Arcanaville

Yes, but technically they don't influence each other multiplicatively: that's just how the math works out. The more straight forward way to look at it is that +Damage increases the damage per attack and -Res increases the damage the target takes per attack independently.

So +Damage will make an attack go up 20% (for example). -Res will make the damage a target takes go up 30%. These two effects happen independently, so an attack that has 100 damage without these effects will get increased to 120 (100 * 1.2), and then when that damage arrives at the target the target will take 156 (120 * 1.3).

It isn't that they affect each other multiplicatively, its that both are independent damage multipliers. Multiplying by two things is the same as multiplying the two things together and then multiplying by the total, which is why it looks like they multiply.

But whenever in doubt, just work it out with one attack step by step, dealing with each factor in turn, and it will generally be obvious which things add and which things multiply.

I was going to respond to your other statements, but I've decided instead to simply state have FUN beating your head against this particular rock, because PvE isn't going to ever get pervasive diminishing returns comparable to the PvP system, ever.

And let me know if you want to bet on it.

Unless pohsyb has been torturing new players again, there's no way your team mates could have known what your recharge was. The only way that statement makes sense is if you were kicked for an exemplared high level empath that was presumed to have superior slotting to you, inventions or not.

In any case, I still don't bother slotting IOs except what happen to drop automatically until basically level 50 (sometimes I start at level 47 for obvious reasons). I've never been interrogated on my build by strangers on a team, and I'd tell anyone who did for the purposes of determining my fitness to be on the team to go climb a tree.

Plus, I can solo Positron. With SOs. If your team mates needed an extra couple of percent of global recharge for Positron, they had problems that DR isn't going to fix.


On the subject of implementing DR game-wide, while I recognize the intent of DR in PvP, I don't agree with the DR curve implementation. As a result, I would oppose implementing it in PvE, and the devs know this. However, that's a moot point because Castle is never going to do that.

On the subject of inventions in general: the devs have stated repeatedly and directly that the point of the invention system is to provide higher levels of performance for players that want to participate in it. Its optional in the sense that they are not needed to be competitive with the PvE content as originally designed. However, if your definition of "optional" is different, then your definition of optional is incompatible with the game's design.

It used to be Castle, but Castle wisely expedited BaB's redname forum account three minutes after he was hired.

In terms of money spent, my guess is about $1300, plus or minus. Relative to the value of the time I've invested in the game, for me the game is practically free to play.

The Arcanaville forum is honestly not all that interesting. Only one person ever posts there, and I think there's only one person reading it as well.

I believe Stargazer tested this (there may have been others, but I remember her article specifically). The wiki is correct. When you debuff a resistance value, the non-debuffed strength is used to resist further debuffs, not the debuffed strength.

Actually, I recall there was a further complication. Unresistable resistance debuffs actually reduced the resistance to further debuffs.

In other words, the game engine seemed to be doing this:

Start: Initial Resistance

Next: Working Resistance = Initial Resistance - Unresistable Resistance debuffs

Last: Final Resistance = Working Resistance - Total Debuffs * (1 - Working Resistance)


I may be misremembering, but that's my recollection.

/Summon Stargazer

I had a chance to profile CoX in more detail on my i7-860. I'm assuming you either have an 8xx or 9xx (4 core hyperthreaded to 8), in which case I doubt GR is going to come anywhere close to maxing out your processor. But there is a lot of potential headroom. The best I could do was run around in a high-density mission (x8) with a lot of stuff going on, while demorecording, with 3D sound, and with high graphics setting and high particle count but low intrinsic resolution (low res because the 4350 I have at the moment is not suitable for high performance graphics and bottlenecks the system at high resolution). I can get the primary compute threads to approach 10% total system utilization (which is approaching one core's worth of workload) but things like the PhysX thread and the sound thread are only using very small amounts of CPU even under high load conditions. If they added features to Ultra Mode that, say, doubled the maximum CPU load for the game client under the top performance conditions, it would probably benefit from an I7 over, say, a Core2 Duo (or even a Core2 Quad). But you'd probably still not have the CPU be a significant bottleneck.

So far, the multi-threaded and multi-core overheads appears to be trivial.

Just a reminder, its not a simple case of CoX using two threads and therefore two cores. CoX is actually using 15 threads, most with very little CPU relative to the primary game client. But starting a demorecord, say, kicked a previously idle thread into just under 1% utilization, as did turning on 3D sound. I did have six threads with measurable CPU performance running simultaneously under some test conditions. Win7, as I mentioned previously, seemed to be doing a good job of isolating those six threads to the four even-numbered CPUs (which are the non-hyperthread ones) without pegging any of them (the highest CPU, CPU zero, only reached about 70% utilization maximum and I did have other things running at the time).

City of Heroes seems to have three main active threads while running, two burning a lot of CPU which are probably the main thread and the physics thread. There is a third thread that burns minor CPU probably related to networking. And then there appears to be a fourth thread that looks to be related to the main processing of OpenGL that is probably system dependent (depending on the OpenGL libraries on your system).

However, City of Heroes seems to open between 14 and 15 threads total, most are just idle most of the time.

That's a good question. You know, based on what I understand about the new phased attribute, I'm not sure its *possible* to have protection against being phased. It might not be, and that might or might not be working as intended. I'm honestly not sure, but I'll try to find out.

Enemies burn more endurance than we do. We don't notice this often because they don't have as much attacks. Without recharge slotting or hasten, try running out of endurance firing only two single target attacks.

Lets take a Rikti Soldier Boss. He's got 200 points of endurance, which means he recovers 3.33 end/sec. The Greater sword attack burns 13.72 end and he can cycle it about every 9.5 seconds (9 second cycle time, 0.5 second AI make up his mind time). That's a burn rate of 1.44 eps. If he also fires his ranged attack, that burns 11.48 endurance with a cycle time of about 9.4 seconds, and a burn rate of about 1.22 eps. That's a total of 2.66 eps, which is lower than the recovery rate, which is why they don't tend to run out of endurance. However, each recovery tick (at 3 second intervals) only recovers 10 end, which is less than either attack takes. *If* you drain this boss to zero, he would need a minimum of two ticks (6 seconds) before he had enough endurance to attack. That becomes important.

Hit him with a -50% EnduranceDiscount, and things change. Now all his attacks take twice as much endurance. His net burn rate jumps to about 5.32eps which is higher than his recovery: his net loss is just about 2 eps, which means the boss would run out in about 100 seconds of fight on his own. On top of that each attack now requires at least three recovery ticks (9 seconds) before it can be fired. We can keep going and debuff lower, until we get the results we want. At -67% (three times endurance cost) his sword attack will burn 41.2 end per attack and 4.33 eps all by itself: draining to zero would then leave the boss idle for at least four recovery ticks (12 seconds) because the ranged attack would need a minimum of 34.44 endurance to fire. If anything, the problem with this effect is not that its too weak, but its too easy to make too strong. It almost certainly couldn't be made stackable or you could make a boss inert.

By the way, any guesses what the EnduranceDiscount debuff floor is? Its 0.0001. In other words, the debuff floor for EnduranceDiscount is increasing the amount of endurance your powers take by a factor of ten thousand (thank god critters haven't figured out how to do this yet). I'm pretty sure somewhere between doubling the cost of your target's attacks and multiplying the endurance costs by a hundred there will be a point where this effect is strong enough to matter. I think if this effect was put into play, that minimum would have to be increased to something saner, like maybe 0.05 or so at worst.

PS: a note on thinking about extra nasty effects for player powers. Those would have a good chance to get backported to custom critter powersets. That might make life interesting for people who run AE missions with custom critters in them.

I'm not sure there are enough ways to do that which aren't tied specifically to numbers, and I'm generally opposed to objectives explicitly requiring a specific number of players on the team for no other reason than to place a numerical requirement on the minimum number of players required. I don't see why, at least in theory, a controller can't throw two switches at opposite ends of a room (if the game mechanics allowed for such things). If there were, I would want to explore those options as well.


Except I wouldn't make death a mandatory factor. Suppose I'm a blaster facing an AV and I discover or conclude I can't beat it. I think its legitimate to use a Power Analyzer on him, collect some "information" and then simply run away to use that information. I don't see this as a "fight him, and if you die we'll rez you with more stuff" mechanic. I see it as a "live to fight another day" mechanic where the player deliberately decides to change the rules. So a stalker sneaks up to the AV and steals a component from his super armor for analysis, say, and simply forgoes combat altogether. That sort of thing is something I think should be an *option* for players when they think straight-forward combat won't work.


Completely independent goals rather than multiple victory conditions begins to drift into the area of making a more sandbox-type of gameplay. The problem with going too far in this direction is that while I think you can incorporate some of it without consequence, at some critical threshold you have to go all-in, because completely optional objectives requires a completely different kind of narrative structure for the game. If you can't depend on players doing certain things, you can't write stories that rely on specific sequences of events. You'd need to think about narrative composition in a completely different way.

I think the concept of multiple victory options is orthogonal to what I'm thinking of rather than an alternative to what I'm thinking of. Even with a mission with multiple paths to success, you still have the potential problem that to make it legimitate "team" content it has to be very difficult in all its victory permutations for a solo player to accomplish. And since the range of capabilities for individual players is high, the range of capabilities for teams of players is also high. And that means its very difficult to make "hard" team content that isn't impossibly hard for too many teams, or too trivial for too many others. You still have a higher emphasis on binary failure than I think is a good idea.

I think multiple victory options is a good idea regardless, its just addressing a different issue to me: that of gameplay diversity in general.

Heals are always perceptually more important for Regeneration because regeneration is always on. As a result, you're only ever seeing heals stacked onto regeneration. Its extremely rare that people see heals without the regeneration and then the regeneration without the heals for extended periods of time to allow for comparison. It would take careful deliberate testing to tease the two effects apart.

This is very much identical to the old saying that you always find things in the last place you looked. That's always true, because once you find it you stop looking. Heals are always the final determining factor for survival, because they are the last - and only - option when regeneration is always on. Either a heal saves you, or failing to use a heal kills you, or you were doomed either way. You can never see a case where regen saves you after a heal fails to do so, because most people don't define failure in terms that allow for that possibility.

That makes my keyboard more important to survival.

If you're saying heals are always better than regeneration regardless of the difference in magnitude between them that's your prerogative. However, mine is to state that its generally without merit. What I can say is that the mechanical difference between regeneration and on-demand healing is something that can be analyzed quantitatively**, and doesn't generate the results you imply for most interesting non-degenerate cases. Which means as a subjective opinion, you're entitled to that perspective in the sense of exploring subjective reasons for valuing in-game capabilities. However, that perspective plays essentially no role in either designing or balancing the game.


** The "on-demand" nature of heals was examined in my irrelevant objective analysis back in I7. Basically, the quantitative advantage is entirely due to frontloading, which can be calculated in a sliding-window analysis. That advantage dissipates rapidly except for the singular case where the scrapper deliberately waits to achieve the state of being at full health *and* fully recharged heals before initiating any fight. Ironically, the only time this doesn't incur a penalty that exceeds the advantage is when you have high regeneration relative to the incoming damage curve. In other words, this "on-demand" perspective turns out to be quantitatively rubbish. Its sole benefit is playstyle-related: some players find it easier to manage heals than pace regeneration. Playstyle preferences have nothing to do with intrinsic value, however.

Positron was giving recommendations for nVidia and AMD cards, because AMD now owns ATI. I'm pretty sure he put "ATI/Radeon" in parenthesis to remind people, not to suggest his recommendation was for an AMD CPU/ATI Card combination.

An effect that would be meaningful is EnduranceDiscount debuff. In other words, debuff the critters so their powers burn more endurance than normal (the opposite of Conserve Power's effects). It means they actually do tap out their endurance and can run out, and it also makes actual endurance debuff and recovery debuff more effective besides.

In trying to hypothesize an MMO that doesn't have required roles, I think sometimes we try too hard to imagine an MMO that doesn't have any requirements *at all* for its encounters and that's actually silly.

I think its bad, or at least limiting, if a game has encounters that require "a healer." But I don't think its at all unfair to make encounters that require a minimum level of damage mitigation of some kind, such that teams below the minimum are likely to fail.


On that subject, I'm someone that believes the whole concept of "failure" in general is something that isn't well explored in the MMO space. Failure is usually binary: you win, or you lose. There's no in-between. And without an in-between, there's less opportunities for experimentation and learning.

Consider the case of an AV in City of Heroes. Its supposedly not intended to be soloable content. That doesn't mean AVs are impossible to solo, it just means they aren't designed to be soloable. If you can't solo any of the AVs in the game, that's not a problem: that's working as intended.

What that means is that if you aren't one of the players with the build and combat skill to solo an AV, there's two possibilities when you encounter one (assuming it isn't downscaled to an EB in your presence). One: you run. Two: you die.

There's no third option, which is actually the most common option in the genre when this type of situation occurs. And that's Three: you learn something from the encounter and come back with a greater chance for victory.

I'm not suggestion what I'm about to say would be easy to implement across the entire game, but suppose that an AR/Dev blaster runs into an AV, and can't defeat it. Instead of being locked into the choice of essentially permanently running away, or fighting until death, what if an additional option was for the blaster to recognize they were not going to win the fight and change the goal from defeating the AV to collecting information on the AV instead. Perhaps they record the fight and then run, and from that recording develop additional weaponry specifically designed to assist with that AV. On the next encounter, the blaster would have a little more damage against that specific target, or a little more resistance to the offense of that specific AV. Perhaps with each encounter, the blaster could use the invention system to craft increasingly stronger temp weapons which would eventually allow the blaster to "solo" that AV. Or alternatively the blaster's primary attacks gained increased effectiveness against that target somehow (increased efficiency, better targeting of weak spots) in the same way Nictus weapons are tuned to affect Kheldians. I don't think it would be conceptually unpalatable to require "temp powers" if they were done right: Martial Artists could learn a special strike to use against the AV, Electric Blasters could learn to tweak their energy discharge to disrupt the AV's neural system, whatever. The point is that this mechanism doesn't have to literally be "craft an I Win gadget." Heck: players might even enjoy collecting these things.

In between successfully soloing the AV and failing (dying) against the AV would be a large grey area of not exactly winning, but not exactly losing either, and building up the ability to eventually win. Instead of having "soloable content" and "not soloable content" there could be "intended to be soloable by anyone" and "may require significant effort to engineer a way to solo" and finally "not intended to be (but could be) soloable even with unlimited effort."

Making this work in a way that was palatable in teams and not just solo would require additional creativity, but I think its still possible. In fact, in teams I could imagine some members of the team trying to keep an AV busy while one or two others were actually crafting the AV enhancements right there in the fight. Imagine I'm an MA/Invuln brute whacking away on a Hero that I'm trying to keep aggro on, while a mastermind was working on an AV enhancement because the team was unable to defeat the Hero. Conceptually, the MM would analyze the Hero and determine its weakspots, and then communicate that to me. "Hey, try this!" Mechanically, he'd craft an invention that he would then use on me to grant me a special power: Mongoose Strike or whatever. Then I would start using that power against the AV, which would significantly increase my damage against him, or debuff his incoming damage so I could survive longer, or whatever. Maybe not this exactly, but something like this would, I think, add significantly to the options available to team balance and soloability.


Getting back to the question of required roles. I think, to oversimplify a bit, that team encounters should not explicitly dictate roles or special individual characters, but its fair game to say that the encounter requires a certain amount of X, Y, and Z, and allow the players to assemble those between the team mates in flexible ways. And if the team finds itself short on Y, there should be sufficient hints that they are short on Y and give them a path to get it, up to a certain point (in Champions Online one form of flexibility are roles/stances, and in Star Trek Online its power distribution settings; City of Heroes doesn't have quite the same type of options yet, although we've come close once or twice).

If you're really crafty, the best of all possible encounter "arithmetic" would be if the encounter required some amounts of A, B, C, and D and success required (at least as a minimum requirement) f(A,B,C,D) > Min, but there were many possible combinations of A, B, C, and D that would satisfy that requirement (or many different ways to get A-D, or both), and thus there would be many different kinds of team composition that would work. That's difficult, but not impossible, but usually requires significant cooperation from the underlying game mechanics to pull off.

I meant to imply that it was impossible for MMO developers to ensure with their game design, not that it was an impossible circumstance for individual players to engineer around themselves. It is of course always possible to feel simultaneously needed by the team and not need them yourself if you choose to only team with people that don't you don't need.

But for a game to attempt to engineer this as a player-entitlement as opposed to a happenstance requires setting up unfair teaming requirements, which is why I said its impossible to do so for all players fairly.

The problem is simply put, we're running out of effects its ever likely we'll be able to get that work predictably.

The singular pervasive effect I would add to MA if it was my singular decision would be -DMG. Something within a binary order of magnitude of -7.5% dmg stackable, with CS approaching -15% dmg (bearing in mind that Dark Melee's attacks have a -5.625% tohit debuff which translates on most critters to a -11.25% incoming damage reduction, and ToF has twice the debuff strength). I'd put an escape hatch for AVs, reducing the debuff to -1.5% on AV-class critters through requires clauses (because AVs generally have very high resistance to tohit debuffs: this adds comparable protection**).

My suspicion is that the strongest opposition from the devs would come from three angles: 1) its too defendery, and they don't want to allow Dark Melee to provide precedent, and 2) its too strong for a stacking effect, again in defiance of the Dark Melee precedent, and 3) that sort of effect is not in the set of effects the devs intended for the MA concept (its not a "secondary effects" in this context).

The fencelines for what MA *can't* have, or rather are unlikely to convince the right people to grant, appear to be extremely tight.


I will say, as an aside, that while I agree with the notion in general that character performance should be based less on random chance and more on skill, you really don't get to assert that very far in this game, because this game is explicitly designed to lower the skill curve and replace many skill-oriented activities with game mechanics. That's why we have a defense attribute, and can't dodge attacks manually. That's why we don't have a blocking mechanic. That's why we have auto-targeting. It would be more interesting if you could decide, as a player, to attempt to debuff your target's accuracy with a carefully targeted strike with thunder kick, and have that effect occur if you executed the attack in just the right way, but there's no real way to do that in City of Heroes. The closest thing to that are combos which are problematic in this context and also currently intended to be unique to Dual Blades.


** Reducing from 7.5 to 1.5 is an 80% reduction. However, damage debuffs would also be susceptible to damage resistances, making the average AV likely to resist more than 80% of the damage debuff, and more inline with the net resistance to tohit debuffing that AVs possess

I'm aware of those effects. Given that I've looked extremely carefully at those mechanical effects as well as all other mitigation mechanisms both numerically and in-game, I reiterate my prior statements.

Furthermore, I don't think this is an assertion immune to objective analysis.

The bolded part is not always true. Some players will see really fun content that requires a team and say "that looks like fun: I should team to see it." Others will say "that looks like fun: how dare the devs force me to team to see it."

The problem is that for some players, the definition of "optional" is "I don't care about it." Or to put it another way, for some (all too many) players, the phrase "fun, optional content" is a contradiction: if its fun, its not optional that they have access to it.

But Kitsune seems to be suggesting something else entirely. Kitsune is saying of the people who at least sometimes want to team when they team they want to feel needed by the team, and yet somewhat hypocritically if they were to choose not to team they wouldn't need the rest of the team. So they need me, but I don't need them. That is a feeling that is logically impossible to satisfy for all players fairly, even though dev teams keep trying to various degrees.

I would tend to say, especially in this game, that if someone claims to have felt like they were not needed, its probably because they were in fact not needed, and leave it at that. The need to satisfy that feeling is not one of the current design goals of this game.

Part of BaB's plan was to get people to spend more time just Walking around, making it seem like there were more players logged in than normal.

Its been such a success that the next expansion after Going Rogue is tentatively titled "City of Tai-Chi."

I don't disagree about the 10% mag 2, but my intent was to suggest that MA powers would have *many* chances for different effects, and under those circumstances its less insulting for the chance for any one of them to be somewhat low, especially on fast-recharging powers. In effect, MA would become the "proc" set (personally, I think MA should have been the "combo" set, but since DB has that I don't want to dilute its unique ability unless I absolutely have to, and I don't think I have to).

Since I think Cobra Strike's stun uptime needs to be about 4 times higher anyway, I don't need TK to stack with it.

Maybe the higher the tier of power, the more potential proc effects it could have, until EC had a veritable rainbow of possible foe effects. I don't know if that is really practical, but it would at least be fun, and MA has basically had to survive mainly on being fun to look until now anyway.

I made Bruce Harlick promise me at least five.

That's not strictly speaking true. Regeneration has, slotted with SOs, about +391.25% regeneration. Alternatively, that's equivalent to 1.63%/sec of health recovery.

Reconstruction slotted with SOs returns 48.75% health about every 32 seconds, or about 1.52%/sec health return. Regen's +regeneration outpaces its heal.

Dull Pain's heal when DP is slotted with SOs returns 78% health about every 186 seconds, or about 0.419%/sec. *However* while DP is up regeneration is boosted while heals are not: relative to base health while DP is cycling you are getting 1.939%/sec of health recovery due to healing, but 1.63%/sec health recovery due to regeneration while DP is down, and 2.59%/sec health recovery due to regeneration while DP is up. The average, assuming about 66% for DP, is about 2.26%/sec.

Whether you account for DP or ignore DP entirely, the regeneration set is slightly more than 50% regeneration and slightly less than 50% healing. Technically, DP itself contributes to survivability in terms of its +health but that's more difficult to express as a single numerical ratio: I would describe Regen as a set in which more than half its health recovery is regeneration, and more than half its total survivability is due to health recovery in general.

I haven't counted Instant Healing in those computations either, which would further swing the scale towards +regeneration. I don't know why there exists a persistent belief that "most" of Regen's strength comes from self-heals, because that's never been true.


Willpower *can* generate regeneration numbers that are higher than Regen's numbers without Instant Healing, but on average Willpower's net regeneration is around the same as Regen's is, if you factor in an average value for Instant Healing.