Arcanaville

[ QUOTE ]
Multiplicative defense, as I follow the math, would be like a +25% Shield instead. Super Burger's damage would be reduced by 25% (so he takes 1.875 damage instead of 2.5) and Electric Mayonnaise would also be reduced by 25% (so he takes 3.75 instead of 5.0).

[/ QUOTE ]

Yes. Already linear effect stacking buffs like Commander Ketchup's damage buffs, or the regeneration of The Cheese Whiz, would be unaffected by this suggestion. They might have other issues worth investigating separately, but probably not addressable by any stacking change specifically.

[ QUOTE ]
<QR>

Arcana, have we ever gotten confirmation on how the Defense Debuff Resistance works?

Castle said that sometimes it effects the magnitude and other times the duration. For when it effects the magnitude is a straight out resist? Like say I have 10% Defense Debuff resistance currently, does that mean I ignore the first -10% Defense Debuff, so if I get hit by -15% Defense Debuff I now have -5%, and any more Defense Debuff stacked keeps going up? Or does it mean I reduce it by 10% so, I am effected by -13.5% Defense Debuff? Or does it work in the wierd manner that Mez Resist does i.e. 15/(1+0.1) = 13.63% Defense Debuff I get effected by?

Just asking because I haven't seen anything new about this, and it's not exactly easy to test I understand. Would be nice if we could get a dev explaining how it works sometime.

[/ QUOTE ]

I don't specifically remember when Castle said that, although if he did, my guess is that is because some defense debuff powers themselves are enhanceable for magnitude, and some for duration, and the resistance keys off of that. I'm just not sure which defense debuff powers are enhanceable for duration.

In any case, for the defense debuffs that are resisted by magnitude, what I've been told in the past is that they are resisted like you would think resistance works, and the way other effects are resisted. In other words, if at a particular level and with a particular set of powers activated, SR had 60% defense debuff resistance, then if it was hit with a -5% defense debuff, it would resist 60% of that, of 5 * 0.6 = 3, and feel the rest: 5 - 3 = 2. A 2% defense debuff would land (actually, I think a 5% debuff would land and the SR would feel 2% of it - the difference is that if you turned off all your toggles suddenly, I think you'd suddenly start feeling more, as some of your debuff protection toggled off, and more of the debuff "got through").

Still haven't had the time to really hard-core test this, though.

[ QUOTE ]
Again, you're misreading me. I mean equivalent performance. The ideal situation is one where there's no difference in survivability between bringing a healer, +res'er or a +def'er. Healing does NOT offer the same performance. Sometimes it's far above, and sometimes it's far below. And you know this.

For one, healing reacts poorly to differences in hp. If a healer can double the survivability (read: time to die) of a defender, they won't be doubling the survivability of a tanker. The amount of damage mitigated via healing is exactly the same, of course. But +res or +def have no such restriction.

Let's suppose this change went through, and at the same time the game were made a lot more difficult, at least in terms of damage dealt. If all you cared about was, essentially, 'time to die', and assuming you can always pick and choose. You wouldn't choose an emp at first, because they just won't add the survivability of an FFer or Sonic. Likewise, once you have a couple of those, you probably wouldn't choose another one, since an emp will do a better job of mitigating what damage does get through, and reducing your downtime. There's an obvious difference in performance between the sets.

[/ QUOTE ]

This has nothing to do with stacking. Regeneration mechanically works differently from defense, or resistance. The suggested change, for the last time, alters the way defense powers are summed, and resistance powers are summed. It does not change how a particular value of defense or resistance works, how strong it is, or how it compares to the other mitigation types. If you think regeneration has the "problem" you're describing above, its not a property of the stacking suggestion: its a property of the way those mitigation methods work, and *no* change in how they stack would alter that. Its irrelevant to the stacking issue.

And why would they make the game more difficult after this change? It makes the highest levels of performance achievable lower, and lower levels of performance sometimes higher, under heavily stacked conditions. Why would that cause the devs to make the game harder?


[ QUOTE ]
I'm sorry, but you do realize you were the first one to use impossible values, right? Here:

[ QUOTE ]
If it wasn't for the archetype caps, defense and resistance could do that: 50% defense and 100% resistance would both be perfect indestructibility. There's no level of regeneration that does that: it would take an effectively unlimited amount of regeneration to get the same approximate behavior.

[/ QUOTE ]

I was pointing out that under a very specific lens, healing can also provide functional immortality when it's high enough. In my case, I discounted all player regen. In your case, you assumed 50% def and 100% res to achieve actual immortality. Neither is possible.

[/ QUOTE ]

There's a difference. My statement was simply a simplification to make a point I could make by keeping the caps and resorting to additional but unnecessary math. My point was that defense and resistance strength rise asymptotically to infinity as they approach a limit, which is why the caps exist at all: the caps exist to prevent basically inifinite survivability. That's not true for regeneration: the regeneration caps are incredibly high: usually 2000% to 3000%. And there's no healing cap at all.

On the other hand, my problem with your comparison is not that the numbers you are using are "impossible" its that they are meaningless. Its meaningless to compare X% defense and zero regeneration to anything, not because no one has zero regeneration, but because that's not an appropriate standard for comparison. Why compare defense to zero defense: that's not the lowest that defense can go: a resting player has -10000% defense. That's actually the zero point for defense: why don't you use that point of comparison instead? You're using zero only because you think zero is always the correct starting point for a mathematical comparison, but that's not true.

When someone asks the question "how strong is X% defense" the logical question to ask in return is "relative to what?" The default answer is usually "relative to not having X% defense." Such a player has zero defense, any amount of resistance (conveninently set to zero, but it doesn't matter), and base regen.

The same thing is true when we talk about Regeneration itself: we don't say 400% regeneration is 400 points better than someone with zero regen: that's nonsensical. We say 400% regeneration is four times better than someone with 100% regeneration: the base.

Anything else lacks justifiable meaning. Its not just about calculating a blizzard of numbers. Each number has to mean something, or the calculations are worthless. Comparing X% defense and zero regen to anything is asking the question: relative to someone with no regeneration at all, how strong is X% defense? And the answer is: who cares: that's only meaningful in certain weird corner cases of comparison, and not the general case.

[ QUOTE ]
The downside being you'd have to learn Calculus to know your actual defense.

[/ QUOTE ]

Actually, no. All you'd have to do is make one tiny shift in perspective, and you'd be able to calculate, say, your net resistance under this scheme easily (I say resistance: its kinda tricky to calculate your effective defensive mitigation even now).

The shift in perspective is this: right now, we give defense and resistance powers a number. That number is a rating that says what it does. You might think that number is obvious, and not really subject to debate, but actually the number that's assigned is just slightly arbitrary, because there's actually *two* different perspectives you can have when looking at a resistance power. You could do what we are doing now, and specify how much (percentage) damage the power *blocks*, or you could specify how much damage the power *allows through*.

When we say tough has 11.25% resistance, what we are saying is that it blocks 11.25% of the damage it sees (for the damage types it works on). But we could say the opposite: we could say tough allows or admits 88.75% of the damage that it sees.

You might think that makes less sense, but actually, in many ways, it makes more sense to specify resistance powers as actually "admittance powers" and specify what they let through: for one thing, its immediately obvious how much damage you take if you have such a power: 88.75%.

Under the current stacking system, such a perspective does have a problem: its more difficult to calculate what the effect of two resistance powers is. If you have one res power that offers 10% resistance, and another that offers 20% resistance, in the current game the stacked combo is 10% + 20% = 30% resistance. Simple enough. If you specified those two by their admittance numbers, its harder: 90% admittance + 80% admittance = 70% admittance: 1 - (1 - 0.9) + (1 - 0.8). Very ugly.

But in a sense, this is the root of the stacking problem. They *shouldn't* have been looking at damage mitigated, but rather damage taken. That's the perspective that survival is based on. Damage mitigated isn't really important to game balance.

If they *had* looked at resistance powers in terms of their damage admittance, there would have been another very logical way to stack them. A 90% admittance power plus an 80% admittance power would stack multiplicatively: 0.9 * 0.8 = 0.72. Why is this obvious? Because if one power causes you to take 90% damage, and another power causes you to take 80% of what's left, then what you are taking is 80% of 90%, or 0.9 * 0.8.

The important thing is that mathematically, this stacking mechanics *preserves* the definition of "80% admittance" in the equation: it actually uses it to derive the stacking equation. 80% admittance means you take 80% of the original damage, period. It will always mean that, even if there is pre-existing resistance.

The current mechanic doesn't do that. When you just add up 10% resistance and 20% resistance, the 20% resistance power suddenly fails to have meaning. It *used* to mean "you take 20% less damage" but that is not how its being used in the stacking equation: that "property" of resistance is being broken. In effect, the current stacking equation is an arithmetical error: it *sounds* like it makes sense, but at a fundamental mathematical level, its wrong.


So: switch from telling players what the power mitigates, to what the power allows to get through, and people start multiplying instead of adding. For resistance, its that simple.

Unfortunately, its not going to quite *that* simple for defense, because unlike resistance, defense has a "sliding scale" of mitigation intrinsic in the fact that tohit buffs can change the mitigative/admittive meaning of defense numbers. But then again, how many people can actually correctly calculate their defensive mitigation currently? They can certainly add up their defense numbers, but that sum alone doesn't necessarily mean anything without properly taking into account tohit buffs/debuffs and accuracy.

But in the absence of all such factors, defense would follow a similar rule as resistance: you'd calculate admittance based on, say 50% tohit, then multiply them up, then convert back into defense. Or not: if you know what percentage of attacks are going to hit you, do you really even need to convert that back into defense numbers, since that's actually what most people want to know anyway?

[ QUOTE ]
[ QUOTE ]

What I really hope is the MUO guys figured this all out before implementing their engine.

You guys know who you are: you *really* do not want me finding these same problems in the MUO engine. You really, really don't.

[/ QUOTE ]

Don't GOOOOooooo! *hangs on to Arcana's leg*

[/ QUOTE ]

MUO is probably at least two years away from launch, and I would bet closer to three. I don't think I'm going anywhere anytime soon either way.

I know next to nothing about MUO really, but I do know one or two little things. And I would love to see the design doc for MUO right now: just imagine all the problems in CoH that could have been trivially solved at this early stage of development that took years - and a lot of wasted work - to correct, if at all.

Think about the I7 defense scaler: it took a lot of time and effort to eventually get in, and before it was implemented, the devs spent even more time trying out other tweaks in the hopes it wouldn't be necessary. There was a moment in CoH's design where the entire problem could have been solved in ten seconds with a number two eraser (or if you prefer, a dozen keystrokes in Excel).

How many such issues exist in MUO right now (perhaps only on paper at the moment), that just a little numerical or mathematical analysis could preventatively solve, but will end up costing the players and the devs serious time to address after the system sees first light, I wonder.

[ QUOTE ]
Heres two simple questions regarding this idea.

1) How hard would it be on players to adapt to?
I'm honestly done revamping characters to deal with game altering changes. After 3 years, I want a stable game base, dagnabit!

2) How many fun additions to the game would be pushed back, yet again to tweak the mathmatical model?
We've allready spent 2+ issues making significant changes to the system that have helped contribute to pushing back fun things like new powersets or ATs, do we really need the Devs to spend more time nit-picking the system?

I'd rather have more toys, than more 'balance', I can ignore the complaining more easily than I can ignore the stagnation.

[/ QUOTE ]

I would argue that *if* this was ever done, it would open the door to a wealth of possibilities for adding new toys. Consider all the things they *can't* do now, because of broken stacking:

1. They can't make power pool protection powers simultaneously useful to squishies, and non-squishies.

2. They can't make well-balanced high-end teamed content, that doesn't involve totally negating highly stacked protections, because highly stacked protections are extreme enough to trivialize high end content.

3. They can't make something *somewhat* more dangerous to certain things: they generally make things *extremely* dangerous or not at all, because of the large range of possibilities that the game has to deal with (i.e. quartz eminators' +100% tohit buff, designed to hurt defense no matter how much of it you decided to bring along today).

4. They can't fix a lot of imbalances in PvP, especially but not exclusively around defense.


Think this is just a problem of the past? Nope:

5. They can't make single set IO buffs very large, because they might stack too high. So they have to make them very tiny, so that those who *do* stack a lot of them don't get game-breaking buffs.

6. They can't allow Defense sets to get very much more defense out of Inventions, even though they make it much easier for resistance sets to get more resistance, and regeneration/healing sets to get more healing.

7. They have to be very careful about handing out mez resistance.

8. They consider debuffing sets potentially dangerous to game balance to implement.


All specifically, *directly* attributable to the stacking problems in the game engine. How many things have we lost in this game due to the stacking problems? And what will the stacking problems take away from us tomorrow, that we might never even know we lost?

[ QUOTE ]
In the suggested fix, res and def are made almost equivalent

[/ QUOTE ]

I don't think you understand the suggestion. The suggestion doesn't change the relationship between defense, resistance, and regeneration at all. It only changes how they stack. Whatever the relationship is between something with 30% defense, 50% resistance, and 400% regeneration today is, under the suggested stacking change, that relationship would be exactly, precisely the same. It has to be: the change doesn't change the definition of "30% defense" or "50% resistance" or "400% regeneration:" those are all exactly, precisely as strong today as under the proposed change.

What changes is what happens when one thing buffs another thing: when 30% defense meets 20% more defense, say. It only changes stacking, not how defense or resistance or regeneration work.


[ QUOTE ]
If you don't want to cause problems you should make all types of mitigation equal

[/ QUOTE ]

No. This always comes up eventually when talking about balance. The notion is, if you want balance, everything has to behave identically, or identically except for minor exceptions. That's pointless: it would be simpler to get rid of two and only use one. The whole point of having three different damage mitigation mechanisms is to have three different damage mitigation mechanisms. Its up to designers to use the proper mathematical techniques to ensure that sets constructed out of heterogenous components are nevertheless balanced according to a particular set of average metrics. It trivializes the design to make the different components transparently identical in performance.

As much as I've worked for balance in the game, I would rather have totally dissimilar and unbalanced mechanics than totally balanced but homogenous mechanics. Because that's pointless from a design perspective.


One additional subtle point:

[ QUOTE ]
You're overlooking something you (and everyone, really) already know. Heals and regen have an immortality line. If players were unable to regen in combat, unless you were achieving 100% resistance or 50% defense, you'd eventually die, no matter what sort of damage you were facing. Not so with regen/heals. (Call it the low-downtime factor.)

[/ QUOTE ]

We compare what different damage mitigation powers do relative to a standard: the standard being something with no protective powers at all. Relative to this standard, we can say that X% defense mitigates a certain percentage of incoming attacks relative to having no defense, or Y% resistance mitgates a certain amount of incoming damage.

The base regeneration of such the standard isn't zero: its base regen: 100%/240 seconds. When you say something with no regen will die eventually even at high levels of defense, that's irrelevant to a discussion of the strength of defense, resistance, and regeneration. We compare +Def to something with zero defense, but we don't comare +Regen to something with zero regen, we compare to something with base regen. Regeneration powers amplify that base regen, they don't add regeneration to something without it.

Comparing X% defense and zero regeneration to something with Y% regeneration is subtly flawed: its using the wrong standard for comparison. In effect, not taking into account base regeneration is like not taking into account the base tohit of the attackers when looking at defense.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Flight poses are an example. No matter how ugly the implementation might be, as long as its not an actual embarassment to the company, I'd say add them, and let the players who want to use them use them, even if they are not perfect.

[/ QUOTE ]

Ironic example. The problem with flight poses was that people who used Fly had no choice but to use them, whether they liked them or not. You were automatically put into one when you started flying. That's why they took them out.

[/ QUOTE ]

I'm pretty sure he's referring to the new optional poses, which are kinda shoehorned into the game in an ugly fashion but at least they are here.

It was clear form the closed beta that those poses were going to be shelved forever until peeps made a big fuss about really really wanting them, so as a consolation the emote-fly option was invented (which works really well if you set up a decent keybind).

[/ QUOTE ]

Actually, I'm sort of referring to both situations, in the sense that I believe that if they have a fundamental flaw (the original random non-optional ones), they shouldn't be added, but if they theoretically work but are just ugly (in terms of implementation, not visual appearance) then its fine to add them for those willing to take the trouble to figure out how to make them work. Because the existence of the current flight poses doesn't preclude making better ones later on (also, I suggested the emote-fly thing pretty soon after the initial versions were released onto test for precisely that reason: if it couldn't be perfected, then a reasonable compromise solution was to hack in an emote-like control to give players access to them).

But not all components of the game are like that. For example, there are game mechanical elements (like the tohit algorithm, or the defensive stacking mechanics) that are unambiguously suboptimal, create a very large number of problems that have to be worked around with very ugly hacks, but cannot be changed because its deemed to risky to tamper with. If you know in advance you're about to add something to the game that you're going to be too chicken to mess with later on, then it better be perfect, or as close to perfect as is humanly possible, because if you're being honest with yourself, you know you're never going to change it ever again, no matter how much problems it causes.

[ QUOTE ]
Ah well. I'm happy to know that the devs understand what we've been saying and would do it if it was possible. It validates my opinion that Castle really knows what he's doing.

[/ QUOTE ]

What I really hope is the MUO guys figured this all out before implementing their engine.

You guys know who you are: you *really* do not want me finding these same problems in the MUO engine. You really, really don't.

[ QUOTE ]
Keep in mind that it's not just +res and +def. Heals would have to be treated in the same fashion, for this to be fair. Heals would always heal x% of the damage you've taken, otherwise you will have a problem.

[/ QUOTE ]

No, to be fair a heal would have to heal the same amount, regardless of how many other heals had landed prior. Luckily, it does so now. There is no problem.

The issue of fairness and balance comes from attempting to normalize the incremental benefits of a particular buff, relative to any pre-existing conditions. This is for the most part already true with heals. Its not true for things like defense and resistance buffs.

Again: under the right circumstances (meaning, for different numbers the numerical change will be different, but this behavior does not rely on any special circumstances in general) an FF defender that can double the survivability of a group (they take half the damage, or they can face off against twice as many foes simultaneously), when they join another FF defender that can also double the survivability of a group, doesn't end up with a group with four times the survivability. Its often ten times the survivability. That creates all sorts of problems when it comes to balancing the game, because you have to balance powers for what you want them to do individually, and what they can do when they stack up, and if the powers can run away like that, you have to constrain them. That's why combat jump is 1.875% defense. The devs don't really think blasters deserve only 1.875% defense: they are worried about what SR scrappers, Ice tankers, and FF defenders will be able to do with 1.875% defense, which is potentially a lot.

When a defender with heals joins a team and can double their survivability, or double the number of things they can take on, and they join another defender with heals that can do the same thing, you generally *do* end up with a team that can take on about four times the stuff [Correction: actually its generally three times the stuff: healing doesn't even scale proportionately like a proposed multiplicative stacking system would do - its even *lower* than that, much less what happens now]. Healing and regeneration do not scale in the same way as defense and resistance do, because of the way defense and resistance are implemented in the game. Heals and regeneration don't suddenly run into a mathematical cap, where you become indestructible. If it wasn't for the archetype caps, defense and resistance could do that: 50% defense and 100% resistance would both be perfect indestructibility. There's no level of regeneration that does that: it would take an effectively unlimited amount of regeneration to get the same approximate behavior.

Its this race to zero that defense and resistance do, and regen does not, that makes them fundamentally different mechanically. And that's why the devs have no problem giving players a power pool that gives +40% regen, but anything more than about +5% defense in the pools is too scary to contemplate (and most are below 3%). That 40% regen doesn't somehow amplify itself and become massive when "stacked" on top of regeneration scrappers: they actually hardly notice. But stack 5% defense on top of an SR scrapper with 30% defense, and they notice.


There is no reason to force regeneration to work in the manner you describe, to address the issue at hand with regard to stacking, because regeneration doesn't have the issue in question. One has nothing to do with the other.

[ QUOTE ]
Since we appear to agree that there are different ways to be "right", I am suggesting in context with CoX that there is more to be gained by compromise for the "stapler" consumer - than doing nothing at all.

[/ QUOTE ]

You can compromise a lot in a design or implementation, but the one thing you should never do is compromise the future. If you can add a feature to the game that isn't quite perfect, but adding it does not in any way impact the ability to improve it later, or add things around it, then you might as well do it. Flight poses are an example. No matter how ugly the implementation might be, as long as its not an actual embarassment to the company, I'd say add them, and let the players who want to use them use them, even if they are not perfect.

But core infrastructure normally doesn't obey those kinds of rules. Adding things to the server environment to allow for cross-server activity is the sort of thing that when its done wrong, is like turning to the Dark Side: forever will it dominate your destiny.

[ QUOTE ]
And I'm not singling out buff/debuff ALLY sets. I think the same thing should happen to Scrappers/Tankers/Brute and Stalker defense sets. In this way you could encourage more diversity among builds.

For example. When the game launched some Invul players used to take only the passives because they were good. But since buffs are additive, stacking them with TI and Unyielding was overpowered. But if you had multiplicative stacking, passives could be good, still less than toggles, but reasonable and people could explore different ways of getting the character they want.

[/ QUOTE ]

My original motivation for the suggestion in the first place.

[ QUOTE ]
There's quite a colorful line you draw between engineering optional entertainment and engineering critical safety items affecting real life.

[/ QUOTE ]

No difference to me. As far as I'm concerned, if you don't believe its your job to do everything the correct way, every time, you shouldn't be making french fries, much less anything more challenging.

[ QUOTE ]
Still, considering how many safe and viable ways there are to engineer and build a bridge, I hope you aren't suggesting your way is always "Right" and trumps your competitor's or that because your boss or his boss bid lower for the contract, you are somehow less attentive to getting it "Right"? Or perhaps your own colleagues are clearly inept - suggesting a line of code would better fit here than there? "My ("right") way or the highway" is an arrogant luxury that few can afford because it undermines more than it gains in efficiency in my experience.

[/ QUOTE ]

There's lots of ways to be right, and even more ways to be wrong. When there exists someone who can do it right, and they cost significantly less than me, I refer people to them. I have a very long-standing reputation professionally for doing that: I have no interest in making people pay more for something just so they can get it from me.

When they can't do it correctly, and I know it, and they attempt to convince people otherwise, I have another very long-standing professional reputation as well, and it involves giving people enough rope to hang themselves with, and then helping them with the knot.

Right or not at all is not an "arrogant luxury" except to the incompetant. I've said that publicly before an audience of my professional peers, more than once. I'll keep saying it until I die or retire.

And I'm almost never the low bidder. I don't work for a stapler manufacturing company.

[ QUOTE ]
Now, there are specific sets (and mainly specific combinations thereof), that ARE capable of performing amazing feats singlehandedly, and those may need to be looked at. But the notion that Trick Arrow deserves a nerf because Instant Healing isn't a toggle anymore is laughably petty.

[/ QUOTE ]

Its worth noting that the mechanical change I've been after since 2004, and the change Castle is basically implying with his example, does not as a consequence of its design involve nerfing anyone. It actually involves *buffing* non-stacked performance in the general case, leaving internally stacked performance identical to the current case, and only affecting multiply stacked buffs in a manner that only returns buffs that can stack exponentially to normal incremental stacking levels.

It doesn't place a hard cap on buffs. It doesn't weaken any particular power specifically. It opens the door to buffing powers specifically (in fact, it practically *mandates* buffing certain powers explicitly). All it does is alter the stacking equations for like buffs stacked on top of like buffs, for the exponential stacking buffs (really, inverse hyperbolic ones, but exponential is a more familiar term).

Trick arrow is specifically one of the sets that stands to benefit from multiplicative stacking. It would allow for a potentially stronger flash arrow for one thing.

[ QUOTE ]
I think your argument is based on a fallacy, that an equivalent percentage of buffing brings the same effect to the team no matter what the numbers. While in your example from a previous post does indicate the percentage effectiveness is the same, the practical effectiveness is not.

It is like saying that I have 2 coupons. The first gives me 50% off. So, I have an item that is $100. So, I save $50. Now, I get a second coupon. It gives me 50% off of the reduced price. I save $25 in this case. So, the second coupon, while it saves me some money, the effectiveness of the coupon is significantly less, even though both coupons save me 50%. There is a point when the effectiveness of the 50% becomes marginal.

[/ QUOTE ]

There is never a point where the effectiveness of the 50% becomes marginal, at least in the direct sense. It only looks that way on paper. It always means the same thing: you can take on twice as many things after you get the buff than before. The fact that the numbers are getting smaller is meaningless, because concurrent with that the amount you *need* to get to stay alive in the face of higher levels of damage is also getting smaller.


Put it this way: suppose you get paid $1000 per month. If I cut the intervals between paychecks in half, you now get a paycheck every 0.5 months, and you now make twice as much money ($24k/year). If I cut the interval in half again, you now get paid every 0.25 months, and make twice as much money again ($48k/year). If I cut the interval in half again, you now get paid every 0.125 months, and make twice as much money again ($96/year).

Now: do you consider this a diminishing return (1, 0.5, 0.25, 0.125...) or an accelerating return (12,24,48,96)? I'm sure there are people who will tell you that it all depends on your perspective, and mathematically both perspectives are equally valid, but would you agree in this case? Would you say that this was a diminishing return situation, and at a certain point it just doesn't matter anymore, and you wouldn't care either way? Heck, what's 0.0625 months: its only 1.9 days. How could shaving 1.9 days off of a pay period mean anything really, right?


The problem in your example above is that you're assuming you only ever have to buy one thing with that coupon. If so, yes, eventually the savings becomes marginal. That's like saying if you're going to fight exactly one minion, and then log out, at some point more defense becomes marginal. That's true: eventually the damage being taken is so small, that reducing it further is not especially helpful or noticable.

But that's not the case. Being able to take more damage means you can theoretically fight more things, and potentially faster without dying. In effect, instead of buying one thing with that coupon, and saving less and less and less, you're actually allowed to buy as many as you want, and those coupons are allowing you to buy more: you might have been able to buy just 2 with the money in your wallet, but after the first coupon you can buy 4, and after the second coupon, you can buy 8, and after the third coupon, you can buy 16. At no time does the next coupon become irrelevant: each time, you are doubling the amount of things you can buy.

Of course, at *some* point your defense can outpace your offensive ability to actually defeat things at a reasonable pace. But that is a separate issue of internal team balance, not a direct issue of effects stacking.

[ QUOTE ]
I'm partially making a point to the OP I replied to but I also want to know what it is about your suggestion that encourages teaming?

[/ QUOTE ]

Buffs would be much more difficult to stack to the ceilings: you couldn't have the first buffer do something, the second one drive you to the rails, and all others be worthless. In a multiplicative stacking environment, each buffer would have the same effect as the first, no more, no less (until *eventually* you reached the absolute caps for certain attributes).

Take force fields. *One* FF defender with maneuvers can make all others redundant, because they can stack bubbles + dispersion + maneuvers and hit the tohit floor. Two can do so without the need to even stay under the dispersion bubble at all. With multiplicative stacking, each FF defender would have the same effect as the previous one, and three FF defenders could still be made useful. Two would have a lesser effect than before, but neither one is being nerfed: each is having the same incremental effect. Currently, the second one is much stronger than the first one, because of stacking, and there's no reason he needs to be.

This means its much more difficult for one buffer to make another one redundant, but the powers themselves aren't being made individually weaker: no FF defender ends up weaker than they are now.

Actually, they could be made stronger: FF defenders can directly benefit from multiplicative stacking even when they are the only buffer around, because right now, the small bubbles and the dispersion bubble are balanced around the fact that they can stack together. If they made either stronger than they are now, they could combine to reach elude levels of defense. As it is, they can already do so with just one power pool power addition (maneuvers).

But for team mates within the small bubbles, and operating outside the dispersion bubble, that means the small bubbles are weaker than they could be. And for the FF defender, the dispersion bubble (their only personal defense bubble outside PFF) is also weaker than it could be, because of the danger of (linear) stacking.

With multiplicative stacking, the small bubbles could be increased, the dispersion bubble could be increased, and the stacked total would still have the same effective strength as they do now. Players (like scrappers) that operate outside the dispersion bubble benefit, the FF defender benefits, and other FF defenders benefit because they are still useful additions to teams that already have an FF defender. Nobody loses.

And there is an additional winner: the devs balance a lot of content around the presumption that some buffs can just be theoretically stacked sky-high. Quartz eminators have +100% tohit buff, (presumably) to cut through stacked bubbles. Lord Recluse (in the STF) has a tower that seems to buff him by at least +50% if not +100% tohit also, again to cut through highly stacked defenses. Hamidon is just plain untyped, to make absolutely sure he can cut through all conceivable defenses. With a more rational stacking mechanism for defense (and resistance which are the two main buffs that have the largest stacking problems), PvE content can be softened to not have to obliterate such defenses. Which means players that don't have such ridiculous levels of defense don't have to be massacred just to get at the other players that do.

There are really no losers in such a stacking scheme. Everyone's powers work the way you expect, which is to say they work the same way, regardless of the presence of other powers. "Stacking" in the sense a lot of people currently mean stacking, would go away, replaced by powers that simply always worked in a simple way. I could give hover 10% defense, and tell someone "if you take hover, you'll take 20% less hits." And if they say "what if I have other defenses?" my answer would be "doesn't matter: whatever you look like now, if you take hover, you'll get hit 20% less often than you do now, period."

Which makes describing powers, in-game and in printed documentation, much easier. Yet another win.


I wouldn't say multiplicative stacking would "encourage teaming." What I would say is that it takes away a lot of penalties to certain teaming combinations. And it has a ton of other benefits besides. And it hurts basically no one. And it makes the powers easier to describe. Its just better.

[ QUOTE ]
Your conclusion is based on facetious facts, supposition and your own opinion. There is no test environment for you to gather evidence to support your claims.

[/ QUOTE ]

Mine's based on math, and I don't really need a test server to test mine, because there's no effect that multiplicative stacking systems grant that isn't easily testable on the current live servers.


[ QUOTE ]
My opinion is that this sort of change would marginalize many sets and would be a general blow to teaming.

[/ QUOTE ]

I'm speaking for my own suggestion, and not any of the many others that have been proposed in the past or in this thread: in mine, this is logically and mathematically impossible. Not just improbable: literally impossible, because of its construction. Every buffing set would have the same effect on something that wasn't pre-buffed, as it would have on something that was already buffed. There is no player-centric or team-centric downside to such a behavior that can possibly exist, unless you believe the entire purpose to your existence is to overstack with a twin.

[ QUOTE ]
Err, wouldn't such a system essentially be going back to the Pre-ED "no one really needs buffs" situation that ED and the GDN were partially intended to fix?

[/ QUOTE ]

It depends on the buffs. If we are talking about things like defense and resistance, then actually the problem with those mechanics is that stacking currently has accelerating effects. Actually, each buff is stronger than the last one: when you stack 25% defense on top of 15% defense, Castle knows that the combination of the two is actually much stronger than the individual components of both. He's fine (in this example) giving two different powers significant numerical levels of defense, because individually they aren't game-breaking. What's game-breaking is when you stack them, and go from one power that cuts damage in half, plus another power that cuts damage by about a third, combining to cut damage by 80%. That's out of whack, and why defense (and resistance) stacking is problematic. You can't make individual powers too big, because they can stack to extremes.

Actually, a multiplicative stacking system - something that *appears* to have diminishing returns *numerically* - actually makes each buff worth exactly the same as the previous one. Its numerically diminishing, but in terms of value, its actually not.

Imagine a hypothetical defender with a 25% defense buff power. Buffing another player with no defense cuts incoming damage in half - critters with base 50% tohit used to hit 50% of the time, and now they hit 25% of the time. Damage cut in half.

The next such defender joins the team, and buffs everyone. Today, everyone would have 50% defense - effectively perma-eluded. The first defender cut incoming damage in half, but the second one cut the remaining incoming damage by 80%- from 25 hits out of 100 to 5. He was worth more than the first one.

In a multiplicatively stacked system, the second one would be worth the same amount as the first. The first one would cut incoming damage in half. The second one would do exactly the same thing. To do that, the second +25% defense, when stacked on top of the first +25% defense, would have to equal a net defense of +37.5% defense. There's math that makes it all work out, but the important thing is not the math: its the effect: everyone's defensive buffs are worth the same amount, to everyone, on every team, regardless of any pre-existing buffs.

And that means defense buffs and defense powers can be more easily balanced, and more generously given out, especially in power pools.

This is specifically true for defense buffs and resistance buffs (and for that matter, critter defense debuffs and resistance debuffs). For other things, like +damage and +regen, they don't work the same way, and probably are in less need of a "diminishing returns" system.

[ QUOTE ]
[ QUOTE ]
I like being quoted by Rednames, but not like this...

[/ QUOTE ]

You do of course realize, if anything ever happens to buffs you will be blamed.



[/ QUOTE ]

You could always blame me instead: I've bugged Castle about this for so long, at this point I just cut and paste my suggestion into a PM, he cuts and pastes his PM that says its too much work, and that basically covers it for another couple of months.

[ QUOTE ]
Sure! Then Weave could give 15% Defense, and Combat Jumping could give 25% Defense (or whatever) and combined, they'd give 30% instead of 40%.

[/ QUOTE ]

32.5%, actually. And you know this would be my first order of business as supreme overlord of the CoH codebase.


[ QUOTE ]
if we were ever to implement a system such as this, which I doubt.

[/ QUOTE ]

That's only because I haven't properly tuned my hyperwave mind control antenna to the proper frequency yet. But poz can't hide in that underground bunker forever: sooner or later he's gotta come up for pizza.

[ QUOTE ]
not only that but often times trying to do a compromise, creates even more work down the road. what it takes to make it work now + perfect it later in most cases is far more time consuming than just doing it right later on.

[/ QUOTE ]

Not only does it usually take longer to fix something than make it from scratch, the effort it will take to fix it can grow over time. That's because the rest of the system will evolve around the broken thing, incorporating its broken-ness as a consequence.

Consider a very trivially simple example: what BaB said about Hamidon. Players were not intended to hold Hamidon to prevent respawns: they thought they gave him enough magnitude to prevent that, but players just assumed the mag was a challenge to overpower. By the time they had the technology to prevent Hamidon from being held and prevented from generating mito respawns, they had tweaked so many other things about Hamidon to work around that problem (and others) that the actual fix itself wouldn't work anymore anyway, thus forcing a complete redesign from scratch.

They can do that with Hamidon: they can make a different one on test. They cannot necessary make a bad cross server infrastructure, then go back later and make another one while keeping the live servers functioning correctly. And the act of implementing a bad one can force them to make other bad things happen in the live server code just to make the cross server infrastructure work correctly, preventing a better solution from ever actually working properly.

[ QUOTE ]
What is it about engineers and programmers having a very hard time working compromise?

[/ QUOTE ]

Right, Or Not At All, is a general rule of mine professionally also, with the exception of handling emergencies (the rule there is Right And Now: Or I Kill You).

I imagine people would prefer if the people who designed their bridges, elevators, aircraft, life support systems, anti-lock brakes, and other optional areas of everyday life tended to think similarly.

Moreover, programmers and engineers know that bad designs tend to get institutionalized. Making it bad doesn't mean you'll be able to improve it later on. Making it bad often means its going to be bad forever, because of the "practical realities" of the system once its in use.

[ QUOTE ]
[ QUOTE ]
Would it be fairly obvious to say that it ill be levels 50-55?

[/ QUOTE ]

I'll go ahead and squash that one now. Level cap is NOT going up in I10!

[/ QUOTE ]
You heard it here first: Castle says level cap increase delayed until I11.

[ QUOTE ]
any statements on the new IO enh/sets regarding defence?

[/ QUOTE ]

Yeah, right-click on everyone before dueling them in the arena: if you see "Enhanced Accuracy" like eleven times in their set bonuses, don't.

There is a lot more to-hit than defense in the sets. Castle just cut them all in half, and there's *still* massively more tohit than defense in the sets: that's how large the discrepancy is.

The best buff for defense powersets in the IO sets isn't generally more defense: its more speed, if you can make use of more speed (which mostly requires building around speed).

It is *possible* to get more defense, but for SR, you need positional defense, and that requires either five or six of a set to get, for the sets that have it, which will make positional defense very expensive to acquire.

One +Tohit IO (at about +6% tohit) almost knocks out the positional defense achievable with about 30 set IOs (out of which you can probably get about +7.5% defense to melee/ranged/AoE).


In PvE, things are somewhat better, mostly because the critters don't get to use inventions, so any benefit you get is a benefit over the critters - even tohit buffs.

[ QUOTE ]
Is accuracy bonuses still supposed to kill defense? I have been doing quite alot of Devouring Earth missions lately and if a Guardian gets to drop that quartz pet of his everything, and I do mean everything, hits me without problem. Same thing with Nemesis if I kill the lieu.

[/ QUOTE ]

Both of those are tohit bonuses, not accuracy, and yes, they are not healthy for defense. I don't know if Nemesis Vengeance is identical to player Vengeance, but I can tell you that the DE quartz eminator is +100% tohit. Basically, if you are not under Elude or three simultaneous stacked defender bubbles, you might as well not have defense at all verses any DE that is standing in its buff aura. If you are a defense set, you need to kill those (or the LTs that spawn them) very fast.