. Will be some tears.Quote:
Originally Posted by SuperStorm103
I do not see how they fix it without increasing search times, tbh.
may want to recheck the current 11v10 meta little buddyQuote:
Originally Posted by littledoctor
I think engineered clans typically have longer search times anyway, so if they are inconvenienced by waiting until there is another engineered clan to match, nobody will really notice.Quote:
Originally Posted by ryryshouse
:)Quote:
Originally Posted by ryryshouse
We'll see if casual engineers start leveling up their clone.
Please remember, it is also deficient when there are no engineered bases.Quote:
Originally Posted by littledoctor
We engineer, and we get engineer wars almost every time anyway so business as usual is what I took out of it. Good interview though. Really glad supercell hired a community manager.
Very interesting times ahead, hope this goes live soon and that supercell are ready to make rapid adjustments If it doesnt work well.
I think that your war search time will be a good indicator on your clan structure. That is how clans will know how close they are to what the MM considers balanced or common. The longer your have to wait for a match, will mean that there are fewer clans that match your structure. Most clan leaders should be able to isolate the issue and begin fixing the problem.
If your a clan with 1 th11 and 9 th8's searching for a war and it takes a long time to match, well it's pretty easy to see that your roster is not really typical.
There is still one outlying issue here, and that would be loading a bunch of th3's at the bottom of your roster. I don't know what they will have in place to check for that, but since apparently they are still working with averages, th3's could generate top heavy match advantages for one clan or the other.
Does dumpin th3s on bottom of war map still work? Plz say no lolQuote:
Originally Posted by SuperStorm103
I agree with most of this. I'll add that engineers are easier to beat than modders, so I'd probably rather face them. However, if I am going to lose, I would much rather lose to modders. At least when you lose to modders, you get interesting replays that might give you ideas on better ways to utilize your troops. When you lose to engineers, it's typically because you made a mistake or 2 on your end and you get to watch the other side pull off a series of attacks that blind monkeys could prevail with.Quote:
Originally Posted by Playforwar
Kim! Good to see you my friend. I have seen a few wars recently where the opposing clan had junk bases at the bottom of the map and it did appear effect the war match. They had a much heavier top then us. I cannot really say if it is a consistent advantage or not.Quote:
Originally Posted by kimchilove
Sup storm!!! Yea long time man. I'm afraid of perm ban so tryin to keep my brash self incognito lol. Dang man so its still somewhat effective? Shame. Anytime u need help wit anything bro wit ur experiments hit me up. Fully maxed th11,10,9 n engineer dless th9 (max walls n 28/30)Quote:
Originally Posted by SuperStorm103
Ive been in more than enough clan wars to know that engineered clans wish to pit their th11s against th9s, th10s against th8s, etc.Quote:
Originally Posted by Shanahan
No engineer should have the option to defend against their equal in multi if theyre unwilling to face them in war.
This is all very interesting, especially, how both sides seem to hear the same words and come up with different meanings. Too many multi-quotes are beyond my skill set.
modders vs engy: thanks for the responses, but I would much rather lose to someone who has a better strategy within the allowed rules (eng), than someone who cheats (mods).
Darian did a great job of stating in PR speak, the same thing he said in the announcement. Making changes, evolutions, better matching. Didn't give any data, any stats, or any real actionable items.
th11 meta with drag, loon, clone is hardly mass dragons.
I still think most clans engineer poorly, and if I believe the anti engineers, are terrible attackers. If you add those 2 together, the superior skill of the non engineers should be a fair match and not produce over at match, unwinnable wars.
They usually end up being close wars. It's just not fun losing to people who aren't as good at attacking as you are. I'd rather lose to people better at the game than me. It helps me get better myself.Quote:
Originally Posted by Ripple1972
It feels like it. Every war we'd have an enemy with 1 or 2 defenceless or 1 or 2 junk, and then stronger up top. So i tried 1 junk base and a th7 to see if it would balance things up for a couple of wars... and both times the enemy had 5 or 6 low bases and a massive advantage up top. So we've gone back to th8 minimum :rolleyes:Quote:
Originally Posted by kimchilove
But this is not not necessarily in-line with what Jonas had said way back with an interview with Ash. He stated that balanced bases within a balanced roster were the most likely to get favorable matches. I guess in this statement there are multiple words that could have a variety of different definitions. What is a balanced base? Balanced roster? Favorable match?Quote:
Originally Posted by SuperStorm103
You say a balanced matches a balanced and the war can go either way but you're still making a plethora of assumptions in that statement. Just because two clans that are both balanced (whatever that means?) match each other does not inherently make the war fair. MM has proven time and time again to match maxish ths with the next th up, rushed bases with both their same th level and lower th levels, and I am sure other things. Personally at this point I am starting to believe that the notion of a balanced base is a fallacy at this point. Can it actually exist if SC does not define it for us.
Smh. Think that form of warring is worst of all manipulators. Darian didnt address that issue but I hope it gets rectified. Sorry to hear u exp that issue.Quote:
Originally Posted by OnyxDS
Do heroes get counted for offensive or defensive weight or is there a seperate type for them?
I'd assume they count for both but who knows for sure.Quote:
Originally Posted by SheerluckHolmes
Agree, seems like they ignore (and will continue to ignore) TH, Labs, etc which are very important in my mind to a balanced match.Quote:
Originally Posted by SuperStorm103
This seems like a perfect conclusion of what engineers think and why they do it, you only want an advantage. Make your advantage by skill not numbers.Quote:
Originally Posted by Shanahan
Will it be a good indicator? War search times are still subjective to a similar clan in the pool at the same time you hit search. Just because your clan is balanced doesn't mean there is automatically a good fit in the available pool. Everyone is trying to say when/if they separate engineered and balanced (whatever that is?) that it will automatically make it take longer for a highly engineered clan to match. Will it? Or will it take longer for everyone to match because you just shrunk the available pool or split in halfish. Hasn't it been proven on here before that decreasing the available pool is the most detrimental thing SC could do to the MM? And what if you're clan is only moderately engineered, with a good common clan structure? Remember engineered and balanced don't appear in my eyes to be mutually exclusive.Quote:
Originally Posted by SuperStorm103
IMO, the best thing you could determine is if you have a really odd clan structure such as the example in your second paragraph then it could take time to match, but there are loads of engineered or balanced clans (whatever that is?) that still fall into a more common clan structure that could match each other quickly. But we already know that the example in your second paragraph is not a common clan structure, we don't need the MM to tell us this.
As far as I'm concerned, the interview with Darian should be the new benchmark. And yes, a lot of what I'm saying is currently theoretical. I wasn't impressed with Jonas' statements about the game. I even called it a perfect example of "fake news". That does not pertain to the personal information that was stated by him about how important the game is to them. But things like "90% of the players played the new BH and so it was a big success" statements are what im referring too.Quote:
Originally Posted by WookieFloyd
I think it will be a good indicator theoretically, as the overlying statement is that your clan will match similar clans. If you assume that most players follow a natural path of game progression and not one aimed at gaming the MMA, they will by default have a very common composition. There may be some anomalies here and there with their break downs. Engineered clans could see longer and longer wait times depending on how far they deviated from the natural path of progression to influence the MMA.Quote:
Originally Posted by WookieFloyd
I used an extreme example to emphasize the point. I know you and everybody else are able to discern uncommon rosters. And the clans with such a breakdown may also be 1 player clans again trying to manipulate the MMA.
Quote:
Originally Posted by OnyxDS
Wow! very late to the discussion, and it looks like a fun one.Quote:
Originally Posted by kimchilove
Welcome back kimchi! FWIW I agree; roster-dumping th3 is even worse than what us filthy engineers do, IMO.
Excuse me, I meant filthy optimizers.
Same as Ripple I'm not good at multi-quoting, and personally I prefer not to get into contentious point-by-point rebuttals anyway. [I suspect when Lunatic gets on, this thread will cause my computer to explode.] So a precis of my thoughts:
Holps, I appreciate you clarifying your early post. Like Chang405 I thought it was a bit, er, immoderate, if you'll allow me bruv, and am gratified to know it was meant as flippancy and humorous exaggeration. I even agree that unchecked engineering is bad for the game. But tbh I am not sure that there will be any movement on it until an MM fix happens, and happens clearly enough for us skeptics to weigh the evidence.
I want a better MM too; and I am willing (hoping!) to interpret this talk of "favorable" matches as equating to "even" matches, which I define as the ones in which the outcome is in doubt, depends primarily on skill, and you lose if you don't perform. Because I find "even" matches much more fun.
But I also believe that, despite all the forum discourse, there still is not much consensus on how to achieve "even"--or for that matter what "even" means.
Recently I posited that Rogue Elite was getting grotesque matchups on paper precisely because they were really good at attacking, and won a lot, and that the current MM penalizes them for both and gives them successively harder matches as a consequence. I didn't mean it as an attack, or as a suggestion that they should change their ways. I'd love to be as good as Rogue Elite and have their record!
Where it got contentious is that I suggested this might be "fair" in the sense of both sides having a chance to win, and their leaders instead posited--perfectly reasonably I might add!--that "fair" would be two competitors using the exact same techniques and tools.
Even though I know that argument by analogy is suspect, let me try this one: horse races weigh jockeys, and weigh down competing horses with lighter jockies accordingly. Is it more fair to weigh horses down, or more fair to argue that any team which figured out a way to make its jockey as light as possible should be accorded an advantage thereof?
I think it's fair to say that the latter situation is what the current MM actually DOES.
But, second question, if you were racing competitively, and you knew the latter situation was what actually pertained, would it be reasonable to "exploit" it, or to continue putting heavy jockies on your horses?
Remember, it is about demographics. so if the large populace has these bases, you are increasing your exposure to these rosters, not limiting yourself.Quote:
Originally Posted by OnyxDS
Lol ty bro. Ive been back but tryin not to post too much bc im troll food n I'll prob b perm banned next time. Lol filthy optimizers. God bless net people eh?Quote:
Originally Posted by posthoc
But your assuming that these so called balanced bases or natural progression (whatever those are?) are the most common and therefore will have more available matches in the pool. What if this is not the case? And a very common base progression doesn't equate to a very common roster structure, which in the end is probably more important than balanced bases. Hasn't it been said that the rusher is the most common player in clash? Well if that's true and the MM has always struggled to differentiate between rusher and engineer then couldn't it be argued that balanced bases are not the most common bases and therefore will take longer to match? Especially if you assume balanced bases will then not match engineered or rusher and therefore reduces the available pool further.Quote:
Originally Posted by SuperStorm103
It is of course yet to be seen bow things will work out. Going only on the fact that it appears offense to offense and defense to defense totals will be matched, OMMMA (Officially Modified Match Making Algorithm) will be able to detect the difference between a rusher clan and an engineering clan. Engineering and Rushing are opposites. Currently, engineering is good and rushing is bad. Rusher clans have defense score higher then offense and engineering clans having higher offense score then defense. So OMMMA, as it appears, will attempt to match two clans with defense over offense scores to other clans with defense over offense scores and vice versa for engineered clans.Quote:
Originally Posted by WookieFloyd
In the same way OMMMA (as suggested) will be able to tell the difference between both rushed and engineered clans as they will would theoretically have matching offense and defense scores. The accuracy may not be super, but I believe player skill can offset any minor advantages caused by 1 or to mid to low th10's matching max th10's.
What this means for match times, is that rusher clans will most likely have the fastest match times with maxxer and engineering clans being the minority will have longer search times. Engineered clans, depending on how intentionally distorted they are, will probably have incredibly long wait times for a match.
Again the effectiveness will depend on how close to reality the revised weighting systems is. As for clans that have 1 or two maxed offense units and selective defense units trying to appear as balanced, Darian stated that they tried to make it so th11's will not match th9's. That is about all I could gather from his comments. They have somehow figured out how to differentiate between a th11 posing as an 8 or 9 from an actual 8 or 9.
Also, Darian said if search times are an issue they have made it so they can easily modify OMMMA to adress it.
Haven't you claimed before that maxed bases are balanced bases, though? There is zero evidence for this being true, other than some purely theoretical negating notion that a base maxed on offense for a given level should be maxed on defense for a given level before jumping, or vice versa. Well...Quote:
Originally Posted by WookieFloyd
...
Look at those poor th5! they don't even get heal spells until they jump to th6, regardless whether they max their defenses or not! How unfair. the MM is screwing those defensive-max th5.
...
Look at those poor th6! they don't even get a fourth army camp, until they jump to th7, regardless whether they max their defenses or not! How unfair, the MM is screwing those defensive-max th6.
...
Oh, I'm sorry, you were talking about higher-level clans? OK...
...
Look at those poor th9! they don't even get a fifth spell until they jump to th10, regardless how big their xbows were. How unfair, the MMM is screwing those defensive-max th9.
...
Look at these poor th10! they don't even get a warden until they jump to th11, regardless how big their infernoes were.
My clan sometimes still recruits on global (the shame!) and we know. The rushers and average players out there, couldn't give a rap about what any of us think is "natural" or "balanced" progression. They want the big stuff, and they want it now. They might've heard some ♥♥♥♥amamie stuff (from me) about how to hold their weight down (from me) but they couldn't care less. Clash is a fun sideshow to them. Which shows maturity and sound life choices, for most of us.
It's anecdotal and therefore unprovable... but it's not a lie. Those of you with junior members will know that.
I have.Quote:
Originally Posted by posthoc
If maxed TH11 isnt balanced, the game is broken.
Before TH11, maxed TH10 should have been balanced, else the game is broken.
We have seen times when the game was broken. Remember witches? everybody dumped witches for 3, they considered that unbalanced so they fixed it.
So, it is not anecdotal, and it is demonstrable in history of the game.
I would agree that highly distorted engineered clans could have long wait times (how many engineered clans truly fall into this?), but I'd also argue that Maxer clans will as well. It appears to me that true maxers are the minority in this game as very few players truly reach max at each th level and then how long do they stay at max at that th level? I would also say that engineered clans tend to not necessarily be so out of balance as you seem to think. In my experience most engineered bases are only slightly off the path of what you call natural progression (whatever that is?). Also how much stock is taken into balanced bases vs a balanced/common roster? Which one is more important in MMing? If you are only very mildly engineered with a common roster wouldn't you likely still see an advantage a lot of times then?Quote:
Originally Posted by SuperStorm103
Do you honestly think that the MM is going to be able to differentiate between all of the infinite amount of ways you can upgrade your base? IMO, the best the MM is going to be able to do is weed out the extreme situations, which would be a huge step in my opinion. Besides the extremes there is far too much overlap between engineered and balanced, it just seems unrealistic to me to be able to account for it all. But I am no programmer or computer guru either so I am by no means an expert.
It seems to me that the .5 progression path was deemed a larger issue by SC than the defenseless base since they seem to be affected more by the last year of MM changes. Most posts on the forum have talked about .5s being far less effective then before and that defenseless bases still seem to be skirting by the MM. Now if the MM changes its role to try to weed out more of these defenseless bases will it lose its handle on the .5 bases? Can the MM truly handle both the extreme situations in either direction (offensive or defensive) and everything else in between?
There's some reason to think that this is true. Is a max TH11 balanced or not? Was max TH10 balanced before TH11? What about max TH9 and max TH8 when they were king of the hill?Quote:
Originally Posted by posthoc
I'm not saying I'm sure it's true, but I'd assume that "balanced" in gameplay terms includes maxed bases. That said, the median base is quite likely tilted one way or another (probably towards offense) and that might be a better measurement for "balanced" depending on the point being discussed.
I would swear that I heard them talking about finding a defensive match appropriate for your offense. (suggesting offense matching to defense, not offense to offense). I'll dig for the quote point later (sorry).Quote:
Originally Posted by SuperStorm103
I personally think we should toss the word balanced out of the window. It's like the end of a rainbow. You can chase it's meaning all you want, but will you will never get there.
We should just stick to rushed, engineered, and maxed.
I have not claimed anything about balanced bases because nobody knows what a balanced base is because SC is the one who defines balanced and they have not told us what that is as far as I know. It could be possible that a truly maxed base for a th level is balanced, I don't know. Here is a thread I made on balanced vs engineered vs lopsided https://forum.supercell.com/showthre...and-engineeredQuote:
Originally Posted by posthoc
I am not really sure if your post was directed at my comments because I was discussing timeliness of matches with SS and whether engineers, rushers, or balanced (whatever these are?) bases could see longer wait times under a hypothetical new MM.
So basically, they're changing the weights. Other than that, not much. So a th11 with maxed troops and th3 defence, may start pulling high th10s instead of pulling th8s in war. Or if they overdo it, it may start matching new th11s. But it is also possible that it will start matching th9s.
This sounds bad. Not sure whether it will be bad for engineers or the balanced bases, but depending on the weights they assign, it sounds like it will be bad for at least one of the groups.
I'm not suggesting matching base by base, I'm just saying that matching two weights, one for offence and one for defence would have been better.
I don't see how engineers will match engineers if they're using one number to represent the entire clan.
Correct me if I'm thinking this the wrong way.
I agree, and might even suggest that true maxers are an almost undetectable--if extremely dedicated and skilled!--minority.Quote:
Originally Posted by WookieFloyd
My previous theory, which I see no need to revise, is that current MM is aggressively non-linear, for any number of reasons.
They might have better data than us outsiders.
They might have started with one, which worked OK until the game grew and evolved, and now are scrambling to catch up.
They might have started with a linear system, but had to add so many tweaks and corrections for boundary cases that it might as well be nonlinear.
In all of my (admittedly unsupported) theories, concern for maxer clans is conspicuously absent. They are, most likely, simply too small a constituency to devote any effort to mollifying.
Now this is something I can get on board with. Balanced bases is a fallacy IMO, at least until the day SC defines what their idea of balanced is. This will likely not happen as it could open the door to further or new version of gaming the MM. But unfortunately we are on a forum where anyone who drops all the defenses for their th level believes that they are a balanced base and getting the raw end of a war matchup because the MM found an engineered clan for them.Quote:
Originally Posted by SuperStorm103
I assume you are an above average player. You will never have much fun using the random spin because most people will not be as good at attacking as you.Quote:
Originally Posted by Taternader
.5 used to be an issue as they didn't count offense. Then they gave offense weight and the MMA was adjusted to account for it. The MMA was unable to accurately account for anything outside the realm of .5 imo. It got much worse with the introduction of new troops/higher level troops/new heroes. The more extreme your offense was, the less accurately the MMA could match you.Quote:
Originally Posted by WookieFloyd
With the new system, I do believe they can actually tell the difference between rushed/engineered/maxed villages. There are markers now if they match on offense and defense instead of just a cumulative average of offense and defense. I agree that maxers and engineers are in the minority. I cannot however really comment on the accuracy of OMMMA.
I still think that Darians reference to "twinks" was a hint. "twinks" also had a marker that allowed them to be identified. That was the option to turn on or off experience gains. Along those lines, clans with offense values over defense values will be marked as engineered, clans with defense values over offense values will be marked rushed. Clan's with = offense defense values will me marked as maxed. I cannot really talk about how accurately it will match within those groups. Will just have to wait and see.
The only outstanding questions I have at this point are:
When will it go live and will it be announced?
How will it handle junk bases at the bottom of the roster?
Maxers lives matter! Seriously though, I'm proud to have 2 maxed bases (just 9, but still proud).And I'm most certain that I am not alone. I also think it is pretty disingenuous for you to assert that our bases deserve no effort in this process. Shame on you.Quote:
Originally Posted by posthoc
What new system? Surely you wouldn't suggest SC might roll out a major update at 3am without notification... :saywhat:Quote:
Originally Posted by SuperStorm103
But in all seriousness I doubt very much any such "twink" identification system is in the works... Or, that if it does, it would an improvement, or even a modification, of what currently happens.
If SC had a reliable way to tell the difference between engineered, rushed, and maxed clans, it would've been implemented long ago. If anything, from the get-go. Weak anthropic principle here.
So the question then is can good engineers such as #6M continue to blur these lines. If the are not matching lane by lane and still are taking an aggregate clan weight it will always be possible to blur these lines imo.Quote:
Originally Posted by SuperStorm103
Also, what is offense = defense look like, are we back to using balanced again? How much can you stray off this equilibrium before the pendulum totally swings in one direction or the other? Is it possible to even maintain this equilibrium?
I think you are slightly off (if i'm thinking correctly). They are going to be matching total offense numbers with each other and total defenses numbers with each others and then matching a clan with similar numbers. So having a max th11 with lvl 3 defenses will skew a clans weight to having more offense then defense qualifying that clan as engineered. That clan should then match other clans that are also labeled engineered. They must be doing something significant with offense weights, because Darian said that a th11 can no longer match a th9.Quote:
Originally Posted by SystematicChaos
So we have a general idea of how it will work, but not much in the way of details. Hopefully, if you don't want to match any defenseless maxed th11's, don't bring any maxed defenseless th11's to war.
Not at all, maxed is just that, maxed. Signified by having very close offense to defense levels. Darian mentioned that th11's wont match th9's and that offense will be included much more in the matching. To me that means that offense will have weight adjustments. One would assume that now having a max lava hound and maxed loons will have significant weight (where he didn't before as he said).Quote:
Originally Posted by WookieFloyd
I (like everybody else) don't really have much information on the Martians. My guess is they will appear to be extremely offense over defense. If that is the case and OMMMA works as intended, they could be in for some very long weight times for war, if they can even find a match. From what we know about ban♥♥♥♥B's roster. They may very well be hosed.
We will just have to weight and see though.
SC will always cater to the masses. The masses provide profit in abundance over the minority. And a true max base is just not apart of the masses. SC doesn't care if you are proud of your base or not they only care how many gems you have purchased, and if your purchased gems make a difference on their profit margins. Plus it seems pretty obvious to me that SC doesn't seem to like true max bases as they stop base progression which means no purchased gems. They show this time and time again through their MM by matching max base clans with the next th level up.Quote:
Originally Posted by Playforwar
FYI, the part about matching offense to offense (as I took it) is at about the 10:50 mark. For those interested in that bit.
If what you're saying is what he really meant, then I'm sure we'll all like that.Quote:
Originally Posted by SuperStorm103
He said it will take total offence weight and total defence weight into account whereas earlier it was very defence heavy. It's the part after whereas that got me confused. It's as if he's saying earlier offence didn't have much weight, but now it's fixed and offence has more weight. At 14:30, he also said that it's not that they're strictly pairing engineers with engineers.
Anyway, from the way he presented it, it's pretty clear that he means mm manipulation is against the spirit of the game. The takeaway for me is that supercell( or may be just Darian ) is against engineered bases, but they have to be diplomatic about this. So, that's good.
I don't think #6M is going to get affected by this. This update is not a Nerf to all types of mm manipulation. I think this will just make offence matter more than it currently does.Quote:
Originally Posted by WookieFloyd
THANK you. I was about to devise a response to the original--well, accusation, I call it advisedly--and you expressed my feelings succinctly--even if you manifestly aren't in accord with my beliefs about engineering.Quote:
Originally Posted by WookieFloyd
Every base and every player deserves SuperCell's best efforts to keep them engaged and satisfied.
TRUE max base? I didn't realize there was something other than max. All I know is that there is nothing left to do. Is that TRUE?Quote:
Originally Posted by WookieFloyd
Yes, I would assume that strictly engineered vs engineered is not totally accurate as there are many degrees of engineering. Just think of it as degrees of engineering will match degrees of engineering.Quote:
Originally Posted by SystematicChaos
I also think that by including offense equally in the matching process, they will be adding weight to offense. So all those clan's that are currently enjoying their offense advantage and free wardens, possibly could be in for a shock when the new weights/consideration is applied.
Thank you.Quote:
Originally Posted by posthoc
If true max bases are a minority and not part of the masses generating revenue, do they really care if you want to stay at th9 forever? It seems they want all players to be happy and play how they want to play. Forcing people to upgrade does not jive with that.Quote:
Originally Posted by posthoc
A lot of these players staying maxed like to play in war leagues in which to be anything other then maxed is bad. Let them stay maxed. And give them other maxed bases to war against. CWL may actually be able to use the in game match maker for their league.