Arcanaville

45 h/s would be closer to average for a level 50 Regen scrapper while fully slotted Dull Pain was up. Dull Pain is another one of those Regen powers that, while optional, is probably a really good idea to take.

Hate decays. If it didn't, a Tanker could taunt a critter, brawl it once, and for the rest of the day all the blasters on the server could attack it without drawing aggro. In-game testing seemed to suggest that while a tanker had a target taunted it was very difficult for anything without a taunt to grab aggro, but the moment the target wasn't either directly taunted or taunted by gauntlet, all bets were off.

So as long as the tanker is actually generating hate, I believe you need to overpower that hate by 100% to grab aggro away. But the moment the tanker stops generating hate, and especially if the tanker stops generating taunt or stops generating damage-based hate even with taunt up, I believe aggro can be lost very quickly.

I think this is a subtle contradiction of the conceptual intent of Granite. I don't think Granite is intended to be a way for Tankers to escape damage (I know its also a Brute set, but its conceptual foundation is as a Tanker set), but to sustain damage. And that implies that Granite will not explicitly make it difficult for the Tanker to draw damage.

That's specifically to contrast against Hibernation, which *is* a power explicitly designed to allow a Tanker to temporarily escape damage.


Mechanically, its difficult to critique an aggro idea without very explicit specifics. For example, the mechanics of taunt and hate may not allow the gentle expiration of aggro you're specifying.

I'm not trying to convince you to change your mind. I'm just pointing out that there's no way for the devs to honor your preference in this area in a consistent fashion, without either targeting the game design literally right at you personally, or honoring a whole range of other thiings that you might not personally advocate, but have no objective separation from your stated preference.

Consider that there are a lot people who consider even endurance cost itself to be too high of a penalty for a "casual" game.


One game design rule that I believe as an axiom is "something for everyone, not everything for everyone." By that I mean its more important to have diversity of experience than for everything to be equally accessible and equally inoffensive. So I think a well designed CoH would have some powersets that have relatively constant out of the box performance that required little work to optimize but also allowed very little head room to improve, some that start slow but can be improved to high degrees with sufficient effort, some that are balanced around opportunity costs, some that are balanced around explicit tradeoffs, etc.

There are examples of different balancing schemes now. Dark Armor is very much an opportunity-cost balanced powerset in multiple (and probably historically accidental) ways. Some people see Oppressive Gloom and Cloak of Fear as being "contradictory" but I see them as the perfect dual mez togggles for DA. Each one separately has a particular strength, but the combination is not the sum of its parts. Using one has opportunity costs in that it will interfere with using the other, and vice versa. But there are still advantages to running both in extreme circumstances. Dark Regen is a significant opportunity cost power in that its high endurance costs require very careful offensive throttling - it essentially costs you future offense to use.

Some people find these properties annoying. I don't. I think they are a legitimate option for players to choose.

Similarly, Regen matures quickly, and mostly without compromise. SR matures slowly, and with significant compromises along the way. But Elude's peak performance, when it becomes available, outperforms Regen's peak performance in most ways. Its arguable if the numbers themselves are correct, but the qualitative distinction between the sets is entirely valid. I don't look at those two sets and say that the problem is SR doesn't have quick recovery and Regen doesn't have Elude. I've never advocated erasing or even minimizing the qualitative differences between the scrapper secondaries, just equalizing their value proposition.

I think there is a place for sets like Dark Armor, Willpower, SR, Regen, and Stone, all of which are designed to deliver completely different tradeoffs during play. So much so that I consider it a failure in design when such options don't exist. It is that very lack of significant tradeoff options that I think is the fundamental error in design in the Champions Online powers system, in fact.


But again, I'm not trying to convince you to change your mind. I'm only saying that while I would not try to erase the particular style of powerset you enjoy, I would specifically attempt to prevent anyone from making it the standard of design.

Animations can be scaled spacially. But I'm not sure that changes the timing or playback speed of the animations. I'll have to look more carefully at that.

Back in I2 the period I was referencing, Regen scrappers could almost pretend to be perma-Elude scrappers with MoG but *without* the endurance penalty, just for giggles, on top of bailing out of MoG and simply running IH and perma-DP and getting virtually the same performance in a completely different way (if you happened to take recon as well, then you'd exceed perma-elude performance back then). But more to the point the relevance here is that Regen had tier 9 performance without having a crash: definitely in the case of IH/DP, and sort of in the case of perma-MoG. Both SR and Invuln had to suffer crashes to get their top level I2 performance.

The post I7 softcap is actually only half the performance of the Big Three from I2/I3 so its an incomparable performance regime. And because there's so many sources of defense and SR is no longer the sole possessor of high defense, softcapping with inventions and pools is no longer unique to SR. SR was probably in the driver's seat in melee defense from about I9 to about I11 when Willpower came out.

Perma-Elude did have certain advantages: for one thing at one time it had base defense and thus did not have the non-positional hole SR has now. But in I2 its peers were perma-unstoppable that *also* had massive defense in invincibility (not as good, but almost) and regen that could flip between almost as strong defense and almost unlimited health recovery. Of the three, perma-elude had the most reasonable cost/benefit ratio. Only perma-elude and perma-MoG had *any* legitimate discussion over whether their costs outweighted their benefits. There really wasn't any serious discussion about whether IH/DP or perma-unstopppable+invincibility was anything other than ideal.

To the best of my knowledge, every time you think you're seeing the game engine speeding up or slowing down an animation, the animators actually made a faster or slower version and instructed the game to play it in that situation. Sort of.

The game engine can speed up or slow down an animation. It just has to be told to do so by the animation sequence entry that calls it. But its those sequences themselves that cannot change dynamically as far as I know. So if the animators want to play a faster version of an existing animation, they can do so by creating a different animation sequence that just happens to call that existing animation, but at a different speed. But there's no way to actually speed up the animation "on the fly" as it were. This has to be hard coded into the game by hand for each individual situation you want it to occur in.

That's a matter of personal preference, and moreover a matter of psychological perspective as well. In I2, Elude's crash was not an out-of-combat crash for perma-elude scrappers: it occured generally during combat and they had to figure out ways to deal with it. And contrary to some people's recollections, not everyone went perma-Elude: some SR scrappers simply found the continuous crashing unpalatable and remained in toggle builds. And they were not the vanishingly small minority: at the time maybe 25% of the SR scrappers I ran into were running toggle builds and not perma-elude. A sizable minority. But while we would be perfectly happy with getting rid of the crash, most of us felt the crash was entirely fair for the performance (as long as we didn't look too closely at Regen scrappers) and an entirely reasonable cost for running Elude perma. It did not make the power or the game any less friendly in any sense of the word.

Yes: Brutes, Scrappers, Tankers, and Blasters.

Only three have access to Fiery Aura as I mentioned, though: Blasters have access to Burn in Fire Manipulation.

Also, while Burn is basically the same for Scrappers, Tankers, and Brutes, its slightly different for Blasters.

(Blaster Burn has the Defiance damage buff and it also has slightly different levels of immobilize protection, although it casts the same pseudo-pet and thus has the same damage. So its not quite correct to say that Burn is the same for all *four* archetypes that have access to it).

"Rapidly" is somewhat of a relative term. Although Android sales seem to have spiked above iPhones recently, they'd need many quarters of that growth to surpass the installed base of iPhones, and iPhone sales itself seems to still be growing. I believe a recent Gardner Group report suggested that Android phones might equal or surpass iPhone installed base sometime in 2012, with both having about 70-ish million devices deployed. Symbian is still projected to hold the top spot at that time.

Two years is a long time to take writing a CoH smartphone app and a long time in general in terms of the mobile market. While Android is definitely the product to watch over the next year, before 2012 comes around AT&T's exclusivity contract with Apple will have expired, and that could change the game completely yet again. For every person I know that owns an iPhone, I know two that don't but would buy a Verizon one tomorrow.

A mobile-optimized web-based CoX application might be the best cross-platform bet. But I'm not sure if that level of web developer exists in NCSoft and available for CoX duty right now.


(Personally, I'm an iPhone user but I believe in the long run competition between Android and iPhone will ultimately be to the benefit of both customer bases, so I'm watching Android very closely. And I don't think WebOS is completely out of the game either with the HP acquisition.)

Scrapper secondary powers do not, as a rule, crit.


Altering the damage cap of the SC pseudo-pet is problematic, and prior precedent has the devs generally ignoring the damage cap issue for such powers. Furthermore, Burn is the same for all three archetypes with access to Fiery Aura, which means the Brute version (and the Tanker version) currently does too much base damage relative to its peers.

And if the Brute version's damage buff ceiling is increased to 850%, the Scrapper version would need to be increased to 500% as well. There is a lot of collateral changes that would be necessary to honor this particular "rule."


(Both the Brute and Tanker versions already taunt).

I guess that works out better for a rated T for Teen game than Pubic Rockets.

I don't consider that to be a problem in and of itself. Some things basically have their full strength all of the time, with caveats other than uptime. Dark Armor is an example. It does not have a conventional tier 9 defensive power. So whatever its full strength is, it has it all the time. And given Dark Armor's design, I don't think that is problematic, because it has other design considerations that make that reasonable.

I don't think the solution is to force people out of Granite. Rather, I think the solution is to balance the strengths and weaknesses of Granite to make running it or not running it viable alternatives. And to be honest, if Granite suppressed Rooted's regeneration, you'd probably be most of the way there. But I don't find that to be a conceptually palatable fix.

I forgot to mention this the first time around, but there is a problem with this design. You can't change the buff component to stack over time like that: Granite has to more or less reach full strength immediately (the debuff can apply over time).

The reason has to do with a lesson regarding Instant Healing. When the devs were monkeying around with IH, one of the things they did to try to enforce a cooldown on IH was to increase its recharge to a very high value: 60 seconds if I remember correctly. By doing so, the devs felt that IH would no longer be used all the time because there would be a definite period of time where it was unusable, after it was detoggled.

The problem was that very long cooldown only encouraged players to never detoggle IH. The cost for detoggling IH was that if you found you needed it again you couldn't have it. That encouraged people to say, the heck with that, I'll run it all the time and learn to live with its penalties.

A related problem occurs if Granite/Calcify requires significant ramp up time. Especially for tankers, and generally for anything likely to need the strength of this power, you're likely to need it most at the start of a fight and not the end, near where you first toggle it on. The movement penalty eliminates the option of toggling it on *in advance* of the fight because if you do you'll never actually make it to the fight. And if the movement penalties are strong enough that its basically impractical to run Granite all the time, the players will have a quandry, because the designed way for Granite to be used doesn't provide the actual protection the player requires in the right way.

I don't think ramp-up tier 9 defensive powers don't make sense in this case. They would only make sense as inverse crashes that were out of the control of the player. For example, if I was making a Blaster epic protection power, I might make something like Foce of Nature but as a toggle which ramped up. Here, the compensating control is that blasters don't have mez protection, so they could be mezzed out of the toggle. If they are, they can retoggle but the protection would slowly build up and not be "instant on." That's a form of mechanical trade off that makes sense in that situation.

In the case of Granite or Granite substitutes for a melee defensive set, I think it makes less sense.


I was thinking about alternatives to endurance-based compensators for Granite, and one of the options I've been toying with is that of movement debuffs tied to attacks. In other words, suppose that when you are in Granite every attack that hits you debuffs your movement. In the heat of the fight, you'd basically be practically locked down. But as the fight tailed off, that movement debuff would weaken, and when the fight was over the debuff would expire and you could move between spawns.

If you think about it, defense when there's nothing attacking you is never overpowered. Mitigation only matters in an actual fight. So if the debuff only exists during the fight, that serves most of the requirements of game balance.

Mechanically, I don't know if that is actually possible: it might require new tech.

That's actually the crazy way I was going to ask Castle about. You could just make the pseudo pets grant a zone-wide buff to the player. But then every single Granite Tanker in a zone would be sending out zone-wide buffs. I'm not sure if that's computationally expensive or not in the general case, especially when something like a zone event happens (it depends on how the targeting system works to locate valid targets for a power effect).

I'm also not sure if something can truly be completely undetectable and untargetable in all special circumstances, like say Hamidon raids. It might make for an interesting corner case exploit if Granite tankers could drop "distractions" in certain unusual circumstances.

There's a lot of potential catches to this kind of thing, and Granite tankers littering the landscape with invisible pets seems to be the sort of thing that would make them queasy.

The biggest problem with using pseudo pets to do this sort of thing is that, to the best of my knowledge, there is no game mechanical way to "lock" the location of the pseudo pet to be the same as your location consistently. The pet would have to be programmed to follow you. It would be amusing, but perhaps not ideal, if players occasionally got their Granite Armor caught on a tree and left it behind.

Fast motion of any kind would also make it difficult for the pet to keep perfectly in sync with you: you could dance out of the armor as well. Ironically, this works best if Granite continues to have massive movement penalties.


I thought of a potentially interesting way to do this, but its the sort of thing that Castle would kill me for saying out loud, so I'm going to ask him first before repeating it. It falls under the category of "incredibly crazy idea that might work, but the devs would almost certainly rather not do."

Just as an aside, ablative armor is on my list of top ten game mechanical additions I'd like to see added to the game.

The Curse Breaker uses the RevokePower effect. It allows you to remove a power that was granted with the GrantPower effect. GrantPower has been around since the beginning of time, but I'm not sure about RevokePower: it could have been added post-launch. But that effect can remove a temporary power you were previously granted (I'm not sure what happens if you target a power that *wasn't* granted with GrantPower: that would be an interesting experiment).

The mechanism used by the ammunition toggles for Dual Pistols is completely different. It can be used to modify a specific type of power effect in every power you possess. That's how ammunition change works in Dual Pistols.

I'll do you one better. I went and dug up the statistics I was collecting on the AE. I have data from more or less release (4/8/09) until 7/11/09 (when I got too bored to collect the data). During that three month period, the arcID increased to 259470, and the total number of actual arcs on the system rose to 49431.

Arc construction was more or less logarithmicly levelling off by that point, and a cursory examination of the current data suggests that pattern continued to the present day (this is to be expected: the initial burst of activity is constrained by a function of the total number of players, but as they fill arc slots the long term activity is constrained primarily by the turn over rate of new players). Nevertheless, even assuming about 10,000 arcs a year, 30 arcs per day (averaged for every day in the year, not just workdays), assuming just 20% of those arcs were submitted for review it would have taken nearly a year just to review the arcs submitted for review from April to mid July. The average wait time for review would have been about 140 days - four and a half months.

In fact, again assuming just a 20% submission rate, you would probably just be catching up with the backlog now, in mid May.

There are certain ironic complications to that simplified calculation. In particular, the wait times for review are so long on average that many arcs could have broken before being reviewed, which would make them trivial to review if they were still broken when the reviewer got to them. But I think its still highly unlikely a human being could have kept up this pace for this long of a period of time, even if we stopped caring if he was even reviewing accurately.

Sounds like something automation would be good at, if one wasn't religiously adamant that automation wouldn't be a good way to filter arcs.

Venture has done a bit more authoring, and far more reviewing than I have, but Venture reviews primarily for content and not exploitability or other mechanical reasons. In terms of knowledge of how the MA functions, I think its safe to say you're severely mistaken.

Venture was in I14 beta, so he knows I'm not knocking him specifically when I say that. While other people were testing the AE by trying to execute specific missions within it, I was the first one to figure out how to manipulate spawn points, the first to figure out how ambushes worked, the first to figure out how storage was calculated, the first to test the strength of critters, the first to manipulate the AI of critters in AE missions, the first to spawn many kinds of ridiculous edge-case missions. I was tinkering with every aspect of the AE from the first day of beta. Half the beta patches were probably designed to stop me from doing the goofy things I was doing in the first couple of weeks. I've actually been looking at AE mechanics and AE exploitability since *before* I14 beta, and I've never stopped looking at it. I even tracked daily usage of the AE from its release until I got frankly tired of doing it.

I'm not sure what *you* think you know about the AE, but in absolute terms its probably not very much. Certainly not enough to pass judgment on what *I* know about the AE.


On the subject of "conjuring numbers" its safe to say that your estimate of five arcs per hour has no basis in reality. Assuming the average arc has three misions, that's just four minutes per mission. There's no way you can do any meaningful checking of a mission in four minutes. What are you estimating for the time to check a custom critter, six seconds?

And if you knew anything about the AE, you'd know there's no magic about checking "the back end" like the devs have some magic perspective on the arcs that is somehow better than actually playing through the arc. "The back end" is easy to see: just look at your local mission files. That's all you'd see when you looked at the "back end." So really, you're saying that if everyone just forwarded you their mission files, you could tell, in four minutes, if any of their missions were either designed to be exploitable or had serious unintended reward imbalances.

Yeah, sure.

As long as we are taking a brief sojourn into fantasy land, lets come up with the job description for this mythological arc reviewer that NCSoft is going to hire. He or she is someone that knows how all the archetypes and powerset combinations function in CoX, knows how the custom critters work as well as the standard critters, knows how spawning works, knows what the exploitable tricks in the AE are, knows how the reward system works and what the reasonable reward rates are, and basically can spot reward problems in a mission just like that.

And this person just wants to read mission arc files all day long for NCSoft and check them for problems.


Its really easy to just throw out numbers like "five arcs an hour" when you don't have to actually figure out how to do that at that rate. And no, NCSoft is not going to pay you to figure out how, so I guess your genius will die with you.

Data point to consider: one of the examples I discussed on the forums (that predates the cottage rule post) was changing Evasion from a toggle to a click (and adding some additional features). At the time, Castle suggested to me (and I repeated on the forums) that the change from toggle to click itself was a potential violation of the power consistency rule aka the cottage rule. It didn't *disqualify* the change from being made, but it was a very strong point against such a change.

The cottage rule encompasses three things: Availability (when the power becomes available to the players aka the power's "tier"), Purpose (what the power does for the player), and Mechanics (how the power is used mechanically by the player: i.e. clicks vs toggles, AoEs vs Location AoEs, foe targeted vs self-targeted). It explicitly does not cover Strength (how large the magnitude of the power's effects are, which are always subject to balancing and tweaking), Synergy (the degree to which the power can be combined with other powers to enhance its effects), or Specific Builds (what effect that power has in a build specifically engineered to require the power to function in precisely the way it functions in any or all respects).


One more thing about the cottage rule: its not absolute. To break it, you have to demonstrate two things:

1. The change is necessary.
2. Alternatives that don't violate the rule won't work well enough.

Of course, whether a change passes both requirements is a judgment call on the part of the devs. And its not easy to pass both: out of all the times I've tried to convince Castle to break the cottage rule to implement a specific suggestion of mine, I've succeeded exactly never**.


** I've got a couple changes to happen in closed beta that would have been breaking the cottage rule, except the cottage rule doesn't kick in until the powers go live. Until then, no one has any expectation of consistency during a beta.

Dual Pistols leverages two effects, one new and one old but unknown until recently.

The old but unknown till Dual Pistols came around is that its possible to create a power effect that only affects certain other power effects by type. "Type" is something the devs can set up independently of any other part of the power, so in the Dual Pistols case they made the different damage types all different power effect "types."

The new thing is that the devs added a new kind of power modifier that can change the percentage chance for a power effect to occur. The combination of the two effects allow one power (i.e. the ammunition toggles) to affect the chance that certain damage effects will occur. Most of the time, this works by making the desired damage effects occur 100% of the time and the undesired ones occur 0% of the time, but it is also used to suppress the knockdown chances of some powers when lethal ammunition is not used.

To the best of my knowledge, it is still not possible to change the base properties of a power in any way: cast time, recharge, endurance cost, etc. Only individual power effects. And in fact Dual Pistols doesn't "change" any of the power effects per se: it "buffs and debuffs" the percentage chances, no differently than it might buff or debuff other aspects of the power.

Also: as far as I know these are relatively speaking computationally expensive effects for the game engine to perform. Its not something the devs necessarily want to do in a wide-spread manner.

This is one of those areas upon which a lot of balance discussions drive over the edge, and it tends to get glossed over. But I think its important to resolve, so here goes.

The phrase "make up for" is a bit fuzzy. There's absolutely no problem in a set that is lacking in X to try to acquire more X to compensate for the lack of X. However, there is a serious game design problem when someone that lacks X can do something that essentially negates the lack of X.

Lets put this into context. Suppose that powerset A has 30% defense, powerset B has zero defense, and powerset C has some advantage that comes with a -10% defensive penalty. And suppose further that combat jumping offered 60% defense. In such a game, two things are true:

1. Powerset A's defense is not very meaningful, because Powerset B can buy a level of defense that softcaps B in one cheap power. While its technically true that there are corner cases where Powerset A's 90% defense with CJ would be better than Powerset B's 60% defense with just CJ, its obvious that the benefit of Powerset A's defense is severely minimized. It would have a significant defensive value that B didn't, but with CJ that value is trivialized. Its just not worth very much.

2. Similarly, Powerset C's "disadvantage" is almost totally negated as well. It can buy CJ and get to 50% defense, which is also above the softcap. And while its true that 90% > 60% > 50%, clearly the defensive penalty that Powerset C received is being swamped by CJ: its just not that meaningful most of the time.

That's an example where a penalty can be basically erased by circumstances. But with Granite Armor, there is a more subtle effect at work. Consider this:

Powerset A has zero damage buff.
Powerset B has -30% damage debuff.

At level one, without enhancements, Powerset B (assuming its paired with an identical offensive set as Powerset A) deals only 70% of the damage of Powerset A. However, at level 40 with +95% damage enhancements slotted into attacks, the situation changes: Powerset A now deals 195% damage and Powerset B now deals 165% damage, which means Powerset B is dealing 165/195 = 84.6% of the damage of PowersetA. Its 30% penalty has been "diluted" to only a 15.4% penalty. And as you stack team damage buffs or invention buffs, that penalty continues to decrease. Couple that with a -65% recharge penalty that can *also* be diluted with slotting, and the total damage penalty can be very high on unslotted attacks and very low on slotted ones, and even lower in the average team with the average level of buffs.

Because the penalty can be diluted it weakens as the buffing situation improves. But since the devs have to target the penalty for a variety of situations, they have to account for things like unslotted attacks (otherwise, they could neutralize the damage of attacks a tanker added to their build before they could aggressive slot them). The issue here is *not* that -30% damage and -65% recharge can be "neutralized" by buying +30% damage and +65% recharge, but rather that those penalties weaken proportionately from their initial strength.

This is no different than the converse problem where Brutes tend to be so highly damage buffed that their own Fury tends to dilute other damage buffs like Assault or even Build Up. Proportionately, those powers numerical strength is weakened due to being in the presence of other strong buffs.

If the devs could apply a proportional debuff, something where, say, the Stone tanker would *always* deal 30% less damage and be 65% slower under *all* buffing conditions with Granite up vs Granite down, then those penalties would be more "inescapable" because they couldn't be diluted away or saturated away.


Ultimately, that is the problem with the Granite debuffs today. They simply don't do what they were intended to do, because the linear additive nature of buffs and debuffs make it impossible** to do what is intended with the current game mechanics. And this has nothing to do with "plugging holes" or "making up for weaknesses" but rather taking a penalty applied to a power and diluting that penalty until its too weak to do its intended job.



** Well, technically difficult, not impossible. I can think of a way to address the damage debuff issue that does not require new tech, and I can think of a couple of tech changes that would address the recharge issue. But I'm not sure its worth it just to tackle a single power.