Arcanaville

[ QUOTE ]
Captain_Napalm:

I think the exponential nature of the equation is going to make that difficult. The problem is that if you add an integer on the "low end" (the 50% to 100% health zone, for example), it's going to inflate the "high end" (the 1% to 10% health zone), too, unless you overly complicate the equation with If/Then statements, though then you'd probably wind up with a "chunky" equation at the if/then statement point(s).

Knowing the equations themselves is good, but it still doesn't really resolve what we need to know. I know most of us say that Defiance "doesn't kick in until it's too late", that it "doesn't matter until we're almost dead". The real question is, however, what kind of average ToHit and Damage Buffs do Blasters, as an aggregate, get from Defiance in-game.

Unfortunately, we're not privy to that data, nor can we be. It's possible that, on average, our ToHit and Damage Buffs exceed Scrappers' damage from Criticals and the +12.5% more base damage they do. It's possible that, on average, ours do less. It's possible that the average completely makes up the difference, but not much more.

The thing is it doesn't take very many instances at the +350% Damage range to skew the average considerably upwards.

[/ QUOTE ]

Which is specifically the problem with data mining. If the penalty for reaching +350% is having such low health that, win or lose, you're stuck resting before the next fight, then that bonus isn't just happening for free, but that +350% shot does, in fact, skew the average upward by a lot. The "average bonus" isn't a direct measure of the value of defiance: the more proper measure would be "increase in kill-speed" because that compares more closely to criticals. Unfortunately, that is extremely difficult to data mine effectively. The notion that data mining is always better than numerical analysis is not in actual fact true. Its only true when data mining produces a better signal than numerical analysis does. That doesn't happen consistently.

In fact, in this particular case, while data mining can *inform* analysis, I would consider it an abuse of data mining if it was used to determine the *actual* performance of defiance. It just doesn't work that way.


I can data mine the toss of a pair of dice, or I can calculate it. The calculations are better if the dice are fair. And this principle doesn't require probabilistic dependancies to hold for other more complex environments to run into similar situations.

[ QUOTE ]
I was reading the above quote and I had a thought: Is it possible to overcome the "sliver of life" by slotting a fourth Damage SO in all your attacks?

[/ QUOTE ]

I don't think so. That fourth damage SO equates to about 5% base damage. At level 50, we're talking about total focus doing an additional 10 points of damage. The "sliver" is ordinarily not *that* much of a sliver. But its difficult to be sure: even assault is only roughly +10% base damage, only twice the strength of that fourth damage SO.

[ QUOTE ]
Also, I was trying to play with the equation a bit to see if I could do the above, but it seems the skill involved in making adjusting the equation to do what I described above is over my head. I was trying to adjust it so that a new equation would yield the same values at 9% and 90% health as the one you posted, but would ramp up sooner in the middle. Can you do that? Just to see how different it would be?

[/ QUOTE ]

If I had to suggest an alternative on the spot, something like this might make more sense:

25% * 2 ^ [(54 - HealthPercentage)/12.5]

(Current: 25% * 2 ^ [(45 - HealthPercentage)/10])

Comparing the current formula to this one, we get this:

Code:[/color]

---

Health	current	alternate

100	0.55%	1.95%

90	1.10%	3.40%

80	2.21%	5.91%

70	4.42%	10.29%

60	8.84%	17.92%

50	17.68%	31.21%

40	35.36%	54.34%

30	70.71%	94.61%

20	141.42%	164.72%

10	282.84%	286.79%

9	303.14%	303.14%

5	400.00%	378.42%

1	527.80%	472.40%

---

The noteworthy points are that 9% health still damage caps a 3-slot (SO) blaster in both equations (the alternate is specifically designed to pivot around that point, but the boost is a little higher at lower health levels). Also, at 60% health, the boost in the alternate is approximately equal to scrapper criticals (at SO slotting levels), because an 18% boost to base damage is about a 9% boost to total damage when you're 3-slotted for damage, which has a certain logical symmetry to me: its the point where SR passive resistances first kick in (the equivalent point in the current defiance formula is about 50% health).

It also ramps up a bit faster, but doesn't get quite as high as fast at the top end. But the top end doesn't matter quite as much to me, since its above the cap for most blasters above level 22, and for blasters under 22 even 300% damage boost is likely to be a massive boost relative to the targets being faced (the current defiance hits the absolute cap for blasters at 5% health: the alternative is only 12% short of that at 5% health).

This assumes you want to keep the basic structure of how defiance works intact, and just are looking to fiddle with the tunable parameters of the current defiance formula. But it does soften the low end, without making it dramatically overpowering it at the high end.

[ QUOTE ]
This issue about the "sliver of health", however, may actually be based on the fact that Blaster damage, at 100%, is considered to be the standard. Since foes are balanced to that standard, it would hold to reason that Blasters are balanced to JUST BARELY kill them. Adding an additional 5-10% on top of that may turn the tide on leaving those "slivers of health". Again, Scrappers' 12.5% bonus may also help them avoid that problem.

[/ QUOTE ]

That's possible, but there is another possibility: the possibility is that there is a subtle selection bias going on that creates that possibility, especially for blasters. Actually, two of them.

To explain the first one, lets imagine that you could keep track of all the fights you experienced in a particular mission. For every target you defeated, there is a certain amount of damage you delivered, X, and that is bound to be at least a little larger than the amount of damage necessary to defeat that target (its health). For each target, there is a certain amount of "overkill" that is Health - X.

Make a list of all the overkill numbers. Now, imagine reducing your damage by 10%, say. Some of those kills would have still been kills anyway, because your overkill is more than 10%. But some of them would no longer be kills: the reduction is larger than the overkill, and what used to be a kill, now ends up short.

It will be short on average by a small amount, much less than 10% of the target's health (its complicated to say precisely how much, though). Lets say 5%. That means a damage reduction actually creates the potential for a lot of targets with just a "sliver" of health left. The question is whether this sort of thing actually happens. And it does.

Many players instinctively or deliberately fight the most difficult fights they can get away with, within their comfort level. They do that primarily using the difficulty slider. And the slider in effect reduces player damage in steps dictated by the purple patch, but in the early stages by roughly 10% increments. In effect, using the difficulty slider to "find" the fights you are most comfortable with can actually induce "small sliver" situations, because those will be the borderline cases where, because you are fighting the strongest things you are capable of, sometimes you'll fall just a tiny bit short.

The second selection bias possibility has to do with how blasters choose their attacks. Its often the case that blasters will instinctively or deliberately pick the heaviest attacks first, in order to take out an initial target first. This reduces the damage they experience during the fight significantly lower. But in using those high damage attacks first, because they are slow to recharge the rest of the fight tends to be dominated by the lower damage attacks. And that means that its much more likely for a target to have just a sliver of health left, as its pelted by lower damage attacks. In fact, critter resistances can increase the perceived instance of small slivers of health remaining, because the average amount of health left on a target before the last attack finishes that critter off is comparable to one half of the damage of the average attack that is inflicted on the target, and that number drops with increasing critter resistances.

Basically, if you are only taking small amounts away from a pile, its much more likely that the last bit of the pile left will be small before you take the last bit away. If you're taking the pile away in much larger chunks, its much more likely that there will be a lot left in the pile before you take that last chunk away.

There are other possibilities as well, some that can be influenced directly or indirectly by the players, to actually unconsciously cause this to happen, separate from the way the attack powers themselves were designed. Its hard to say for sure if the game is designed to coincidentally create this problem for blasters, or if its built into the overall playstyle for blasters.

In either case, analysis of the SR passive resistances suggests that the actual effect of a power like this, and its casual glance effect, can be radically different. In analyzing the SR passive resistances, I was able to determine three numbers: the average experienced resistance, the average effective resistance, and the average surivivability effect.

The average experienced resistance was the average level of the resistances given a set of combat parameters. That number was about 15%. The average experienced resistance was the average amount of damage deflected by the resistances: that number was about 3.5%. Then there was the effect on survivability: the amount of fixed resistance necessary to improve SR survivability by the same amount as the passive resistances: that number was about 24%. All three ways of "measuring" the resistance seem logically sound, and yet they all disagree with each other dramatically.

Similarly, the damage boost might, or might not improve things more than it first appears. But I have reasons to suspect they *don't* improve things by much more than it appears on paper. I just don't have a way to prove it quite yet.

[ QUOTE ]
Arcana,

Your 'interesting data points' chart for the damage buff, could you whip up a similar one for the to hit buff? I'm pretty sure it will look dismal but I'd still like to see it, if you have the time.

[/ QUOTE ]

Not exactly, no. The reason is, as hinted in the original post, there is no trivial way to confirm the tohit buff of defiance with high degrees of accuracy. Basically, what I had to do is this: hit a blaster with -regen and -dmg, and then knock his health down to levels I wanted to test, and then let him swing away and watch for streakbreaker changes which told me the blaster's net accuracy had crossed a streakbreaker window threshold. Needless to say, this does not produce nice numbers like the damage buff ones, and it took the better part of a day just to confirm two data points on the tohit buff curve to any degree of precision at all.

So, I'm very certain the damage buff equation is correct. I'm only reasonably sure the tohit buff one isn't radically wrong, although as I said, I have confirmation on both expressions from a red name.

[ QUOTE ]
I'm extremely disappointed that an image of Statesman would be used in such a way. I know that we designed City of Heroes to appeal to "children of all ages." My proudest accomplishment with the game is that many parents (including my brother) play with their children...sharing, I hope, the same joy for the comic book world that I've had ever since I was 8. I cannot express the shock that I felt seeing that a character created by Cryptic would be used in the same sentence as Columbine.

[/ QUOTE ]

I too was shocked at the implications that video made. As soon as I'm off work, I'm superjumping over to the producers of that documentary and unleashing my fire imps upon their base. Let's see them try to finish a film production when they have to farm all that prestige again.

Oh, look at that, its time for my medication again.


My opinion, for whatever its worth, is this: the issue of violence in video games is a smoke screen. Its a non-issue. I'm sure if you conduct a perfectly up and up, legitimate, controlled, scientifically solid study of video games' impact on the game playing public, you'll find some evidence that video games promote violent behavior, to some extent or another, in much the same way that if you look hard enough, eventually you'll discover that everything causes cancer.

The question is whether the potential amount of violence it might promote either excuses that violence by being promoted to a direct cause, and equally important, whether the ability for video games to promote violence isn't overriden by much stronger environmental influences.

I'm sure there are lots of things that promote violent behavior to a greater extent than video games. Heck: soccer does.

Probably the thing that promotes violent behavior the most, by several orders of magnitude, is alcohol consumption. So why are video games talked about as a promoter of violence significantly more than alcohol. Mainly, because video games are a convenient target.

That's all this is. Video games may or may not have a net negative influence on society. But for the most part, we don't *judge* voluntary activities in a free society that way. It only happens sometimes when the activity is considered marginal, in the sense that it has few supporters. Which means it only takes a few detractors to beat up on it.

Columbine is the perfect example of this. Kids wear trenchcoats, play video games, and shoot up a school. Can't attack gun ownership, because that is a losing battle: its not a convenient target. Can't attack trenchcoat ownership, because people will rightly assume you are insane. So attack video games, because that is - or appears to be - an easy target.


Why is City of Heroes, in the persona of Statesman, in that video? Probably because the people making it really don't *care* if any particular video game is good, bad, or ugly. The video game industry is an easy target, and when you are beating up a relatively defenseless target, you don't really need to fight fair.

Understand: "fair" in this case is not whether they represent all sides of this issue. For me, the inherent unfairness of the situation is that video games are actually being *made* into an issue at all. There is no good reason to do so, that cannot be applied to a great many other things but aren't, because they are not easy targets.

In other words: Bully was a target because ironically the ones doing the targetting thought it could be bullied.


I would be more concerned than I am now if I didn't believe that because video games are being targetted only because they are an easy target, this problem is very likely to moot itself in relatively short order, as video games become as pervasive as television watching, or if you prefer, alcohol consumption. At that point, it will cease to be an easy target, and the people who currently believe they can beat up on it will simply be forced to go away. They will be obsoleted by societal evolution.

Ultimately, I think we have to tolerate these social dinosaurs, primarily making sure they don't stomp on too many things, before they flicker out. We can fight them, but time is on our side: we only have to contain them long enough for them to be marginalized out of existence.


In case my position isn't obvious, I'll state it bluntly. Suppose you could prove that (some) video games actually made (some) people more violent. That would only mean to me that psychiatric medicine needs to catch up to help the disturbed individuals for which this is true. It means nothing to me in terms of the video game industry: nothing. Why? Because if you could prove to me that certain books made some people more violent, like Catcher in the Rye, I would have the same response. I wouldn't ban Catcher in the Rye, and I wouldn't conduct a government study to find out what future writers need to not do, to avoid replicating Catcher in the Rye, and bar them from doing it. To me, that's a place I won't go.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
just got a reply.
"It is NOT resisted. This is as intended."

[/ QUOTE ]

is it intended that it kicks ice/ in the teeth while the other sets get a poke with a stick?

[/ QUOTE ]

Invulnerability gets hammered pretty well. Most of that set's non-S/L mitigation comes from Invincibility. Slotted up and surrounded by enemies ( needs 10 to cap it ) gets you right around 30% Defense. When Rage drops it is a -20% to Defense. The Rage drop causes Invincibility to stop effecting mobs so it stops granting the Defense bonus. This takes you from 30% to -20% Defense. I'm not sure about Ice but if that set gives you 30% with your armors you would still have 10% Defense.

Rage can amount to up to about a -50% Defense Penalty for Invulnerability Tankers.

[/ QUOTE ]

Keep in mind that another part of the damage mitigation in the Ice set comes from the Chilling Embrace damage debuff, which also stops affecting enemies during the Rage crash.

[/ QUOTE ]

Theoretically, Ice Armor can duck the rage crash by hibernating. Since you can't affect anything else anyway, there's no cost to hibernate, and you are invulnerable while hibernating. And it cycles fast enough to be available every rage crash, with appropriate slotting.

Also: hibernate has +recovery. My guess is that ten seconds of hibernation gets you close to a full bar of endurance.

[ QUOTE ]
I'm a bit late to this party, but fan-[censored]-tastic. 2 hours well-spent plowing through this guide. You mentioned a link to download the simulator code?

DS

[/ QUOTE ]

Yes, but that link might be broken now: I'm in the process of making files such as that available for download again from another place. When that is available, I will post a link here.

As discussed previously, the sim is in python. Its as cleaned up as I can make it, or rather want to spend time to make it. It does have graphics dohickies, but plastered on just for fun. I'm thinking of rewriting it as a project to familiarize myself with the XNA framework, which would make it a bit more user friendly. But that's quite a ways off given the free time available.

Originally, it was just a bunch of code to run calculations for me, then it grew into a script to evaulate the behavior of dull pain, then damage ticks, then the passive resistances, and then suddenly it was a damage/defense calculator. So I threw out the original code and rewrote it as a per-tick attack/mitigation simulator. It isn't a *game* simulator, its more of a discrete calculator: if you want to know what happens when you have damage that looks like this, and defense that looks like that, but you don't know what happens if reconstruction is still recharging then, while an attack lands over there, this is designed to do that. Millions of times if necessary, to determine the average case.

I'll probably be posting a link to the original python-based simulator within a couple of weeks in this thread, if you are interested.

[ QUOTE ]
Now the question is whether or not the exponential equation it's based on is appropriate or whether it should be adjusted, or whether it should be an exponential curve to begin with.

[/ QUOTE ]

The exponential part is not really a problem: asking for it to be linear doesn't necessarily make the low parts higher, it might just make the high parts lower.

What might make more sense, though, is to give it a "floor." If it had a 5% floor, say, then the buff would start off as about as good as scrapper criticals on minions, unslotted. Then, "shallowing" the curve could allow the boost to reach 20% by 50% health, which would be close to the average buff of scrapper criticals, slotted (because scrapper criticals are post-slotted boosts, while defiance is a base damage boost). Everything above that would be gravy.

Its not really that much different from the current curve, its still exponential, but it offers a slightly better curve at lower levels.

[ QUOTE ]
[ QUOTE ]
Even at full health, you are getting somewhere in the neighborhood of a 1% damage boost due to defiance. You're not likely to notice it much until you reach about 50% health, though.

[/ QUOTE ]
Enough proof for me. You lost some where in the area of 600 points of damage from 1500 to get 900 and gained 1.xx of damage. Thats 5% if you want to average it out to a round number, no thanks. It still sucks, i am going to fight my hitpoint bar every time.

[/ QUOTE ]

At 911.9 health (for me), that was 66% health, and about 6% damage boost. That is approximately the benefit of scrapper criticals on minion targets. Its not trivial.

I wouldn't head to "criticals are still better because its better for an occasional big number than a lot of slightly bigger numbers" quite yet. One of the things I'm looking at is whether criticals end a fight quicker than a comparable defiance boost. It is not as simple as you might expect, because one odd coincidence I'm starting to see, and I cannot explain, is the "sliver of health" oddity. Alarmingly frequently, the damage numbers of the attacks blasters get combine to leave a target with a sliver of health, at unbuffed damage levels. Its happening at a higher frequency than can be explained by probability. And that serves as critical context for whether defiance can just get "lost" as a tiny buff that doesn't actually help in practice. It can, at surprisingly small levels, sometimes reduce the number of attacks necessary to kill by one (sometimes at levels of only 8%).

I suspect it has something to do with the fact that damage numbers were designed around a particular "metric" of number of attacks necessary to defeat a minion, and were scaled from there. That might have left a statistical "footprint" in how damage interacts with targets' health bars: how much "wastage" damage is done to a target (the amount of damage over and above necessary to defeat the target, and subsequently the likelyhood that one less attack might leave the target with a small sliver of health) might not be remotely random at all.

[ QUOTE ]
Good job Arcaneville.

[ QUOTE ]
1. Blasters are defiance buffed constantly

Even at full health, you are getting somewhere in the neighborhood of a 1% damage boost due to defiance. You're not likely to notice it much until you reach about 50% health, though.

[/ QUOTE ]

This is true, but is that small damage buff for 60% health and above is almost meaningless. It only adds a few more points of damage that will still need another follow up attack to finish, most likely will already result in excess damage without the Defiance boost.

[/ QUOTE ]

Well, its hard to hover at 10% health for a blaster, but its not hard to hover at about half health before popping respites, and the defiance buff there is comparable to assault. Its not bad (but smaller than the average bonus due to criticals). I think the message I'm getting from the defiance numbers is that its about 2-3 times too weak on the low end, not totally worthless, if we assume that blasters should get a similar benefit from defiance as scrappers get from criticals. But its hard to say: blappers especially tend to get *much* higher defiance numbers: those quarter-health blappers you see destroying everything are doing so because they have a damage buff comparable to perma-rage.


[ QUOTE ]
[ QUOTE ]
Because the relationship between critter damage and blaster health changes with level, defiance means different things at different levels. I'm still looking at defiance for blasters at levels other than 50 (defiance itself doesn't change, but I'm still looking at the damage numbers verses health issue at different levels).


[/ QUOTE ]

Have you also considered resistance buffs on the blaster? Higher resists will slow down the damage the blaster recieves, but if the blaster is taking less damage, then he's living longer where normal damage will be enough.

[/ QUOTE ]

Yes, indeed. Sturdies can be amazingly bad news for blasters to have in PvP, unless they are heavily debuffed. It also adds some context to the whole "would you trade your tier 9 attack for MoG" question. MoG would start by making the blaster perma-build up, and somewhere around "one third health" would make the blaster perma-capped. That's a serious damage buff coming out of MoG.

[ QUOTE ]
Nice stuff Arcana.

Wish I could confirm the bar lag/defiance lag thing... the number of times my defiance bar has skyrocketed with the damage increase coming several shots later (if I'm not dead) has been a marked one. This seems to be a prevalent complaint. Is it all just player perception?

[/ QUOTE ]

I don't know. What's odd to me is that in all of my testing, not only have I not seen the issue you describe, if anything, I've seen the *opposite* happen, because the bar itself is slightly laggy: the damage buff happens *first* and then the defiance bar jumps a fraction of a second later. I've never seen it happen the other way around. That doesn't mean its not happening, but especially where such things are likely to occur if at all (during much faster paced combat than my test conditions), its also very difficult to actually *see* what the defiance boost is moment by moment.


I'm of the opinion that defiance damage should be a separate floating number, like criticals. That way, blasters can actually see the defiance boost in a way more psychologically satisfying (and perhaps a bit more obviously). This was done recently for defense: the various "dodged" and "deflected" messages which show that defense is actually "working" were added purely to give a psychological boost to the visibility of defense. Defiance damage should probably get similar treatment.

*Especially* since defiance is one part numerical buff, and three parts psychological buff, its important to make sure its highly visible to get the psychological boost its intended to provide.

Well, here's how defiance actually works, based on some testing I was able to do in between clicking presents like a maniac, and some discussion with the devs.

First of all, here are defiance's two operating equations:

DamageBoost: 25% * 2 ^ [(45 - HealthPercentage)/10]
ToHitBoost: 1.25% * 2 ^ [(45 - HealthPercentage)/10]

HealthPercentage, as far as I can see, always rounds *up*. So 58.22% health is 59%, as far as the equation is concerned.

Lets see defiance in action:

[ QUOTE ]

You Brawl with Harlequin Fencer and deal 20.12 points of smashing damage!
Brawl is recharged.
You activated the Brawl power.
Harlequin Fencer Parries you for 101.19 points of lethal damage.
You Brawl with Harlequin Fencer and deal 20.18 points of smashing damage!
Brawl is recharged.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 20.2 points of smashing damage!
Harlequin Fencer Parries you for 101.19 points of lethal damage.
Brawl is recharged.
You activated the Brawl power.
Harlequin Fencer Stabs you with her Rapier for 120.47 points of lethal damage and reduces your defense!
You Brawl with Harlequin Fencer and deal 20.22 points of smashing damage!
Brawl is recharged.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 20.37 points of smashing damage!
Harlequin Fencer Parries you for 101.19 points of lethal damage.
Brawl is still recharging.
Brawl is recharged.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 20.55 points of smashing damage!
Brawl is recharged.
Harlequin Fencer Parries you for 101.19 points of lethal damage.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 20.64 points of smashing damage!
Brawl is recharged.
Harlequin Fencer Parries you for 101.19 points of lethal damage.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 20.68 points of smashing damage!
Harlequin Fencer Stabs you with her Rapier for 120.47 points of lethal damage and reduces your defense!
Brawl is recharged.
You activated the Brawl power.
Brawl missed!
Brawl is recharged.
Harlequin Fencer Parries you for 101.19 points of lethal damage.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 20.89 points of smashing damage!
Harlequin Fencer Stabs you with her Rapier for 120.47 points of lethal damage and reduces your defense!
Brawl is recharged.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 21.1 points of smashing damage!
Brawl is recharged.
Harlequin Fencer Parries you for 101.19 points of lethal damage.
You activated the Brawl power.
Brawl missed!
Brawl is recharged.
Harlequin Fencer Stabs you with her Rapier for 120.47 points of lethal damage and reduces your defense!
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 22.04 points of smashing damage!
Brawl is recharged.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 21.45 points of smashing damage!
Brawl is recharged.
Harlequin Fencer Parries you for 101.19 points of lethal damage.
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 21.66 points of smashing damage!
Brawl is recharged.
Harlequin Fencer Stabs you with her Rapier for 120.47 points of lethal damage and reduces your defense!
You activated the Brawl power.
You Brawl with Harlequin Fencer and deal 21.91 points of smashing damage!


[/ QUOTE ]

Here, I'm brawling a level 50 fencer, with my level 50 blaster. My level 50 blaster has, with accolades, temp powers, and whatnot, 1566.2 health at the moment. I did not keep track of my health throughout this particular run, but what I can tell you is that it starts at full health, and ends at 911.9 health. Lets see if this matches the defiance equations.

911.9 health is 911.9/1566.2 = 58.22% health, which I said earlier is rounded up to 59%.

(45 - 59)/10 = -1.4

2 ^ (-1.4) = 0.3789

0.25 * 0.3789 = 0.094732

Now, blaster brawl at level 50 is 20.01, if I remember correctly pre-defiance (not precisely 20, but pretty close). It might be 20.005 or something like that, that the game is rounding, but we'll use 20.01 for now (if you use 20, you'll get pretty close to the same results for reasonably large values of defiance). So, if defiance is giving a 0.094732 boost, your brawl should hit for:

20.01 * (1 + 0.094732) = 21.9056, or about 21.91 rounded to two decimal places (which the damage chat always does).

(Note: the defiance calculations appear to be actually carried out to a significant number of places: if you do them to only two or three decimal places, you tend to see rounding errors, but at five places, they tend to match much more accurately what the combat chat says)

So at 911.9 health, the actual damage of brawl matches the predicted damage of brawl, based on the defiance equations. Lets push our luck: lets work backwards and see if the previous brawl (21.66) actually also matches. We have to work backward: the fencer hit for 120.47 damage on the previous strike, so if we were at 911.9 health at the end, we must have been at 911.9 + 120.47 = 1032.37 health prior to that, and therefore the defiance boost at that point would have been based on 1032.37/1566.2 = 65.91% ~ 66%

Defiance then becomes: 0.25 * 2 ^ ((45 - 66)/10) = 0.058315

The total damage then becomes: 20.01 * (1.058315) = 21.18


Whoops, something went wrong: the actual damage of that brawl was 21.66. And the thing that went wrong in this case is that a health tick occured at that point in time. Can we correct for that?

Yes we can. Health ticks are always 5% of health (powers like Health only speed up the ticks). So a health tick is 1566.2 * 0.05 = 78.31 health. Our health was actually 78.31 lower than the calculated 1032.37, or 1032.37 - 78.31 = 954.06.

(Why lower? Because we had to be lower, so that when the health tick actually happens, we end up at the correct final value of 911.9. If we actually *were* at 1032.37, the when the health tick occurs, we would be even *higher* than that, and then after the fencer damage we would be even higher than 911.9.)

954.06 health is 954.06/1566.2 = 60.92% ~ 61% (always round up).

Then defiance becomes 0.25 * 2 ^ ((45 - 61)/10) = 0.082469

The calculated brawl damage is then: 20.01 * (1.082469) = 21.66

Which matches the brawl damage shown in the combat chat precisely.


I have about a hundred different data points for defiance, and they all seem to match this equation so far, so I believe the damage equation itself is on pretty solid ground. I have less testing of the tohit buff equation. I have three streakbreaker-related tests of the tohit one, by very carefully engineering accuracy tests under conditions where health doesn't move (by being regeneration-debuffed). They *seem* to confirm the second equation, although I wouldn't say those tests are 100% definitive.


In any case, both equations have been verified by red name, so they are the best ones we have.


Just for fun, here are some interesting defiance points:

Code:[/color]

---

100.00%	0.55%	full health

57.00%	10.88%	assault (10.5%)

45.00%	25.00%	small rage

42.00%	30.78%	circa one-shot death from average boss

41.00%	32.99%	Fortitude (31.25%)

35.00%	50.00%	Large Rage (50%)

31.00%	65.98%	Aim (62.5%)

28.00%	81.23%	Perma Rage (80%)

25.00%	100.00%	Build Up (100%)

15.00%	200.00%	Circa one-shot death from average LT

9.00%	303.14%	Damage Cap (assuming 3-slot damage)

8.00%	324.90%	Circa one-shot death from average minion

---

The interesting stuff, as you might expect, happens at or below half health. But you don't need to be at a sliver of health to get a significant benefit. Also, in spite of conjecture to the contrary, my own testing indicates that the defiance bar lags, but defiance itself doesn't: if you have X% health when you fire a shot, you'll get that much defiance boost, period. If you are there when you hit the activation button, you'll get the boost (at least as far as my testing so far indicates).

I stuck in some data points on where you become in danger, on average, from being one-shotted by the average LT and Boss, in very rough numbers (these are my own ballpark estimates). Meaning, if you are fighting one of those, any defiance number lower than that is a buff you don't really want to ever have to see: its genuinely a desparation buff, inless you have resistance buffs operating. Also, +health accolades help (those numbers presume base blaster health at level 50, verses typical level 50 critters).


The important things to note about defiance:

1. Blasters are defiance buffed constantly

Even at full health, you are getting somewhere in the neighborhood of a 1% damage boost due to defiance. You're not likely to notice it much until you reach about 50% health, though.


2. Defiance is not "laggy"

All my testing to date seems to indicate that if you have health X at the moment you activate an attack, then that's the defiance boost you'll get. Much like all damage-boosting abilities. The defiance bar is very laggy, though, and also has a maximum movement "speed."


3. Don't rely on the accuracy (tohit) buff for defiance

Its there, but you aren't likely to notice it until you are very nearly dead. This is probably intentional.


Because the relationship between critter damage and blaster health changes with level, defiance means different things at different levels. I'm still looking at defiance for blasters at levels other than 50 (defiance itself doesn't change, but I'm still looking at the damage numbers verses health issue at different levels).


[i]Edited to fix the order of some numbers in the defiance chart[i]

[ QUOTE ]
[ QUOTE ]
i need to start writing on more controversial topics. the most I've ever gotten in reply to a topic or guide I've made is when Castle replied and people flooded into ask him stuff.



[/ QUOTE ]

Never be afraid to be vilified my friend.

The EVILGEKO is loved, hated and feared...and that's just by me and Arcanaville. I don't know what the rest of you lot think!

[/ QUOTE ]

I know I'm afraid that I love to hate him.

[ QUOTE ]
That doesn't jibe with my understanding of the power, which is that the ticks are not independent. I think that the 50% tick is only considered if the 75% tick occurs. This matches all the player testing I have ever seen going back to I2 (when I rolled my En/En), which shows a 25% chance of 0 ticks, a 37.5% chance of 1 tick and a 37.5% chance of 2 ticks. If the probabilities were independent, it would be 12.5%/50%/37.5%.

[/ QUOTE ]

To be honest, I've never tested the probability of nova's waves to hit. Its not a difficult thing to do, so I'll probably do so at some point (like, after I catch up with all the I8 badges). As I said, pending actual testing, I was suggesting where Iakona's 87.5% calculation came from.

He might have also just done a quicky calculation and not even checked his own data to see if the second tick was dependent on the first, and simply made an error.

[ QUOTE ]
[ QUOTE ]
Nova with 3 damage SOs + Aim + Build Up + 1 small damage inspiration = 87.5% chance of killing a non-resistant +1 Lieutenant. I think this is a perfectly reasonable amount of damage for the drawbacks of the power. I don't think there should be a guaranteed one-shot AoE kill for +1 Lieuts; I think it should stay as a high chance of one-shot kill as it is currently.

[/ QUOTE ]
After thinking this over some more, I'm not sure I agree. I've run through the math and I don't see how nova w/ 3 SO's, BU, aim, and 1 small insp (1+0.95+1.00+0.65+0.25=3.85) defeats a +1 lieutenant if only the base ticks hit. At least the first optional tick needs to hit to defeat a lieutenant. That tick occurs only 75% of the time. So, in addition to a 5% chance to miss entirely (assuming capped tohit), you also have a 23.75% (0.25x0.95) chance that nova will hit but fail to defeat a given lieutenant. That suggests only a 71.25% chance of killing a +1 lieutenant, not 87.5%.

Code:[/color]

---

Level	Dmg_Mod	Scale	Dmg	Enh	Enh_Dmg	Lt_HP

33	43.84	3.0	131.5	3.85	506.4	650

40	50.90	3.0	152.7	3.85	587.9	780

50	55.61	3.0	166.8	3.85	642.3	860

 

---

[/ QUOTE ]

I haven't spent time verifying Iakona's probability numbers for nova, so I'm assuming for now they are accurate. I think what Iakona is saying in his nova numbers is that the first tick has a 75% chance of hitting, and the second tick has a 50% independent chance of hitting, so there is an 87.5% chance of at least one tick hitting (i.e. the second tick is not contingent on the first).

That would still mean there was a net 83% chance of defeating that +1 LT, because of the mandatory 5% chance to miss, as you mention.

[ QUOTE ]
Nukes need "slightly" increased damage, if right now they are 87.5% odds of taking out a +1 lt, those odds need to be raised to 95%. A bad roll can still happen, call it a critical failure, but it should not happen more than 1/20 times, preferably a bit less.

[/ QUOTE ]

Essentially, that's the same thing as asking for the base damage prior to the two probabilistic waves being able to take out a +1 LT, because none of the nukes are autohit, and therefore all of them have the 5% minimum chance of missing altogether.

[ QUOTE ]
[ QUOTE ]
I even posted something called "all fights should be AV fights" or something like that, as a part of this idea. The notion was that the stuff that *really* counted, should be like one AV fight, not like 1000 minion fights.

[/ QUOTE ]

Where was I for this? Man, I really missed out by missing that thread. I would have been behind the concept whole heartedly. Honestly, the big fights are what make comic book action entertaining, it isn't the hordes of minions. I will say that if you are going to fight hordes of minions, it should be done well and the gameplay should reflect and provide the visceral never ending combat that you are in. I always fault CoH for not really enhanceing it's strongpoint in my opinion.

[/ QUOTE ]

All I remember is that I got a couple of "that's interesting" replies, and a whole megaton of missing-the-point "it'll take forever to level if we had to fight nothing but bosses and AVs all the time" replies.

I think if I redrafted the idea today, I'd probably get a much more animated discussion around it. Although, who knows: perhaps half the responses to my "Open Letter to Cryptic" post were of the form "that's basically impossible" (fortunately, and predictably, I don't think the devs themselves concurred).


[ QUOTE ]
In other news, what do you think of Alan Wake? It looks like a great game, part shooter, part puzzle solveing, part RPG and the pace is kinda left up to you the player. I think it may be a model for how a non-linear game should be designed, but I may be getting ahead of myself and reading too much into the reviews.

[/ QUOTE ]

I only know what I've read: its by Remedy, it looks way cool, and the game play is episodic in nature. If they pull the episodic part off well, that could make it great.

[ QUOTE ]
The great games I have come to find, mix challenges with relaxing moments in the game or give the player time to come down off of his edge and steel himself for the next challenge.

[/ QUOTE ]

A long, long, time ago, I suggested that the most important MMO concept that CoH decided to keep, that they should have jettisoned, was the concept of XP itself.

In my pie-in-the-sky, naive view, what CoH should have done, which in my opinion would have fit the genre better, and satisfied by hardcore and casual players better, was to eliminate the idea that you kill things for XP, and then when you have enough XP you go up one level, and go kill other things for XP. What they should have done instead is designed the game so that it took a specific event or set of events - call them task forces or trials if you like - that gained you a level. If you wanted to level fast, you could focus on those. Passing the right trial(s) took you to the next level. If you didn't want or need to level as quickly, there would be all sorts of other activities you could do, all designed to be enjoyable in and of themselves. Missions that earned you things you could only use at that level, say, or story arcs you could run that would put you up against special villains that only existed at that level. The rewards for these activities would be something other than XP, but still fun. Unlockable things, if you will. Today, we'd call these things badges, temp powers, accolades, special enhancements, maybe even inventions.

The idea would be if you wanted to relax and cruise in the game, you could do cruising missions that weren't designed to kill you, they were designed to be Superman flying through Metropolis knocking off the occasional bank robber - not a challenge at all, just something fun to do. When you really wanted to put your nose to the grindstone, you could do the "get to level 22" trial, and continue on.

Because the content of the game would be specialized to each level (or level range), there would be no point to racing to 50. You would play at level 21 for as long as you wanted, and no longer. Once you were bored, you would hit the level 22 trial, and you'd be 22.

I even posted something called "all fights should be AV fights" or something like that, as a part of this idea. The notion was that the stuff that *really* counted, should be like one AV fight, not like 1000 minion fights.


A very amateurish idea, with a lot of hanging threads that would need to be addressed for it to be at all workable. I find, though, that as time passes, the idea looks increasingly better, and not worse. In fact, mayhem and safeguard missions seem to echo at least some of these ideas, if not the central one.

[ QUOTE ]
[ QUOTE ]
If they were:

1. Really fast, so you could use them in an emergency




[/ QUOTE ]

Just out of curiousity, do you feel that Dreadful Wail is well-designed then, given its one-second activation time?

[/ QUOTE ]

My highest sonic is like 20, so I don't have first hand experience with Dreadful Wail, but nofuture lists DW's activation time as 1.97, or about 2 seconds. I think an emergency power should have an activation time of about 1 second. DW also has a stun, which lasts longer than knockback typically incapacitates for (11.92 seconds, apparently) which is good, but lower than the crash incapacitates the blaster in return, not so good.


DW does bring up the interesting point, though, that some nukes seem better designed to be alpha strike weapons, because they have disabling effects (what's the point of disabling a dead target) which would be useful if you could continue to attack the remaining targets. If I were designing the nukes, because I like diversity, I would strongly consider making two, possibly three types:

Type 1: Blizzard, Dreadful Wail, Thunderous Blast

Design: designed to do heavy damage (basically as now) and incapacitate in some fashion the targets; the crash does not crash endurance to zero, it just stops recovery. This allows follow up attacks to eliminate remaining targets.

Type 2: powers such as nova, inferno

Design: does somewhat more damage than currently (i.e. circa 20% more damage), crashes to zero, stops recovery for a short period (circa 5 to 10 seconds maximum). Activates in one second or less, but can be designed as a DoT to allow for longer more interesting animations; first wave of damage and associated secondary effects strike within one second.

Type 3: powers such as full auto, rain of arrows

Design: not true nukes; designed to offer significant firepower more frequently, but at a much lower damage level than the high end nukes. No significant crash.


Something like that. I'd probably want to fiddle around with the mechanics of the powers (like range) a little more, but that's what I would do if I were designing them.

[ QUOTE ]
I think the real problem with the nukes have always been their activation time. They need to trigger a little faster to work as that last ditch effort to save yourself.

[/ QUOTE ]

This is part of what I was saying about the nukes. We don't exactly know what they are for (and Castle is implying they might not have been for anything at all specifically). But if they are meant to be alpha strike weapons, they should have enough damage potential to get the job done all by themselves, because their crash precludes finishing off crippled targets. If they are meant to be last-ditch finishing blows, they don't need to necessarily have as much damage to one-shot things, but they should activate very fast.

Which is the dillema regarding them. In my opinion, they are too slow to be finishing blows, and they are too weak to be alpha strike killers given their crash. Which means the combination of attributes makes them poorly designed.

If they were:

1. Really fast, so you could use them in an emergency

or

2. Strong enough to ensure there was almost no chance of something being left behind to fight you while you were crashed

or

3. Didn't crash endurance AND stop recovery (did one or the other, at worst), so you could continue to fight (albeit in a possibly degraded fashion) after the blast


They'd at least be *specifically* designed to function in a certain way, without players having to attempt to bypass the powers' restrictions in some way.

[ QUOTE ]
Probably the best way to answer this is to refer to comics -- When The Human Torch 'goes nova' he puts everything he has into it and pretty much knocks himself out. That's the 'feel' these powers are meant to reflect.


[/ QUOTE ]

I'm not questioning the conceptual validity of the crash, but rather whether they do enough damage given the crash.

I would suggest that this would represent the *lower* limit of how much damage nova should do: since a solo heroic mission can spawn a +1 LT, nova should be guaranteed to kill such a critter, if its designed to incapacitate the blaster completely.

Doesn't it seem reasonable, Castle, that if the big bangs are conceptually designed to incapacitate the blaster, at the very least they should be capable of delivering a *guaranteed* kill on a heroic solo spawn, assuming no extraordinary resistances? Extra damage beyond that would be extra damage beyond that: useful for people who run on invincible, for increasing the utility of the power in teams, etc.

Edit: assuming it hits at all: I'm not advocating for nova to be autohit

[ QUOTE ]
That other discussion was heading for nuclear territory, so I'm spawning a new thread about the "nukes" alone.

Arcanaville, you said you were having trouble killing even-level (yellow) lieutenants with Nova. That strikes me as very odd and outside my own extensive personal experience with Thunderous Blast which should actually be doing less damage than Nova does, if City of Data is to be believed (though I'm not entirely convinced the numbers City of Data has for Thunderous Blast and/or Nova are correct).

I've had a handful of +2 level (red con) lieutenants survive a Thunderous Blast, but it is the exception to the rule rather than the rule itself.

How do you have it slotted? Are you using Aim and/or Build-Up first?

[/ QUOTE ]

What I said was, its not a guaranteed kill. The maximum damage potential of nova is enough to kill something like a red LT, but its not guaranteed. That was important to my point that if the power leaves you drained, but leaves a reasonable chance of leaving threats behind, its not really designed to be a very good alpha strike weapon.

And I've used nova as an alpha-strike weapon literally thousands of times. I'm not asking for help in maximizing its usefulness: I'm suggesting that nova and the other big blasts might be useful in spite of their design, instead of because of their design.

[ QUOTE ]
heh, maybe it is a BUG, but what I find in I-7 is that the only thing that can't attack at zero endurance is ME! Or other players. Mysteriously, at level 40-44-ish, PvE mobs still have alot of devastating attacks with zero endurance and I am guessing they would even after being Drained by 8 Elec/Elecs since Electric Drains can't do negative side draining (meaning negative integers). It's not in the Coding.


[/ QUOTE ]

No, the problem is that there are actually two endurance draining effects: actual endurance drain, and endurance recovery suppression (debuff). Drain to zero is not a big deal for many critters, because they really don't have a lot of attacks in the first place. Usually, they hit you with their one and only attack (or two, maybe), and then they are going to have to wait about 5 seconds (depending on the critter) regardless of what you do while it recharges. If you drain them to zero, but they still have recovery, they can recover enough endurance in a few seconds to have enough to fire that one attack again. That's all they would have the endurance to do, but that's all they can do anyway. If you only had one attack, you wouldn't have much endurance problems either.

To really stop a critter cold, you need to drain them to zero, and *also* debuff recovery so they can't get that endurance back on a time scale comparable to how long their attacks recharge anyway. And that's the hard part.

It isn't really a question of "unfair" in the sense that critters play by different rules, because they do play by our rules. Rather, its almost never the case that players are really fighting one on one fights.


Here's one way to look at it: if players are "balanced" around fighting three minions, then in fact, those three minions have 300 endurance: triple the endurance, and triple the recovery rate, of the average player. Its as if players were constantly fighting masterminds, who also have the same endurance circumstances (namely, their pets are almost never out, so its as if they have almost unlimited endurance).

In a certain sense, if players are balanced against multiple critters, critters should actually have less endurance per critter. Consider a level 50 blaster facing off against 3 even minions. The blaster has about 1200 health, the minions have about 400 (430) each; in a sense, the blaster is facing off against something with approximately the same amount of health as he does (but that "thing" is more vulnerable to AoEs, of course). But he's facing off against something with 300 endurance. I think its possible someone didn't think about endurance balancing in the same sense they thought about health bar balancing between players and critters.

That doesn't necessarily mean critters should all have 33 endurance, but it does suggest that in actual fact, its the *fact* that critters play by the same "rules" as players, but players aren't balanced against fighting them one on one, that is part of the fundamental problem with endurance drain.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
More then likely, other plans are in the works too, possibly good or bad. Hopefully, not bad, as Castle did post some of his ideas here first and should they bad, the mobs will likely look to burn blasters at the stake first

*Straps on his helmet* Luckily, I am prepared for the worst should it happen. *slams door on his underground bunker* CLANG!

[/ QUOTE ]

*mutters* Dam it Fusilier, locked me out AGAIN! *finishes setting detonation cord* Hope you like the new door chime I installed *sets timer and runs for cover*....*waits*....*waits* Damn Timebomb takes forev..*BOOM!*

[/ QUOTE ]

*Pops head up from his bunker* What the heck is all that racket?...*sees a plume of smoke a short distance away* Dam it Outrider, do you really feel it was necessary to blow up my garage door?

[/ QUOTE ]*Sneaks past Fusiller into his bunker while he checks his garage.*

[/ QUOTE ]
*Helps Sgt_Paine "liberate" Fusiller's supply of Spam and Funyuns and SlimJims*

[/ QUOTE ]

Worst abuse of the Global Frequency, ever.