Arcanaville

[ QUOTE ]
[ QUOTE ]
Do the +Regen/+Recovery Unique IOs (Regenerative Tissue, Miracle, Numina's Convalesence) stack with themselves? They are/were listed in game as having a 120 second duration...

[/ QUOTE ]
Those are Unique now ... so stack with each other, sure; but you can't use more than one of any of them.

I figured the "120 second" text was referring to the 120 second duration on those types of special IO effects in click powers like Heal Other, but that shouldn't apply at all to an auto- or toggle-type Healing power.

[/ QUOTE ]

I think the question he's asking is, if you were to slot it into aid self, say, and cycled aid self twice in 120 seconds, would you get the buff twice. And I think the answer to that question is no, or at least I don't think you are *supposed* to get the buff twice. If you do, I believe that would be a bug.

Two more Pool A drops:

Stupefy: Acc/Rech
Lethargic Repose: Acc/Sleep

Only one new drop for me tonight (it was a kinda weird testing night for me), but it fills a hole:

Pool A:

Cleaving Blow: dmg/rech

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
One thing I think is wrong, I wanted to test out how good minerals was, so my stone tank went out with minerals and weave running and found an even con illusionist lieutenant. 63 attempts later, she finally hit him, so I'm fairly convinced that villains don't get the benefit of a streakbreaker.

Mr Minotaur 50 stone/axe tank Freedom

[/ QUOTE ]

Minerals is 25% defense to psionic unslotted, and 39% defense slotted. Weave is 5% defense unslotted for tankers. Combined, you are basically perma-eluded against psionic-typed attacks. Anything lower than net 20% tohit, and the streakbreaker doesn't kick in until 100 consecutive misses, as indicated in the guide.

The streakbreaker affects everyone. I've directly measured it affecting NPCs. In fact, my current hyper-precision method of measuring defense values relies on NPCs being affected by the streakbreaker, and would not work at all if NPCs were not affected by it.

[/ QUOTE ]
This is strange, because psychic clocks were hitting me like 30% (subjective estimate) of the time, do they have a huge +acc/to hit ? That was the basis of me doubting the minerals figures in the first place.

[/ QUOTE ]

I do not believe they do. I can't explain given the information how they could consistently land at that rate through slotted minerals and weave.

Pool A:

Mako's Bite: Acc/Dmg
Tempered Readiness: End/Rech/Slow
Call of the Sandman: End/Sleep
Positron's Blast: Dmg/End


Pool B:

Sovereign Right: Dmg/End

[ QUOTE ]
One thing I think is wrong, I wanted to test out how good minerals was, so my stone tank went out with minerals and weave running and found an even con illusionist lieutenant. 63 attempts later, she finally hit him, so I'm fairly convinced that villains don't get the benefit of a streakbreaker.

Mr Minotaur 50 stone/axe tank Freedom

[/ QUOTE ]

Minerals is 25% defense to psionic unslotted, and 39% defense slotted. Weave is 5% defense unslotted for tankers. Combined, you are basically perma-eluded against psionic-typed attacks. Anything lower than net 20% tohit, and the streakbreaker doesn't kick in until 100 consecutive misses, as indicated in the guide.

The streakbreaker affects everyone. I've directly measured it affecting NPCs. In fact, my current hyper-precision method of measuring defense values relies on NPCs being affected by the streakbreaker, and would not work at all if NPCs were not affected by it.

[ QUOTE ]
[ QUOTE ]
Wait... Posi goofed?

[/ QUOTE ]

I take credit for the mistake, but blame the fact-checker for being sick that day.

I think I can now safely say that I have forgotten more about this game than most will ever know, and I proved it.

[/ QUOTE ]

Its a shame that in a week, you won't remember proving it.

(But this is why we don't purge dev posts)

[ QUOTE ]
[ QUOTE ]
One comment on the article: now that we know the strategy that geko had in mind, its worth noting that there were actual players discussing the possibility of just that strategy way back in 2004. In fact, the article describes attacking from one side, letting the mitos spawn, and then switching to attacking from the other side. The player version of this strategy recognized that there are *two* respawns, and the attacks would have to come from points on a triangle: first evading respawn #1, and then evading respawn #2. The problem was one of coordination: it took so many players to successfully attack Hamidon (at least in a relatively open raid without carefully crafted characters) that coordinating the "jumps" was a non-trivial problem.

But the actual *strategy* was, in a sense, discovered by the players, just not to my knowledge executed.

[/ QUOTE ]

Yep, and the main reason it was never successfully done in an open raid was that it was not only more vulnerable to griefing than the hold strategy, but it was extremely vulnerable to ignorance. Half a dozen people who didn't know what was going on could conceivably spread the Mito spawn to areas where it would be fatal.

On top of this, every *fix* to Hami proceeded to make this stategy more and more unlikely. Today, it's outright impossible, as Hami and the Mitos simply have too much range for it to work.

[/ QUOTE ]

It actually predates the hold strategy by several months. Its main problem I think was coordination. I know it was *attempted* on at least a couple of occasions, but I wasn't a direct witness to any of the attempts (not my server).

Timing is critical: so critical, I'm not sure if geko actually stepped through it moment by moment himself. Once you get Hami below a certain health level, he spawns on you. You have to be able to relocate rapidly, and continue fighting, so he doesn't regain ground on you and go back above half-health and stay there too long (or you'll get another respawn).

The thought was that timing needed to be so perfect, that the full version of this strategy involved at least three strike teams. Three attack positions (A, B, and C) are selected. Strike team one attacks from position A, strike team two waits outside the goo near position B and specifically doesn't attack (this prevents them from drawing respawned mitos), strike team three waits near position C. When attack team one reduces hami to half health and causes him to respawn, they ditch and strike team two immediately moves to position B and continues the attack, out of range of the respawned mitos, giving hami no chance to recover. Strike team one then regroups, joins strike team two, and continues the fight. When that group causes a respawn on position B, strike team three moves in at position C and continues the fight, while strike teams one and two regroup and join them, to finish the fight.

Imagine coordinating that on your server.

One comment on the article: now that we know the strategy that geko had in mind, its worth noting that there were actual players discussing the possibility of just that strategy way back in 2004. In fact, the article describes attacking from one side, letting the mitos spawn, and then switching to attacking from the other side. The player version of this strategy recognized that there are *two* respawns, and the attacks would have to come from points on a triangle: first evading respawn #1, and then evading respawn #2. The problem was one of coordination: it took so many players to successfully attack Hamidon (at least in a relatively open raid without carefully crafted characters) that coordinating the "jumps" was a non-trivial problem.

But the actual *strategy* was, in a sense, discovered by the players, just not to my knowledge executed.

[ QUOTE ]
How would this be any easier than just giving hostage NPC's max perception so the stealth wasn't an issue?

[/ QUOTE ]

You'd have to change all the NPCs in the game, for one. For another, this issue came up specifically in CoV beta, when the complaint was that stalkers *had* to drop Hide in order to escort kidnap targets. At that time, the devs said that they felt it was working as intended for players to have to drop stealth to escort unwilling NPCs. Giving the NPCs max perception was already rejected once as a solution for players wanting to remain stealthed while escorting certain NPCs.

[ QUOTE ]
[ QUOTE ]
Along this line... What can you tell us about the various +Stealth IOs from the Travel Sets?

Specifically, can they be slotted in passive powers? If so, how is this going to interact with NPC hostages and such that won't interact with a stealthed character?

[/ QUOTE ]

Currently:
<ul type="square">[*]They can be slotted into passive movement powers.[*]You can have one of each type on your character for some massive bonuses.[*]Escort NPCs will not follow you if these are slotted on your characters.[/list]
Now, before the uproar starts, I have to say that unless we find a satisfactory solution for the last item on the list, these IO's have a high probability of being cut before reaching the live servers. Even if that problem is solved, it is still unlikely that you will be able to slot all 4 versions in a single character -- either that, or the stealth aspect will be reduced greatly.

[/ QUOTE ]

The simplest suggestion I can think of, that would require very little real work, would be to hand out to everyone a new inherent power, called "LookAtMe" (well, something cool), that was a zero endurance burning toggle that suppressed self-stealth. Basically, the presumption is that you should always be able to suppress your own stealth, by just jumping up and down and waving your arms, if necessary. Players that had passive stealth could toggle this during escort missions that required visibility, and temporarily suppress their own stealth.

Seems to cost nothing but a free give-away toggle power, and it doesn't alter mobs, affect other teammates, alter the mechanics of stealth, or modify the mechanics of the invention IOs.

[ QUOTE ]
So again I take your little "hoarfrost will not be perma" remark and say... yes it will, just like I said before. So if you want to add in all the "factors" for your rage... well then I will add in all the "factors I mentioned" for hoarfrost also.

[/ QUOTE ]

Have fun with that, under whatever color sky you dwell.

[ QUOTE ]
Now I am confused. In the past you have spoke of "perma rage" in SS.

Rage 240/1.95 = 123 second recharge... and a 120 second duration. So it is off 3 seconds every 120 seconds or in 240 seconds off by 6 second which right online with hoarfrost. Now going straight by your &amp;#8220;If I say something is perma, everyone should know that I mean "absolutely no gap between cycles." I am truly thrown off by you speaking of perma rage.

And again this is all without a single set bonus&amp;#8230; By your very own definition Hoarfrost is going to end up being &amp;#8220;closer to perma&amp;#8221; the rage you have called perma.

[/ QUOTE ]

1. I use the term "perma-rage" because its the generally accepted term, so people know what I'm talking about (the use of rage as continuously as possible). However, to the best of my recollection, I've never explicitly stated how much recharge is necessary to achieve permanent status of rage.

2. However, I'll do so now. Rage has 240 second recharge, 120 second duration (and 1.17 second activation time). It also has a 10 second cool down. As a practical matter, all levels of recharge from +0.863 to +1.020 have the same overall benefit, because of the cooldown: it allows rage's benefits to accrue 120 seconds out of every 130 (those numbers factor in activation times).

Recharge levels higher than +1.020 would theoretically allow rage to stack, and therefore have a potentially different overall net effect.

Therefore, all levels of recharge from about +0.863 to +1.020 can be said to make rage "perma" in the sense of the rage benefits being available as often as the cooldown period would ordinarily allow. But in another sense, barring exploits, no amount of recharge actually allows rage to be literally perma in the sense most people mean, because the cooldown prevents rage's benefits from being useful for certain unavoidable windows of time.


And to be honest, I don't think you're confused. Everyone else knows what I mean when I refer to perma-rage as a generally accepted term. And they know what I mean when I say that I don't personally describe anything as being perma unless its actually perma, because I can actually calculate how much recharge is actually required to make it literally perma, and saying anything else would be injecting an unnecessary subjective opinion into something that can be easily calculated and stated for other people to make up their minds on. So if you're confused, you're in a very small minority.

[ QUOTE ]
unless you want to split hairs... thats perma.

[/ QUOTE ]

For me, perma is perma. But if its different for you, you might want to let me know what's your threshold for perma in terms of seconds of downtime.

If I say something is perma, everyone should know that I mean "absolutely no gap between cycles."

But if you think that's perma, without the +0.375 speed buff at all, dull pain with hasten and 3-slot recharge averages about 30 seconds of downtime. Is the difference between perma and not perma somewhere between 30 seconds and 9 seconds?

[ QUOTE ]
[ QUOTE ]
7.5% recharge only stacks to 37.5% which very few set combinations could do and even then that's only a bit more than half of Hasten.

[/ QUOTE ]

/SR with hasten

.375: from IO's
0.70: from hasten
1.00: base
0.95: enhancements
0.20: quickness.

Total: 3.255
Time needed for hasten to be perma: 3.75
We also have yet to add in "perma IO set + recharge"

Not an advantage... hardly. You looking at hasten have 20 seconds of down time tops... so 120 on 140 off. Even better apply it to ice/* tank with the same numbers - 20% for quickness. In which you will get perma hoarfrost. Or even here for other sets: manuever, CJ, Steath, Weave, grant invis. There is 37.5% on its own. Add that to regen or any set with a dull pain for pretty close to perma dull pain with hasten.

[/ QUOTE ]

Its a recharge SO. Sure, its better than nothing, but its not all that great relative to the other buffs out there that exist.

Its all fine and good to say what you could do with +0.375 speed stacked on top of all this other speed, but it begs the question of whether or not the +0.375 itself is all that important relative to all that other speed.

Its the consolation prize for defense sets, because if the devs only gave us the +7.5% run speed enhancer, they would never be able to stop laughing.

Also: Hasten + 0.375 speed + 3-slot recharge is not perma-hoarfrost. Not since I6, anyway.

[ QUOTE ]
It's an old question around here. Would you rather have:

- A large, non-stacking bonus, or
- Several small stacking bonuses, or
- Both, and a hard cap to how much it can help, or
- Some wierd mix?

It really seems like the simplest solution should be to figure out the biggest bonus that any set of set bonuses should be able to give to an aspect, and have a cap.

For instance, the maximum +defense from set bonuses is +10% (or whatever). The maximum total ToHit from Set Bonuses is +12.5% (or whatever). Done. Then you can eliminate all the 'same bonus doesn't stack above 5 instances' code.

[/ QUOTE ]

Hypothetically, a power set with two heals, two resistances, and two defenses might conceivably be able to hit the caps on the bonuses offered by all three sets, while a power set with 6 defenses would get capped by the caps and be unable to buy anything else.

Hypothetically, of course. The devs wouldn't really do that, would they.

[ QUOTE ]
[ QUOTE ]

Castle is saying that didn't work out too well in testing, so its being changed so that when you slot those powers into pet casting powers, you'll *always* get the buff - to your pets - instead of only getting the buff when you activate the pet casting power.


[/ QUOTE ]

"Always" in this context meaning "as long as the pets are within the buff effect radius."

Not trying to be pedantic, just aiming for completeness of information.

_Castle_, this is great news. Thanks for the update!

Scrap

[/ QUOTE ]

I should have said "you'll always radiate the buff" when I said "you'll always get the buff."

[ QUOTE ]
(Also, on the Paragonwiki page, the Res/+ 3% def is marked as unique, which means I won't get to slot one into each of my passives like I had planned.)

[/ QUOTE ]

I'm guessing that happened like today.

[ QUOTE ]
[ QUOTE ]
Hmm. Well it looks like Invuln took quite an IO-beating in the last two days. DA and Regen still look pretty good though. Time to go to plan C for my Invulns on build optimization.

[/ QUOTE ]

What happened to Invulnerable? I don't see what changed with this information.

[/ QUOTE ]

I posted this *after* it was discovered that some IOs were marked "unique" and therefore you could only slot one of them, and *before* it was discovered that (at least for now) the Res+Def IO wasn't one of them (the smart money said it was going to be one of them).

Edit: also, the duration-based click-buff thing.

[ QUOTE ]
Just a little confused here. Is he saying that is one slots an IO that has res to taunt of placate it will work 100% of the time if said IO is in a passive?

[/ QUOTE ]

There's only one such IO currently in existence, and it buffs pets, not players, when slotted in pet casting powers. All pet casting powers are clicks, not passives or toggles. The original intent was for those buffs to buff the pets when the pet casting power was clicked, and last for a certain amount of time before expiring. Castle is saying that didn't work out too well in testing, so its being changed so that when you slot those powers into pet casting powers, you'll *always* get the buff - to your pets - instead of only getting the buff when you activate the pet casting power.

[ QUOTE ]
Nice guide. Consumed virtual seconds is a useful tool indeed.

It's worth noting (if you did, I didn't notice) that some of the really long recharge powers (which frankly, are what we worry about most), may well have recharges close to 232 seconds; which makes it feasible for them, at least, to assume a flat .36 recharge out of hasten. A convenient shortcut in the rare case its useful-- e.g., recharges on click accolades and perhaps, leaving long-recharge powers like EMP and controller AoE Holds unslotted or minimally slotted.

Even if you're close to 232 seconds, it can be a convenient kludge. If you have 1 recharge in a 240 second hold, for example, you know you'll burn at LEAST 68 real seconds at the increased recharge.

[/ QUOTE ]

My motivation for posting the guide was to lay out the basics of looking at recharge in potentially unconventional situations. That's why I didn't just say "this is 3-slot hasten, this is 3-slot dull pain with hasten" etc. My suspicion is that when I9 goes live, a lot more players are going to be experimenting with recharge, and asking questions like "can I get perma-DP back"? or "what's the best possible uptime for Elude?" or "if I slot up these IO sets in order to get these set recharge buffs, what does that do to hasten?" There's going to be a lot more "weird" recharge out there, and that will mean the "standard" speed questions might not be the norm anymore.

Hopefully, this guide will help a lot more players be able to figure their recharge questions out for themselves, and will allow more players to help more players that can't.

[ QUOTE ]
a = 0.3 * 120 * (1.95 + 8a) / 422
a = .522

[/ QUOTE ]

Yikes. You're right, I must have pushed the wrong calculator button on that one. Also, it didn't feel right either, but I let it go. Grr...

Just to make sure:

a = 0.3 * 120 * (1.95 + 8a) / 422
422a = 36 * (1.95 + 8a)
422a = 70.2 + 288a
134a = 70.2
a = 0.524


[ QUOTE ]
Of course, this is wrong, since the factor [a] for a buff cannot be higher than its maximum buff;

[/ QUOTE ]

To be specific, the calculation is implying that this level of recharge is so high that the powers are self-stacking: if the recharge buff doesn't self-stack (and AM doesn't) then its actual average recharge buff contribution cannot exceed its base recharge buff. So 0.524 implies its actually up all the time and giving 0.3 all the time.


Thanks for the correction: I've updated the guide with a proper rendition of this calculation and an explanation for what's going on.

Guide to Calculating Recharge v1.0

How to calculate recharge.


The question of how recharge works comes up a lot, as does the question of how to calculate how fast a power will recharge, especially given the complexities of hasten. Moreover, I suspect more people are going to be fiddling with recharge when I9 goes live and players start experimenting with the recharge buffs that exist in the Invention System. I have a technique that works pretty well: its not the fastest computational technique, but it has the advantage of being relatively easy to understand and adapt to a lot of different complex situations.


How does recharge work?

First of all, recharge boosts are expressed (like lots of things in CoH) in terms of percentages, like 33% recharge buff (terminology note: speed buffs generally refer specifically to movement speed like running. Recharge buffs refer specifically to the speed boosting that reduces recharge). But in actuality, those buffs are really decimal numbers: in this case, 0.33. The way recharge buffs improve the recharge of a power is by formula:

NewRecharge = OriginalRecharge / (1 + RechargeBuffs)

So if you have +33% recharge and a power takes 10 seconds to recharge normally, with that recharge buff it will take 10 / 1.33 = 7.52 seconds to recharge. Recharge buffs are additive: if you have a +0.33 recharge buff, and a +0.2 recharge buff, the new recharge will be the old recharge divided by (1 + 0.33 + 0.2) or 1.53.

Something important to know: powers have an activation time and a recharge time. And recharge buffs only speed up recharge time: they do not affect activation times at all. So if a power takes two seconds to activate, and eight seconds to recharge, it can be used every ten seconds. A 0.33 recharge buff does not mean the power can be used every 7.52 seconds: it means the power recharges in 8 / 1.33 = 6.02 seconds, and can be used every 2 + 6.02 = 8.02 seconds. Notice that because recharge cannot work on activation times, its always a little weaker than you might expect in terms of speeding up things like attack powers.

Actually, its a little more complex than that: powers can have interrupt periods, activation times, cast times, root times, and all manner of complex mechanics. But for our purposes, we will call the time from the moment the power is activated, until the moment it fully discharges and begins recharging the activation time.


How to Calculate Recharge

Now, that's how recharge works. How do you calculate how long a power takes to recharge. Well, the simple way is what I just said: add up all the recharge buffs operating on the power, and divide the recharge time of the power by the factor (1 + TotalRechargeBuff). But what if it isn't that simple? What if you have variable recharge buffs.

When do you have variable recharge? When you have recharge buffs that aren't on all the time. And there are examples of that in the game. The most prominent example is hasten. Hasten is a recharge buff that has 450 second base recharge time, and 120 second duration. While its up, it offers +0.7 recharge buff. Because hasten isn't up all the time (as of ED its no longer perma-able with just slotted recharge) we can't just divide hasten's 450 recharge time by the recharge buff operating on the power: that recharge buff is one thing when hasten is up, and another thing when hasten is down. So how would we calculate hasten's uptime?

Answer: we use an alternate method for calculating recharge. We tend to think of recharge buffs as reducing the total time necessary to recharge a power. But there is an alternate way to think about it. When a power has 450 seconds of recharge, we can say that without speed buffs, that power earns one tick every second towards recharging to full. After 450 seconds, the power earns 450 ticks, and is fully recharged.

If the power is operating under a speed buff, it earns more ticks per second. With a +0.33 recharge buff, it earns 1.33 ticks per second. Theoretically, after 338.35 seconds, the power would earn 450 ticks, and be fully recharged.

Trust me: it works. And its useful. Suppose we have hasten, with no slotted recharge at all. How long does it take to recharge? Easy:

When hasten first activates, it offers a +0.7 recharge buff for 120 seconds while its up. Then it drops, and offers no speed buff until it recharges.

For the 120 seconds it is up, hasten earns 120 * (1 + 0.7) = 204 ticks towards being recharged. That means it has 450 - 204 = 246 ticks to go. It has to earn those at one tick per second, so it has 246 seconds to go until its fully recharged. So the total cycle time for hasten is 120 + 246 = 366 seconds.

So hasten, with no slotted recharge, takes 366 seconds to recharge. Its activation time is 0.73 seconds, so hasten's total cycle time is 366.73 seconds. Thats how often you can use it.

What if we slot it with recharge? Well, if we slot 3 even SOs of recharge, that is a +95% recharge buff, or more properly a +0.95 recharge buff. Now, its cycle time changes: during its 120 second uptime, it now earns 120 * (1 + 0.7 + 0.95) = 120 * 2.65 = 318 ticks. That means it has 450 - 318 = 132 ticks to go. These remaining ticks have to be earned without hasten being up: it will earn those ticks at a rate of 1.95 ticks per second (because of the slotted recharge). So it takes 132 / 1.95 = 67.7 seconds to earn those ticks. That means hasten recharges in 120 + 67.7 = 187.7 seconds. Its total cycle time is 187.7 + 0.73 = 188.43 seconds.


Okay, time to move to slightly harder problems. How long does it take for Elude to recharge, if you have hasten, and hasten is 6-slotted with even recharge SOs for recharge (+110% recharge buff), and Elude is 3-slotted with even recharge?

First, we tackle hasten. Hasten earns 120 * (1 + 1.1 + 0.7) = 336 ticks when up, leaving 450-336 = 114 ticks to go, which takes 114 / (2.1) = 54.3 seconds to earn. Hasten's total cycle time is 120 + 54.3 + 0.73 = 174.03 seconds.

Now, in one hasten cycle, Elude is going to earn a certain amount of ticks towards its recharge. When hasten is up, Elude will earn 120 * (1 + 0.95 + 0.7) = 318 ticks. When hasten is down, which it will be for 54.03 seconds, Elude will earn 54.03 * (1 + 0.95) = 105.36 ticks. So in one complete cycle of hasten, Elude will earn 318 + 105.36 = 423.36 ticks. Now, Elude and Hasten will not be perfectly lined up all the time: sometimes Elude will see more of hasten's uptime than at other times. But what will Elude's recharge be on average?. Well, on average, Elude will need 1000/423.36 = 2.36 cycles of Hasten, slotted the way described, to fully recharge (423.36 ticks per hasten cycle, so that's how many cycles it needs). So Elude will recharge in 2.36 cycles of Hasten, or 2.36 * 174.03 = 410.71 seconds. Elude has 1.5 second activation, so Elude's total cycle time will be 410.71 + 1.5 = 412.21 seconds. Basically, Elude can be used once every 412.21 seconds, and its up for 180 seconds, so it has 412.21 - 180 = 232.21 seconds of downtime: 3 minutes up, 3 minutes, 52.21 seconds down.

We can do this for other powers, like dull pain, even when the average number of hasten cycles required is less than one: consider hasten 3-slotted with (even) recharge, and dull pain slotted with 3 even recharge enhancers also.

Hasten with 3-slot recharge has 188.43 cycle time, as mentioned above. Dull Pain earns 120 * (1 + 0.95 + 0.7) = 318 ticks when hasten is up. It earns 68.43 * 1.95 = 133.44 ticks while hasten is down. But it only needed 360-318 = 42 ticks to fully recharge. That means while Dull Pain earns 451.44 ticks per hasten cycle, it only needs 360 ticks to recharge, and it therefore only needs 360/451.44 = 0.797 hasten cycles to recharge, or 150.18 seconds on average. Sometimes a little faster, and sometimes a little slower, depending on how Dull Pain and hasten align.

One little bit of subtle complexity for the mathematically inclined: when recharge powers are up, powers like Elude earn more ticks faster, so all other things being equal, its much more likely that a power like Elude will complete its recharge while a recharge power is up, rather than down. That means the *number* of times the power will recharge at below average times (faster) is going to be higher than the number of times it will recharge at above average times (slower). But the times when it is slower will weigh heavier on the average, because they will obviously happen for longer periods of time. Overall, this is a wash: the average recharge rate per unit time is going to be the computed average. But it might *seem* like its faster, because you'll see more instances where its faster than slower, and are less likely to factor in the fact that when its slower, you are spending more time waiting (especially for long duration/long recharge powers like Elude).


Multiple Recharge Boosting Powers

Time to graduate to the hardest problem. Suppose you have accelerate metabolism and hasten? Each has recharge buffs. How do you calculate the recharge times of either, when each speeds the other up?

Unfortunately, I don't have a nice simple calculation that determines this. I'm forced at this point to switch to algebra. AM is a +0.3 recharge buff for 120 seconds, and has 422 base recharge time and 2.03 activation time. Hasten, as before, is a +0.7 recharge buff for 120 seconds, and has 450 base recharge time and 0.73 activation time.

First of all, we presume that AM has an average recharge buff, call it a, and hasten has one, call it h. We can then create two expressions based on those average recharge buffs.

The average recharge buff of AM is based on its average cycle time. Its 0.3 * 120 / (cycle time). And its average cycle time is based on the average recharge buff it experiences. If its 3-slotted, it always has a 1.95 recharge. On top of that, it gets the average buff from hasten, and also its own average buff. So the total recharge it experiences on average is (1.95 + a + h). That reduces its recharge from 422 to 422/(1.95 + a + h). Its total cycle time is 2.03 + 422/(1.95 + a + h). So its average buff is 0.3 * 120 / (2.03 + 422/(1.95 + a + h)), which reduces to:

a = 0.3 * 120 * (1.95 + a + h) / (2.03 * (1.95 + a + h) + 422)

Now, I like precision as much as the next person, but I'm not really interested in solving quadradic equations just to find the recharge of AM. I'm willing to estimate. AM's activation time is very small relative to its cycle time, so I'm going to assume its zero for the purposes of this calculation. It will only alter the numbers by a second at most. That reduces that expression further to:

a = 0.3 * 120 * (1.95 + a + h) / 422

Similarly, you can derive an expression for the average recharge buff of hasten:

h = 0.7 * 120 * (1.95 + a + h) / 450

Two equations, two variables. All you need to do is solve. Having a calculator helps multiplying all the terms through. What you get is:

h = 0.500, a = 0.228

So the average recharge buff of hasten is 0.500, and the average recharge buff of AM is 0.228. That makes it easy to calculate average recharge times.

Hasten: 450 / (1.95 + 0.500 + 0.228) = 168.04 seconds
AM: 422 / (1.95 + 0.500 + 0.228) = 157.58 seconds

The total cycle times become just activation time plus recharge time:

Hasten: 168.04 + 0.73 = 168.77 seconds
AM: 157.58 + 2.03 = 159.61 seconds

This algebraic method is extensible to any number of speed boosting powers operating simultaneously. You could use it to see what happens when 8 rads all use AM on each other, for example.

a = 0.3 * 120 * (1.95 + 8a) / 422
a = 0.524

When you calculate the average recharge buff of a power to be higher than its actual base buff, that's saying the buff is recharging fast enough to be self-stacking. What the math doesn't know is whether or not the buff is actually capable of self-stacking. In this case, AM doesn't self-stack. Therefore, its maximum average buff is the case where its up permanently, in which case its average buff is equal to its base buff, or 0.3.

AM recharge: 422 / (1.95 + 8 * 0.3) = 97.01 seconds
AM cycle time: 99.04 seconds

More than perma, with no hasten.

The algebraic method above pretty much works all the time, with one caveat: if the recharge buffing powers you're looking at have downtimes that are low relative to their uptimes, what you see in game will tend to be about the calculated average. But the larger the downtimes are relative to the uptimes, the more likely it is you'll see wilder extremes in recharge, with sometimes much longer and sometimes much shorter ones. A good rule of thumb is that if the recharge buffing power is up at least half the time, you'll likely see recharge rates near the computed average most of the time.


Summary

How to calculate recharge:

* With no recharge boosting powers, or with constant recharge buff, take the base recharge of the power, and divide by the total recharge factor, which is (1 + All Recharge Buffs)

* With one recharge boosting power that is not perma (up and down), use the tick method. Calculate how much ticks the power earns while the recharge boosting power is up, how much when its down, and then calculate how many total cycles the power in question needs. Even if it turns out to be only a fraction of one cycle, the method still works.

* With two or more non-perma recharge powers, you have to resort to the algebra method above.


Edit: Thanks to Plasma for correcting a major calculation error.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I think that I would be happy if they did this:

A. Slightly reduce the effectiveness of IH as it stands, but turn it back into a toggle.

[/ QUOTE ]

I would be happy if they slightly reduced the effectiveness of Elude but turned it back into a toggle also.

[/ QUOTE ]

What if it were 30% defense but cost the same as the old IH? I think that would be interesting. I'm not even sure that would be overpowered.

[/ QUOTE ]

I'm...*wheeze*...too fast for you...*huff* *puff* Sur..render now, and *gasp* I'll...spare you...

[/ QUOTE ]

It would actually be rather like I3 perma-elude, with all the constant END crashing and no extra recovery. Without careful endurance management, perma-elude was actually rather like that.