Originally Posted by
CasimirEffect
I mentioned the dangers of providing blanket statements, or a list of bullet points, covering "correct or ideal roster compositions," as is the purpose of this thread, a while back. People are ingenious and given enough time and a template, those interested will find a way to exploit any system and common roster design philosophy.
Mossack, you might need to constantly update your thread based on our experiences over the past months to avoid walking people into a trap...
So after much discussion, our clan took much of the advice provided in this thread and rid itself of all lopsided bases (ALL!). For some time we did see a marked improvement in matchups, but the engineering learning curve is catching up to October updates and people are now designing rosters which apparently match with clans like ours (5 TH11, 5 TH10, 15 TH9s, +/-1-2 in any category given upgrades and such) in less than 25 minutes. For the last 2-3 weeks we have seen engineered rosters at at least twice per week while restricting search time, babysitting our roster, keeping reasonable weight distribution (TH9 w/ infernos - max TH11). While numbers are always slightly different in each war there is one consistency: our opponent is able to trade lower weight at the bottom of their roster for addition Warden, Eagles, infernos, and often 9.5s up top. The additional full TH11s are the ONLY asset which tips the scales and frustrates our group (myself included), all other differences can be managed and I accept as variance. We have seen additional TH11(s) on the enemy side in all 8 of our last wars despite varying our TH11 count from 2-5 in an attempt to fix.
My point remains from last month: Until SC has a MM algorithm that accurately weighs the impact 2 or more additional TH11 defenses and 2x the attacks has on the outcome of any given war (skill being more or less equal) any post like this will typically lend itself to exploitation if the community embraces it. I appreciate your efforts, but SC has some work to do if posts like this are to bare fruit rather than frustration.
Issues to be addressed before I recommend anyone follows this upgrade path:
1. How MM handles balancing of key TH11 structures like Warden and Eagle
1A. How MM handles weight trade at top for weight at the bottom (likely fixing #1 will fix this as well)