Demon Shell

Two weeks late, but I have to chime in and say both congratulations and thanks for all the effort and time put into this project.

So I finally got a chance to test Demons. Ember Demon/Demonlings are not fighting at what I would consider to be ranged. They would engage only a few steps away from an enemy. I believe it had to do with this. So, they're mid-ranged. They still stay relatively close by. It was different, but it felt alright. They tend to be further apart, so they get hit with cones/AoEs less. And that's probably the point. Kinetics are probably going to need to force them into melee for FS.

Most of this sounds great. And the implications these types of changes have for future enemy encounters/AI is a major plus. However, I don't know about Demons suddenly becoming predominantly ranged. This is mostly with regards to Ember Demon, who I generally assume will be near me in mid-range. Something I'll need to test when I have a moment.

Coralax set Freakshow set Knives of Artemis set Maybe Outcast Could probably get away with Midnighters Basically some more CoH canon-linked options.

If you're sure, then I'd go full AoE click to bridge as much distance as possible between Dark Melee and TW. Looking at the numbers, that seems to be /Rad. If TW and a few other sets have trouble sustaining at the SO level with 2-3 sustain-less armor sets under the condition that they don't use any Power Pools (and with SR and Shield heavily benefiting from Aid Self/Field Medic synergy) then who cares? There are tons of underperforming combinations in this game at the SO level. And if TW drops out of the top spot on /Dark (which may not even happen given that, at the SO level, /Dark can turn Murky, CoD, or both off for the vast majority of encounters with little difference in survival), then I doubt that drop will be very far. My only concern in sustain was the impact of click sustain versus passive +rec. If the answer is "practicality nothing" then it's not worth a dedicated trial to investigate sustain issues.

I would have liked to avoid /Ice because, unlike Power Sink and other similar powers, Energy Absorption is a proactive ability due to the +def. The entire point of picking a set with a click sustain power is that some sets (like TW) will need it and have their DPS reduced in using it. Other sets that don't need it, won't use it. You can compare those results to the Willpower test and measure the difference between a constant +rec and a sustain click. Does, for example, Claw or Staff begin to close the gap with TW (assuming they don't need extra sustain)? But it is the nature of Energy Absorption to use it when its available for survival. Apart from impacting TW the most, every set is going to use it and have similar reductions to DPS with a similar boost to sustain. Based on your criteria and the sets you listed: Bio, Rad, or Fiery. I won't pretend to know anything more about Bio or Rad than what Mids shows me (I have no experience with either). They have additional AoE options and passive Recovery like Willpower. Fiery is going to be weird. And ultimately sets that need extra sustain are going to use Consume, and sets that don't need the sustain are typically going to take any AoE available (provided enough targets are present). So, like Energy Absorption, every set is probably going to use Consume anyway. That defeats the purpose, in that area (though maybe the impact of an AoE Armor is a higher priority). There isn't a perfect choice. A concession is going to have to be made somewhere.

I never said it wasn't. I said: You are not going to take the clear times of Willpower and Dark Armor, average them, and get anything meaningful out of it. A set with nothing to manage and a set with everything to manage is just going to give you two, unequal extremes. You need an average performing set. A power or two to click, that's it. No concern given for when to run what toggles, like DA, because no other set traditionally does that. And to that end, I don't think it should be SR. The first set I suggested was /Electric, which has just as many end-return/heal tools as /Energy but is a resist set and should provide more stable survival across all testing. But if I could go back in time and tell you two trials were going to be run, I'd have said to do /Electric and /Energy. They're the two most similar sets. And from those you could then draw very strong conclusions, such as do some sets perform better with a damage aura or a buff/debuff aura? At the SO level, is it better to stack resist/def on every set that can, or layer? If things like these don't matter (and, for example, every set just clears 10% faster on Energy), something is wrong with the attack set balancing and you have strong numbers for where and why this is happening. If you want to test them under endurance stress, don't use Power Sink/Energy Drain. Is Energize with a 50/50 uptime enough for even TW? If so, grab some greens and don't use either move.

Invulnerability's +def scale from Invincibility is so extreme the set's survival goes up the more enemies are around you. Perfectly normal from an end use stand point, but it's survivability is an oddity. That's going to mitigate the value of the survival tools from your melee sets. Stone (in addition to not being available on Scrappers) has both fewer toggles to run due to Granite, and a reduced recharge resulting in your more expensive attacks being used less. That is (forced) reduced end usage. And, if you're not running Granite, you're end usage is bizarrely high on SOs. Why does Mud Pots cost so much? Why do Rock/Brimstone/Crystal/Mineral have above average end costs each when you also need to run Rooted? Outside of IOs, the set has a lot more wrong with it than just the extreme -dam/-rech on Granite. Shield can't run all weapon sets so it's not suitable for testing. And while /Dark is a high performance, end-game set (I know, I have one), it's not balanced at the SO level (unless you're a Tanker). Dark Regeneration costs more than a nuke to overheal you by the hundreds. CoF's accuracy is too low. OG would heavily benefit pure DPS sets because the enemies in the test only have one melee attack, mitigating the value of most CC. And, at the same rate, it's also going to lower clear times on sets without KD/heavy AoE as the mobs just stagger out of range (and heavily benefit sets with Stuns as you can stack them on a boss to reduce incoming damage and avoid using Dark Regen). If you're looking at sets with no end management tools, that don't have some strange outlier mechanics going on, that leaves SR. SR has fewer toggles, but Quickness will increase frequency of end-heavy attacks. I'd rather a set with at least some level of end management and a damage aura be used, but SR doesn't have an aura at all on Scrappers. That should at least create a different atmosphere than Willpower.

Fire Melee can't sustain on /dark. I know, I had one (a Brute) pre-IOs. So of course Titan Weapon can't. All that proves is something is wrong with /dark at the SO level (it's Dark Regeneration, it costs 20 endurance with 2 end redux). That's not news. If your concern is wholly based on TW, then you've missed something important. None of these tests are going to prove where Titan stands until you incorporate Epic Pools. Every other set can benefit greatly from Epics, except Titan. Titan might gain from throwing a Ball Lightning at the start of a fight while jumping in, but anywhere else is going to interrupt Momentum. If there's more than one enemy alive, Dark Melee is going to be using Ball Lightning on cooldown. That is a radical shift for a set with no 360 AoE spammable. My point is to apply pressure to see where the other sets stand relative to one another. Is Claw's resource management so good it never has to stop attacking to get endurance back? How good is Parry when it's your only source of defense/tohit reduction? Does Ice Melee ever have to heal? Do these things influence clear times to any noticeable degree? You don't (or at least shouldn't) need to remove endurance/heal sustain as an option to test this. You just need to be penalized from a DPS standpoint for using them. And if none of this is true and none of these things matter, then something is wrong (because that would mean DPS really is the only thing that matters, and all the extra tools melee sets have aren't good enough). Either way, this is something you can directly compare with the previous Willpower results. There is nothing you can compare between Willpower and DA; they couldn't be more different. But if you want to apply as much endurance pressure as is reasonable, then /SR. It has no endurance recovery and, like Electric, Quickness will increase your recharge, causing you to use your more expensive powers more often. The testing parameters don't necessitate it, but you'd just have to keep Practiced Brawler up and pretend they do. Being constantly interrupted with PB would also hurt Titan. Additionally, in theory, /SR is balanced to not need a heal (I know mind had one, but I'm not going to pretend I was a very good /SR). This would be a good opportunity to test that. And if some additional mitigation is necessary, how do the sets that have that compare to those that don't.

Well, yeah. Double the DPA on Disembowel and Swoop and that'd pretty much be the end of this discussion. I think it's better to compare BS to Katana though, since they're the Parry sets. And the main issue there is Lotus/Soaring are both superior to Whirling/Disembowel by a wide margin. The answer may be as simple as buff Disembowel's DPA up to Clobber levels so BS is for ST and Katana for AoE (which is how I feel WM and BA should specialize).

I had thought about it a bit and feel that /Elec or Ninjitsu would be about as close to neutral as armors can get. Electric has a bit more end management tools than average, but its toggles cost more. And without the ability to perma Energize at the SO level, I think some sets would still need to use Power Sink. Ninjitsu has, in addition to a heal and end return, a survival power. It's not the best, but maybe some sets like Fire would use it while sets like Ice would ignore it. If Fiery didn't have Burn, it would probably be the best option. As is, I would think the damage aura of Electric would be more valuable to test and compare against than Ninjitsu.

I would specifically avoid Dark Armor. Not just because going in the complete opposite direction isn't going to create an average by comparing Will to Dark. But Dark Armor is brutal without IOs. It is unfun. You're not going to want to play half these characters.

Well, there is the other option. Whirling Axe and Mace both have pretty low DPAs (they're not Stone Melee Tremor bad, but very much on the low end). What if Whirling Axe DPA was brought up to compensate? Let's say Cleave has an arcana-time DPA of 70 per target. It doesn't (I think it's slightly higher), but let's just say it does. You can expect to hit between 1-2 enemies per use. I was a career BS Scrapper so I'm very familiar with the cone as it is currently. You can, with reasonable reliability hit 2 targets with it by a quick quarter-second nudge to the left or right (and from a DPS perspective, that's absolutely worth it), if they're next to each other. And while you may on rare occasion hit 3 (or 4 if you're fighting Rikti Monkeys), there's also the high chance that two enemies just aren't close enough together, or that there is only 1 enemy. Let's average this all and say you hit 1.5 enemies per swing (you don't, it's lower, but lets pretend). With these assumptions in place we can say your average DPS per swing with Cleave is 105. What we're discussing is increasing the cone size, such that instead of hitting between 1-2, you'd be hitting between 1 and (realistically) like 5 or 6 enemies (it has a target cap of 10, you're not hitting 10). Again, BA is weird. Almost all of its attacks have the same DPA and Cleave is no different. Ordinarily Cleave is either not in, or only rarely used, in your ST rotation (which would mean its rarely ever going to only hit 1). But, if there's a specific power you want and you need to trim a move and you have the endurance, you can drop Chop or Gash and replace it with Cleave and have very little difference in ST DPA (alternately if you're hit with -rech, Cleave is your backup). Because of this, if Cleave were a Shatter-like, you're probably looking at an average of either 3 or 4 targets per swing. If 3, that's 70 * 3 for 210 DPA per use. If 4 that's 280. Between HC and this proposal, that's a difference of 105-175. Let's say Whirling Axe has an arcana-time DPA of 20 per target. It doesn't (I think it's slightly higher), but let's say it does. It has a target cap of 10 and you can totally hit that. Usually though, you're looking at a high of only like 8-9. It is absolutely not an ST power and especially for BA, to only hit 3 enemies would technically be a DPA loss. So your average is going to be high; probably like 5 or 6 targets. So that's 20 * 5 for 100-120 DPA per use. Now I played fast and loose with a lot of these numbers, but the point is to cause Whirling Axe (Average DPA 100-120) to near the proposed change to Cleave (Average DPA 210-280) would pretty much require the DPA of Whirling Axe be doubled. This would take it from a low DPA 360 AoE, to something closer (or even slightly higher than) Claw Spin. That would further separate BA from WM with similar results to a Cleave cone increase (although that wouldn't exactly solve the issue of "Cleave has a target cap of 10 and why").

Well there's your problem. There are 13 armors, 4 of them have recovery passives, and the others either have +end clicks, which lower DPS, or nothing, both of which create a greater emphasis on endurance management. And in choosing the armor set with all passives and no click powers, you didn't just pick the least invasive set, you choose the deviant. Every other set has click powers, even SR. It's part of how armors work. It lowers DPS, it messes up Momentum, and it makes sets that have survival tools that also deal damage (or trump, in effectiveness, the support powers in your armor pool) more attractive to use instead. You choose Willpower, but then you didn't leverage it on every set. TW leverages Willpower naturally. It offers increased Recovery (TW is the most expensive set) and does so with no clickables to interrupt Momentum (Electric, Fire, Ice, Energy, and Ninjitsu have end return, but they’re all click powers, if you have to activate those during Momentum (especially Power Sink) your DPS dives). Why then wouldn't you slot the other attack pools to also benefit from Willpower, via the stable, non-DPS interrupting, increase to Recovery? They have the endurance to spare over TW. That's their advantage to leverage in slotting. And this is the key, some sets will leverage it more so than others. And if there are sets that still have a surplus of endurance after that (Claws or Dark maybe?), why wouldn't you throw on Assault to boost it even higher? If you were using a defensive set like /Dark where that end redux is necessary on attacks, wouldn't Dark Melee need less? Wouldn't TW need more? If you were using /Electric, clicking those heal/end powers would reduce DPS on TW more than any other set by interrupting Momentum, but how much per set would depend on how much damage you take/endurance you use. You have to play the attack set to the strengths of the armor pool and I feel you may have focused too much on standardized enhancement slotting instead of getting the fastest clear with just SOs. That's how to determine a set's performance at the SO level versus the performance of a singular style of slotting. Case in point: Dark Melee has no value as an attack set if you aren't running extra toggles to leverage Dark Consumption, which itself has no value in slotting an End Redux over additional recharge (DC has 180 recharge but only costs 0.35 end). Would that have pulled it out of the bottom slot? On Willpower, probably not. But that alone would have increased its performance with 0 downsides. By no means does that invalidate the effort and data you've accumulated, which is vast and eye opening (I thought Kinetic was better than that, but then again my Kinetic was /Fire). Cleave should definitely be a Shatter-like cone. I don't know if that would be enough, but it's certainly a change that should happen.

Due to the risk of Cascading Defense Debuff, unless you're /SR Parry-likes should never be skippable. Your testing with Battle Axe probably showed you what the devs saw when they balanced it that way (again, if I'm remembering right, it's lacking in detail so I'm only assuming it was BA due to the numbers it has, but I distinctly remember reading a devpost on the old forums during a balance pass about balancing an underused melee set based on DPA as a test, but then they never came back to it). At low recharge levels, when you're using all/most of your powers (and no Epic Pools), BA is fine. On the other hand, WM Clobber has a DPA nearly double that of anything BA has. When you're using all of your powers, that's balanced against WM having lower values for its other attacks. But CoH is cooldown reduction, the game. Once you get your cooldowns low enough, you don't need those weaker powers lowering your average. And while we're on that subject. I read during your testing (I apologize if this is wrong, I haven't read through the entire topic) that in addition to always using the same enhancement set up (1acc/redux/rech, 3dam), TW was the only set that had any endurance issues, and even that never slowed your attacking. This leaves me immediately wondering, did you replace the Endurance Reduction with a second Recharge Reduction on some of, or all of, the powers in the other sets to boost the recharge on their higher DPA attacks (until it reached similar end consumption levels as TW)? I understand the desire to maintain a control for testing, but it's not realistic. If you don't need that extra end redux (and it's not being given by IOs while chasing a set bonus) there's no reason to slot for it. If you did use alternate slotting on some attacks, I'd be interesting in seeing those, especially the ones on Dark Melee (though obviously not in this topic, and again I apologize if this was already covered).