Arcanaville

Its unclear based on your appraisal above how Controllers are unique. The only thing really unique to Controllers based on the above appears to be they uniquely have access to the Illusion Control set, and that's problematic for two reasons, one is that all non-Illusion controllers don't get that benefit, and two you're in favor of proliferating it anyway. That would leave it without a unique aspect as you appear to be defining uniqueness above (the other unique element is containment, which is a form of crit).

Architect. I was planning on spending at least some of the next few weekends working on a new performance testing arc. Dialing in the balance can take a lot of iterative testing, and sometimes I want to do things the architect doesn't really present a straight forward way to do. I also have the basics of missions 1 through 4 sketched out, but I'm still thinking about mission 5 (technically, its likely to be mission 4).

I'm also spending a bit more time fleshing out the thing to have a coherent story and all the important blanks filled out, which challenge arcs don't really need but I could use the practice.

This is stealing from my AE time (and its been a busy weekend so far) but I squeezed in one very ugly pylon run. I'm apparently out of practice on pylons and also really need to get with slotting those KB protection IOs. Still, ugly and all, here it is, if you want something to compare to. In the future, not getting hit in the face first and maybe toggling on resistance toggles from the start would be better, but I get about 3:27 or 207 seconds from my first activation of PA to when she goes boom, which is an effective 313 dps (~128 from -regen, 185 from actual damage). With a little practice tuning it looks like I could eventually turn in a sub-3min time, but that might be close to the limit with my current build.

Now going to go see Hunger Games.

With tier 4 interface (reactive) and Destiny refilling my end bar I got the time minus the time to clear the vicinity down near six minutes. That's a rather dramatic difference considering I forgot to turn on Enervating field until about a minute into the actual fight with the Quarry, duh.

Hmm, I should probably test this build against a pylon next, that's actually a pretty decent time for not using Lore pets.

But those things get *better* as they add tools, they *start* off pretty good.

Also youtube finally did its thing: Ill/Rad vs Quarry. And in not-too-terrible-on-the-eyes 720HD (the original is about 1040). Gonna see what happens with a proper interface power next.

My Mind/Psi is too low, but I have a Mind/Earth I could try to do some measurements with. She doesn't have blockbuster recharge, but she is perma-Dom. I might need to boost the build by a significant amount to make her perma-mass confuse. I'll have to investigate.

No, but I don't think we're going to come to agreement on this. You view the blaster archetype as containing powerset options for the players. The decision is to play a Blaster, and then its up to the player to pick the "right" powersets to deliver the performance they want. If even one single powerset combination delivers a particular desired performance level, you view Blasters as an entity to be "fine" overall, and the only problem really is the diversity of options available.

But I don't view Blasters that way. I think they should be represented by their average, not their highest possible options. The blaster archetype should have a set, or a set of sets of features that every powerset should contain and be able to leverage. Otherwise, assuming Archery and Mental didn't exist, you could say Blasters were not the king of damage, but if I buff just one powerset, well now they are.

In one sense, that's mostly a semantic difference in terms of describing the archetype: its either underperforming with some exceptions, or it performs fine except for all the exceptions where it doesn't. But that difference also contains an approach difference that is significant. I believe when an archetype underperforms, you should try to fix the archetype with archetype-wide solutions. In your approach, there's nothing wrong with Blasters that individually buffing every primary and every secondary wouldn't fix.

The problem is one of those is practical, and one is not. The devs are not going to attempt to solve an archetype-wide blaster problem by handcrafting solutions for each primary and secondary, nor are they going to add massive tier 9 aoe and a drain psyche equivalent to each secondary. But these are such extreme outliers that there's no *other* way to really approach what they can do. You can't emulate -regen with other effects: other effects would be stronger in some cases and weaker in the extreme ones - the very extreme situations you are using to demonstrate its power. Similarly the devs aren't going to just hand a Rain of Arrows equivalent to every blaster: even if they did that wouldn't address problems below level 32, and it would force blasters to build for ultrahigh recharge to see most of the benefit. And there's a huge difference between perma DP and non-perma DP: when its perma its up when you have to use it again. If its non-perma, using it exposes most blasters to extreme vulnerability. That's probably a large part of its intrinsic design tradeoff.

And even after all of that, that would only solve the problem of making maximum potential roughly equal across the board. It would leave behind the problem of addressing normal players' blaster problems during normal leveling, which have a higher priority to resolve. I think its pushing the limits of reasonableness to say that in exchange for blasters dying more often and giving up all other options besides damage, they *eventually* and *sometimes* become the king of damage. In my opinion, nothing is worth that particular prize.


We added containment to controllers, we didn't individually tweak each powerset's damage to solve that problem. *After* we added containment as part of the baseline, *then* we can look to tweak individual powersets, like Gravity was. We added gauntlet to tankers, we didn't tweak each tanker powerset to manage aggro. In keeping with the way we do this everywhere, *first* we should establish a baseline of features that all blasters should have, and *then* we tweak individual powersets if that becomes necessary once the average is more or less where we want.

Tried it last night and managed it in just under nine minutes (8:42 or so) including clearing some DE around the quarry - I put a Quarry into an AE arc I was working on and did that instead of running out to Monster Island. I capped it, but my fraps is set to cap at full resolution, so I converted it to something youtubeable before I went to bed and now I'm uploading it to youtube, which will take about an hour and a half (its still plenty big: compression reduced it from 13 gigs to 550 megs). Youtube will then do its processing thing, and it should be viewable a bit after that.

Its been a while since I tried this with an intrinsic level 50 GM (I've taken down the Jack/Eochai pair solo not that long ago, but those are far lower intrinsic level monsters) so I almost messed it up: I got greedy right at the end and thought I could finish off the Quarry without using one last PA. Mistake. Fortunately I deployed them fast enough to prevent a catastrophe.

Although it would have been less of a problem if I had been running my defense and resistance toggles. I deliberately attempted to do this using no insps and without using my Destiny (Ageless Partial - End/Rech). With unlimited endurance, I could have gone faster, but I wanted to see what the build could do intrinsically. You can see spamming fireball slowly drains end, not spamming it slowly recovers end, so there's a balancing act there. This might be a good place to consider using enhancement boosters: specifically targeting endurance reduction. Maybe I can get away with going full speed even without Destiny.

Also, with smart placement of the pets its actually more efficient to let the Phantasm die and recast him than try to keep healing the moron, which is why I let the fool die in the video. He has no melee attacks at all but he insists on standing almost directly under the Quarry. Maybe he's aiming for the Quarry's crotch or something: he did seem to knock the Quarry down often.

Incidentally, I really need to spend more time on Aurora. Not only did I not finish slotting her (I mostly finished that last night, although she's still missing some inventions I thought were not necessary for that test) I also have a gigaton of incarnate components and yet I haven't finished up her incarnate powers either. She only has Spiritual Total Radial (tier 3) and Reactive Interface (tier 1!), plus Lore pets that aren't used here and the aforementioned tier 3 Destiny. I really need to craft a tier 4 reactive and try again at full strength.

Its not that I don't play her, its that I keep forgetting to finish her because its normally hard to notice her build isn't finished in regular play. I can't believe I played this long with a bunch of placeholder SOs in Blind (that was rectified with Apocalypses before the attempt).

In any case, 8:42 implies about 136 DPS plus regeneration debuff (or 490 DPS if you prefer to count that way), which is so-so because it includes the enervating field debuff. I gotta believe that among regen debuffers that's probably just average. I'm thinking a soft-capped Fire/Psi dominator could beat both builds.


Another thought occurred to me last night. If we're going to consider -regen to be equivalent to damage in the single target case, then we have to consider another potential sideways challenger to the AoE case: mass confusion. Suppose we had a perma-dom mind dominator that could perma mass confuse (its possible). That could translate into huge kill speed and effective damage. You would not get 100% rewards for that, but if you tossed in your own damage enough you could get the majority of the rewards, at essentially zero risk.

If -regen on a GM counts as single target offense, shouldn't mass confusion count as a form of AoE? It kills the targets, and if used correctly you still get most of the rewards. If, say, you only get 75% of the rewards per kill but you kill 33% faster, shouldn't that essentially be a tie? (0.75 * 1.33 = 1.0).

Which then brings up this question: is a really odd but actually legitimate challenger to max recharge Archery/Mental Blasters a max recharge Mind/Psi/Fire dominator?


Edit: youtube link now working: http://www.youtube.com/watch?v=UDaUBCT_GM4

Well, Drain Psyche isn't "balanced" because the devs never get regeneration correct. If it was, Dominators wouldn't have a stronger one, given they have far more control to use to leverage it.

Rain of Arrows isn't balanced as an offensive attack in the extreme recharge case because it was given to Blasters back when Archery really sucked. Archery's been buffed twice since then, but at one point Rain of Arrows was the consolation prize for having extremely slow and low offense outside of that. Whenever you trade single target for AoE, there will always be a point where given high enough recharge your AoE damage will outstrip any single target sacrifice you could have made, or in the case of archery not even made as much anymore.

The devs just don't care enough about the extreme recharge case to nerf that case if it would cause problems with the lower recharge cases.

As soon as I'm finished with it: actually THB deserves credit for me thinking about making something that isn't a challenge mission in the classic sense of being able to outlast or outmitigate a monsterously lethal mission, but to design a set of missions that represent something perhaps slightly higher than normal difficulty, but based on in-game challenges so as not to be too outside the realm of testing for things that have any practical value. Something that tests offense, defense, and tactical options.

I've decided to combine that with Aeon's request for ideas on making scaling missions whose difficulty doesn't require throwing a bunch of AVs at the players: I'm using most of the scaling tricks I know (including with AVs, but also without) so the arc also works as a -1x1 through +4x8 scaling test.

I hope to have it ready for initial testing in a couple weekends.

The "multiply offense by mitigation" theory is something I wrote up in late 2006 and early 2007, shortly after my PeakDR series (because it used PeakDR, which despite the name was actually a damage metric). To say its "your" theory sort of ignores the fact that that very thing has been discussed by me and others even in the most recent threads on blasters as a component of the problem.

Its also not a theory, just an assertion, until you put real numbers on it and attempt to actually prove it convincingly.

It depends on how. Direct damage mitigation, for example, increases your survivability in one fight by decreasing your damage over time in that fight, which is extensible to multiple fights. But increasing survivability by increasing kill speed does not decrease your damage over time, it reduces the length of the fight to increase your survivability. And that's *not* directly extensible to an entire mission of those fights.

I'm probably wasting my breath here anyway, but I'm in the process of writing this up for a dedicated article so I might as well outline the thought process here. The problem with increasing survivability with kill speed is that reducing the length of the fight doesn't mean the player spends less time fighting. Players tend to move from one fight to the next: they don't tend to wait around. If the increase in kill speed retains the same average damage per second *during the fight* and then shortens the fight, the net result will usually translate into a higher, faster kill speed, but with more fights in less time it will also translate into roughly the same average damage per second over time. If the player used that higher kill speed to slow down, pace themselves as if they had killed slower, then that extra kill speed *would* translate into higher overall survivability. Per window of time, they would kill the same number of things and take less damage (because the targets were alive for less time). But that's not reasonable to expect to happen. Instead, the duration of the fight drops, the damage drop off becomes steeper, but the average damage remains similar within the fight, and bookending those fights together results in a similar damage per second and a similar, not lower survivability per unit time.

In the scrapper secondaries comparison, a mention is made of "sawtooth" damage curves that represent a simplified damage over time curve. The average damage of those triangular damage curves is 1/2 the height of the triangle. Increasing kill speed steepens the triangle, shortening the base, not not changing the average. If you speed up the fights but don't pace them apart to the same distance, damage doesn't go down, survivability doesn't go up.

It can get worse. Using AoEs at the start of a fight can temporarily steepen the damage curve downward, and then cause it to level out. That shape curve has a lower area under the curve, which corresponds to less average damage over time. But using AoEs *later* in the fight creates a different shape: a shallow drop followed by a steep one. *That* curve has more area under the curve relative to the baseline, and that means that generates more average damage per unit time, and thus less survivability.

Bursty or random damage increases won't align to the optimal place for the to occur, which is the very start of the fight. They will occur much more randomly. And when you have periods of higher survivability and periods of lower survivability, that doesn't always average out. It often nets to less survivability.

The reason for that is that you cannot bank health. During periods of high survivability you may find yourself at full health. All the regeneration you could be generating goes to waste. Then when you are at lower survivability periods your health will drop faster than average and you won't be able to average that out with the regen you lost for being at full. If your survivability is more even, you have less periods of full health and low health, and more periods of averaged health when you are most likely to get the most from regeneration over time.

When I get home I'm going to try this on my Ill/Rad after I finish her slotting. Seems I've been playing her for about three years without all her powers slotted: must have slipped my mind and honestly I haven't noticed. But for this sort of thing I would want to run at full power. I'm pretty sure I can beat 20 minutes without incarnate pets if I can sustain maximum activity without running out of endurance, which is sometimes a problem running everything. Which happens when you forget to slot them all.

Contrary to popular belief, Illusion's damage ain't that great (its ability to tank hard targets with the PA is its real power) so I would consider this to be a mediocre debuffing example. I would expect some builds to do worse, and others to do better against a single giant monster class critter.

Been thinking about adding a giant monster to my performance testing AE arc anyway: good time to do it.

In many other MMOs you could probably charge real money for recall friend taxibot service alone.

That would I guess force me to become _@rc@n@v!113_!!!

Actually, I originally asked for the 54x8 because I thought that would be more instructive. I asked for the GM one because I thought it would be too clumsy to use Drain Psyche in that situation relative to its single target benefit, but it seems with a high enough smashing defense it will at least work against a Quarry. Although it takes a really long time because the recovery just isn't there to power offense fully.

As can all archetypes. Soloing a GM is something less than 1% of the players are ever likely to do, or be able to do. That's an accomplishment unto itself. That any Blaster can do it is great. But that's entirely besides the point of whether the blaster archetype is actually as a whole designed correctly.

To put it another way, if the devs nerfed Drain Psyche tomorrow, if they increased the recharge of Rain of Arrows tomorrow, what would be your complaint? There's nothing in the blaster specification that mandates it have those tools. Rain of Arrows has been known to be broken for years: the devs only let us keep it because historically Archery has had low damage otherwise: it was really bad before it was sped up. Drain Psyche is something we get because the devs always, always, always get regeneration wrong. They could learn to add one day.

You keep saying that there's nothing wrong with Blasters because you could just bring every other set up to Archery and Mental. But that's not true, because you would also have to slot them for the extreme recharge you have. You cannot sometimes be able to survive x8 and sometimes not: you can't set a mission that way. People need to be able to level Blasters, and play them at all combat levels and with normal slotting. The fact that "Blasters" work when you happen to take archery and happen to take mental manipulation and happen to slot them for extreme recharge is an interesting anecdote, but we cannot design the entire blaster archetype on the assumption that Blasters will always have cycling tier 9 rains and capped regeneration. Archery/Mental doesn't even get that now for most of its existence.

As an accomplishment, its a great accomplishment, regardless of the powers used. As proof Blasters don't need help, its a) not relevant and b) 100% guaranteed not to influence the devs in any way.

Even at the -res cap you'd be hard pressed to bring the single target damage of even a well-slotted blaster up to 400dps+.


Its not having it both ways. We call the damage dealers "damage dealers" because they are designed to deal more damage in most circumstances. But there can be corner cases when a non-damage dealer will do better, and -regen is one of those cases that has been known and accepted for years. If you're going to compare damage output to the conventional damage dealers, you have to compare in those normal circumstances. If you purposefully aim for a corner case where the normal damage dealers might not even beat the support characters consistently, then you have to compare across the board and not just to them.

As I said, damage isn't that straight forward to compare. Just because A beats B in one circumstance, doesn't mean it will beat B in all of them. You're picking the area where melee archetypes are weakest and claiming victory in damage output without comparing to the real Kings in that one area: the debuffers.

In any case, as I've consistently stated, in general Blaster performance should normally be compared to the other stated damage dealers: Scrappers, Brutes, Stalkers, and Dominators. But its fair to compare them to other archetypes in specific circumstances. That comparison makes the most sense when talking about whether Blasters deliver enough on their offensive output role to compensate for their general lack of survivability.

Which would at least make your single target damage closer to 90dps not 55dps, which seems less weirdly low (average blaster single target damage should be closer to 120 dps at level 50).

As I said, that's not true, but I'm working on a more... interactive demonstration of the principle.

Wait wait wait. That's going too far. The notion that you can buff the other powersets to compete with Mental is not true, because it seems at least on the surface that almost 90% of the damage you display in those videos is actually the regen debuff of Mental. How do you intend to do that: by giving every blaster secondary drain psyche? That's not going to happen.

If you're going to try to claim the King of Damage crown with a GM take-down leveraging -regen, you have to compare to other regen debuffers specifically. Because of the mechanics of regen, apparent damage is not transitive. You can't say that just because you beat the melee archetypes - who tend to happen to not have -regen - you're damage is clearly the best.

At the very least, you should compare to the other damage dealer that also has Drain Psyche - and a more powerful version of it - Dominators. The archetypes (excluding epics) that the devs have declared to be designed to be damage dealers as their primary or balanced role are Blasters, Scrappers, Brutes, Stalkers, and Dominators. The King of damage should at least beat the best of those, and also tend to beat most of the rest. But defenders, controllers, and masterminds all have -regen also, and could achieve similar or superior results.

I can see Positron practicing his Peter Cullen impressions in a mirror going "One shall stand, One shall fall" and then ordering the art team to make him a giant statue of Positron to replace the globe in Atlas Park.

Its complicated. Knockback involves three separate systems each of which does something different. The boolean mez system is used to determine if you are "knocked" or not. Anything in positive territory means you are in the "knocked" state just like holds. But the knocked state doesn't actually do anything in terms of the powers system. Holds do something. Immobilizes do something. Knock causes you to be passed to the physics engine to be thrown about. That does whatever the physic engine wants. On top of that, the animation system is instructed to play the fall down and get up animations at the appropriate times. And the animation system is also free to do whatever they want.

A knock with duration longer than a combat tick would basically tell the game engine to knock you, and then knock you again during the next combat tick. But that's squirrelly. For one thing, knock stacks within a certain time window. So a knock with duration might stack with itself. Its also unclear what happens when you're knocked continuously while you're moving: the knock magnitude might originate from the caster, or it might simply loop the same knock over and over again. The difference might be significant: it might be the difference between looking like someone on the ground hit you with a knock over and over again and having a jetpack strapped to your butt.

Mostly, though, I think something weird but not necessarily spectacular is likely to happen. Although given what happened when they accidentally created a heal with duration, its hard to say for certain.