DPS vs. DPA and EPS vs. EPA
I think it really depends on what you're looking at fighting and how. Against things like AVs and GMs where the fight will be lasting a significant time, the DPS and EPS metric is much more significant than DPA/EPA. One the other hand, if you're soloing with a blaster on -1/X1 and you're one-shotting everything you come across, then the DPA/EPA stat will be more relevant.
In general, though, the best answer is probably both. Finding the median between DPA/DPS and EPA/EPS is as important as either stat alone.
My story arcs: #2370- Noah Reborn, #18672- The Clockwork War, #31490- Easy Money
Sartre once said, "Hell is other people." What does that make an MMO?
It depends on whether your endurance cache or your recovery rate is the limit you're hitting. Consuming end faster than you can replenish it in a fight is not necessarily a bad thing; running out of end before the fight is over almost always is.
@SPTrashcan
Avatar by Toxic_Shia
Why MA ratings should be changed from stars to "like" or "dislike"
A better algorithm for ordering MA arcs
It depends on whether your endurance cache or your recovery rate is the limit you're hitting. Consuming end faster than you can replenish it in a fight is not necessarily a bad thing; running out of end before the fight is over almost always is.
|
Years ago, I used to slot the highest cost powers first. Until I realised that often high-cost powers will actually cost me LESS than multiple faster powers used in the same time, or within the same cycle. But even supposing I break my fool's rule of slotting highest cost first and foremost, what then? What will help me run out less - slotting high EPS powers or slotting high DPS powers? We're assuming a static attack chain here. I'm not looking to optimise my power usage, only my power slotting with a static usage.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
Oh, if you're assuming a fixed attack chain being cycled indefinitely, and just trying to figure out which power should be slotted most heavily for endurance reduction, then it's easy. Slot the one with the highest total endurance cost in that chain first.
For a really simple example, suppose you have power A costing 5 end, B costing 7 end, and C costing 8 end, and your static attack chain is ABAC. Then the highest total endurance cost is A at 10 end.
Now, if you're trying to build a more endurance-efficient attack chain, then you're in for some bad news. Damage, endurance cost, and cast cycle for single-target powers is by and large on a fixed ratio, so the only way to consume less endurance... is to deal less damage. If you want to maximize your activity while minimizing your endurance consumption, then you want to choose attacks with the lowest DPA.
Incidentally, the above is also why increased recovery is so very good. Characters are machines for transforming endurance into damage, and there's a finite limit to how efficient you can make them - after which, the only way to increase performance is to add more fuel. What's more, the most efficient method for transforming endurance into damage, by an incredibly wide margin, is AoE. Basically, AoEs tend to be 2.5x the cost of ST attacks, but deal up to 16x the damage. But AoE also consumes a lot of endurance, and if you don't have the means to cover the cost of an AoE attack chain, that efficiency boost is unavailable to you.
(The above should not be construed as "take stamina", just singing the praises of +recovery in general.)
@SPTrashcan
Avatar by Toxic_Shia
Why MA ratings should be changed from stars to "like" or "dislike"
A better algorithm for ordering MA arcs
*blinks*
Im way out of my depth here.
*leaves*
@Damz Find me on the global channel Union Chat. One of the best "chat channels" ingame!
Now, if you're trying to build a more endurance-efficient attack chain, then you're in for some bad news. Damage, endurance cost, and cast cycle for single-target powers is by and large on a fixed ratio, so the only way to consume less endurance... is to deal less damage. If you want to maximize your activity while minimizing your endurance consumption, then you want to choose attacks with the lowest DPA.
|
Also, AoE and Cone powers are actually more efficient endurance-wise, as you mentioned. Basically, you can grab their single-target DPE and multiply it by the number of targets you hit with them. AoEs are balanced at around 4 powers to break even with single-target attacks and cones at only 3. That's one of the reasons I prefer to play against more enemies of a lower level than fewer enemies of a higher level.
Basically, the way I build endurance-saving attack chains (if that were my goal) is to pick and emphasise high-DPE attacks, as that allows me to deal the same amount of damage more cheaply. That's why I would argue that Blaster Mastery immobilize powers shouldn't count as attacks, as their efficiency is HORRIBLE.
---
More to point, when I said I had a static attack chain, I may have misspoken. My attack chains tend to be static, but... Indescribable That is to say, I know what I'm doing step to step, but I honestly couldn't tell you how many times I use which attack to the number of times I use another attack. I usually let power recharge, availability and situation guide my attacks, which is actually quite consistent, just... Not really something I've actually tried to describe, or something I'm sure I even could.
That's kind of where my conundrum comes from - I can't count the number of times I use an attack, hence why I can't just sum up its cost and go from there. I'm trying to use EPS and EPA to kind of suggest just that.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
Actually, I have to kind of disagree here. Different attacks have a different DPE - (points of) Damage Per (units of) Endurance - which depends on specific power balance and AT mods. Now, as a general rule of thumb, damage and endurance are on a fixed ratio... But that's not always the case. For instance, a regular Blaster attack deals around 12.031 DPE, but powers like Snap Shot and Aimed Shot deal rather more, at around 13-14, and Bitter Ice Blast for some reason deals rather a lot LESS, at around 10. In fact, I was just running numbers on Scrappers just yesterday and I realise something quite odd - Fire Blast from Blaze Mastery has a DPE of around 14.154, which is... Surprising, considering it's a ranged attack in a SCRAPPER Epic.
|
(On Fire Blast: all ranged attacks in epic and patron pools for melee ATs now use the melee mod to calculate damage. Fire Blast, like many Fire attacks, also gets DoT "for free".)
Also, AoE and Cone powers are actually more efficient endurance-wise, as you mentioned. Basically, you can grab their single-target DPE and multiply it by the number of targets you hit with them. AoEs are balanced at around 4 powers to break even with single-target attacks and cones at only 3. That's one of the reasons I prefer to play against more enemies of a lower level than fewer enemies of a higher level.
|
More to point, when I said I had a static attack chain, I may have misspoken. My attack chains tend to be static, but... Indescribable That is to say, I know what I'm doing step to step, but I honestly couldn't tell you how many times I use which attack to the number of times I use another attack. I usually let power recharge, availability and situation guide my attacks, which is actually quite consistent, just... Not really something I've actually tried to describe, or something I'm sure I even could. That's kind of where my conundrum comes from - I can't count the number of times I use an attack, hence why I can't just sum up its cost and go from there. I'm trying to use EPS and EPA to kind of suggest just that. |
@SPTrashcan
Avatar by Toxic_Shia
Why MA ratings should be changed from stars to "like" or "dislike"
A better algorithm for ordering MA arcs
To attempt a metaphor: this is why it's cheaper to be wealthy. If you have cash on hand and a reliable income, you can afford to buy in bulk, which makes the price of each individual item cheaper and lets you have more things for your money overall. Similarly, if you have endurance reduction (which multiplies the value of your endurance cache) and recovery, you can afford AoEs, dealing significantly more damage overall and for a lower DPE. (And as you know this, it boggles me slightly that you would avoid Stamina - but that's another discussion.)
|
Interesting. So that gives me an answer that depends on what I do with the powers. Cool, I should be able to ascertain that
Actually, you can count the number of times you use an attack. Go and fight for a while, then copy-paste your combat logs to a text editor and search for the activation message for each of your attacks. Repeat until you feel you have enough data. Calculate accordingly. |
I guess may have to fall back on my back-up - trying to figure out which powers I SHOULD use the most, which powers will cost me even if I don't use them much and then slot from there.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
The devs spent a long time setting up a flexible system so we didn't have to worry about what is best only what will suit our style. If you want to tank then look at dpa and slot end, if you want to scrap then look at dps and max your burst damage.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
Perhaps you can reach endurance efficiency by increasing your damage output? Killing the same amount of critters with one or two less attacks would decrease your endurance consumption per mob/enemy.
|
Granted, I could improve my damage/endurance efficiency by slotting for more damage, but I'm already kind of doing that by default, and it isn't enough. I'm looking for alternate means, and endurance slotting is the other obvious solution. The question is what to slot first and what to omit altogether.
By the way, what people said before is true - if you want to deal more damage, you have to spend more endurance. Sometimes disproportionately more.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
That's... Actually brilliant! OK, if we assume I can get enough targets in my AoEs to overcome their lower efficiency (call it three or four), then I'm actually getting a pretty good deal on those powers and I shouldn't worry about their endurance slotting. On the flip side, if I use AoEs on single targets, like I tend to do with, say, Slice, then I'll want to slot those powers more heavily as they'll be less efficient for that purpose.
Interesting. So that gives me an answer that depends on what I do with the powers. Cool, I should be able to ascertain that |
TL;DR: Slot more endredux in AoEs so that you can use them more often without running out of end and get better performance.
@SPTrashcan
Avatar by Toxic_Shia
Why MA ratings should be changed from stars to "like" or "dislike"
A better algorithm for ordering MA arcs
That's... actually not at all what I meant. You got the part where AoEs are more efficient than STs - that's the "buying in bulk" part. The thing is, you need to be able to afford to buy your damage in bulk. If you compare defeating 3 enemies with pure ST or pure AoE, in the ST example you'll spend 3 units of endurance for every 2.5 you spend on AoE. But if you use ST, you'll also take 3 times as long to defeat all the enemies - which is bad for performance, but good for recovery because you'll recover 3 times the endurance during the fight. If you only recover one unit of endurance in the time it takes to defeat one enemy, you can fight using ST attacks continuously, but will have to spend 1.5 units of time not fighting for every unit of time you spend fighting if you use AoE. That's why you have to be "wealthy" - that is, slot endurance reduction to increase your effective endurance cache, and take recovery bonuses to increase your endurance income - to "buy in bulk" and get the most benefit out of the per-enemy efficiency of AoE.
TL;DR: Slot more endredux in AoEs so that you can use them more often without running out of end and get better performance. |
Thing is... Few AoEs are actually much worse for EPA than single-target attacks, simply because so many of them are so damn slow. The last set of numbers I did put things like Frost Breath and Ice Storm on the low end of the EPA scale, with Bitter Ice Blast and Freezing Touch on a much higher level, at almost twice the EPA. If I'm looking to save endurance short-term, I'll definitely look at slotting the highest EPA. I'm not sure where "short term" morphs into "long-term," however. Not that it matters for this example. BIB and FT still have higher EPS, too.
Come to think of it, that should make slotting my Ice/Ice Baby a lot more straightforward... Not the character I was targeting this contemplation at, though.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
Granted, I could improve my damage/endurance efficiency by slotting for more damage, but I'm already kind of doing that by default, and it isn't enough. I'm looking for alternate means, and endurance slotting is the other obvious solution. The question is what to slot first and what to omit altogether. |
Not awe inspiring but another 20% damage on my fire/fire blaster makes me feel better.
By the way, what people said before is true - if you want to deal more damage, you have to spend more endurance. Sometimes disproportionately more.
|
Perhaps I am not fully understanding the intent here.
I'm confused... are you trying to minimize the EPS of a chain or maximize it's DPE?
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
Ideally, minimise EPS in such a way that doesn't cost me DPS. Which means "just stop attacking" or "run away to rest" is not an option. I'm not sure what's left but to minimise DPE, but if you have ways of achieving this in other ways (that doesn't involve Inventions), I would very much be interested to hear it. If I'm missing something obvious, I want to know about it
|
Yeah, you want to maximize DPE. Calculate DPE (Damage Per Endurance) numbers for all your attacks, using reasonable assumptions on the number of targets for AoEs. Then figure out the optimal chain using the attacks with the largest DPEs (the same you do using DPAs to figure optimal DPS chains) as often as possible. This is the chain that will give you the best amount of damage per endurance point spent. Which is another way of saying that it costs the least endurance per point of damage dealt.
Of course, that chain will assume a certain number of targets hit with your AoEs, so you'll actually need three or four chains, each optimal for a certain number of targets. And you'll have to switch on the fly as the number of targets decreases.
As for tricks to boost your actual DPS/EPS ratio (which is what DPE essentially is), look at ways to reduce end costs (Conserve Power) or boost you damage (Assault, Build Up).
Finally, there's ways to increase your end recovery, which means a reduction to your net end loss(up to now, I've talked only about your brute end loss). Stamina is the obvious choice here, but powers like Power Sink and Consume have the similar effects.
Because of the AoE advantage, the attack sequence with best DPE is not the attack sequence with least EPS. You cannot optimize both at the same time.
Well, actually, I take that back. There's an attack in the game that has unusually high DPE, and also unusually low EPS and EPA. It's even a pool attack, so you can add it to the attack chain of any character, and it does a great job of filling any sized hole in an attack chain to keep you constantly active.
That attack... is Flurry.
You may accuse me of being facetious here, and I kind of am. But the point here is important: the best-performing, highest-efficiency attacks have a high up-front cost, and if you can't cover that cost you'll be locked into low performance and inefficiency. The reason why measures such as IOs and Stamina are so popular is because they mitigate up-front costs (EPS) to a level where you achieve not only higher performance (DPS) but also higher efficiency (DPE). A wealth of endurance (through recovery and lowered endurance costs) enables purchase of damage in bulk, which gives you more damage at a lower unit cost. To the best of my knowledge, there is no way around that.
But this conversation is about enabling your playstyle, not pushing mine. In which case, Flurry becomes a dead serious suggestion as one of the cheapest ways to generate both damage and activity. You should also be pursuing all powers and situations that give "free" damage: Fire attacks, Fury, Containment, Criticals, Scourge, and Defiance all spring to mind.
But I can't recommend many of the usual methods I'd use to increase efficiency, because most of them have unavoidable up-front costs. Hasten, for instance, increases the uptime of click damage buffs and of your most efficient powers - but it comes with an unmitigatable endurance cost, so it's out. Assault can be well worth its price in endurance, but because it consumes part of your desperately needed out of combat recovery, it's out too. The same applies to damage auras, which are AoEs with the endurance efficiency of ST attacks.
Actually, you know what'd really hook you up? Masterminds. They have an effective endurance pool of 700 end (!) to draw from, and get sevenfold (!!!) returns from Assault and other force multiplication tools. Try a Demons/Traps out for size - you have lots of things to do, the whip attacks are actually worth using because they debuff resistance, and Demons are aggressive and tough especially with the buffs and debuffs from Traps covering them.
@SPTrashcan
Avatar by Toxic_Shia
Why MA ratings should be changed from stars to "like" or "dislike"
A better algorithm for ordering MA arcs
Actually, you know what'd really hook you up? Masterminds. They have an effective endurance pool of 700 end (!) to draw from, and get sevenfold (!!!) returns from Assault and other force multiplication tools. Try a Demons/Traps out for size - you have lots of things to do, the whip attacks are actually worth using because they debuff resistance, and Demons are aggressive and tough especially with the buffs and debuffs from Traps covering them.
|
---
But that's just anecdotes. Something you said was more relevant to the thread and actually more enlightening. I couldn't find the right quote for it from the post, so I'll just explain it, instead: The differentiation you make between cost up-front and cost over time. This is what I'm struggling to actually grasp at the moment and what I'm failing to account for. If you can, could you please elaborate on that? What does it constitute, how does it affect endurance consumption, which situations lend themselves to minimising which cost? That sort of thing would REALLY help me.
To be honest, I'm currently looking at a spreadsheet for my Scrapper, and while I have the numbers handy... I'm not really sure what they're telling me. Most of them have an EPA of around 4 or 5, but their EPS is all over the place, from 1.0 plus to below 0.8. I'm not actually sure if that much difference is even significant.
But let's take something simple. Let's take Head Splitter vs. Slash. Head Splitter has an EPA of 5.803 and an EPS of 0.828 whereas Slash has an EPA of 0.976 and an EPS of 3.910. What is this telling me? Which power will cost me more? When will it cost me more? Which power do I want to slot for endurance? Let's just say we're using Head Splitter as a single-target attack for the moment. Maybe I'm fighting hard, single elite bosses in this hypothetical situation. What are the numbers telling me?
---
In fact, while I'm at it, here's something I've never actually thought about - how can I compare the endurance use of one character to another character? Would I just build an attack chain from uptimes and just grab the resultant combined cost, add it to toggle cost and compare that? I... Guess I could do that. But will it help? At least help me see that "Yikes! This guy is using up far too much!" I've always felt that some characters were a LOT worse off than others, but I've never had a good way to SHOW it.
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
In fact, while I'm at it, here's something I've never actually thought about - how can I compare the endurance use of one character to another character? Would I just build an attack chain from uptimes and just grab the resultant combined cost, add it to toggle cost and compare that? I... Guess I could do that. But will it help? At least help me see that "Yikes! This guy is using up far too much!" I've always felt that some characters were a LOT worse off than others, but I've never had a good way to SHOW it.
|
Couple of things:
It's probably just your wording that's throwing me (this talk of building from uptimes for example), but to confirm, the way to calculate the end cost of an attack chain is to add up the end cost (not EPS or EPA) of each power used (obviously add it twice if the power is used twice), then divide it by the total time taken to complete one cycle of the chain (reached by adding up the time taken for each attack including Arcanatime). This will give you a nice simple EPS of the chain, to which you can add toggles.
Also, rather than just comparing end use, I'd compare the net endurance change, i.e. end consumption minus end recovery. Lets you include any endurance increasing powers: Stamina, QR, Power Sink etc.
To be honest, I'm currently looking at a spreadsheet for my Scrapper, and while I have the numbers handy... I'm not really sure what they're telling me. Most of them have an EPA of around 4 or 5, but their EPS is all over the place, from 1.0 plus to below 0.8. I'm not actually sure if that much difference is even significant.
|
EPS, as you are defining it in terms of Endurance per Cycle Time, becomes increasingly meaningless as your attack chain approaches full for exactly the same reason that Damage per Cycle Time does: you cannot activate two attacks simultaneously. The instant you have more than one recharged attack in your tray, it flies straight out the window.
Look at it this way: you're calculating EPS by taking the endurance cost and dividing it by the total cycle time (activation + recharge). Here's the thing, though: that is making the assumption that you are using that power as soon as it recharges, every time it recharges. In point of fact you are almost certainly not doing that, not for every attack you possess; if you have manged to construct a build that lets you use every power exactly as it recharges you could probably also solve a Rubik's cube blindfolded with your feet. The instant you let a recharged power go unused you are altering its effective cycle time, which means you are altering its effective EPS (as you're defining it). In other words, the EPS of a given power is almost never going to remain stable. So that's kind of worthless as a metric.
EPA is a little closer to useful, but honestly it's telling you something you already know, in a slightly different form. If you graphed the DPA of all your attacks and superimposed it on a graph of the EPA of all your attacks (accounting for the AoE factor for AoEs)...they would follow suspiciously similar curves. Baldly put, within a given archetype, Higher DPA Equals Higher EPA. Can you find exceptions? Yeah, probably, but isolated data points are not going to change that basic truth. That's simply an unavoidable result of the rules of power design in this game, which SpittingTrashcan obliquely alluded to earlier: damage has a set cost in endurance, and that cost is 5.2 Endurance Per Point of Scale Damage. Of course DPA and EPA are going to look similar, because they are the exact same thing expressed in different units.
It's like looking at something going 50 miles per hour and something going 80 kilometers per hour: different numbers, same speed.
Yes, yes: exceptions, Snap Shot, blah blah blah. Thing is, those exceptions? They're exceptions. Isolated data points. In the grand scheme of things they aren't going to make that much difference to any single, specific character, since within a given powerset the damage/endurance ratio is going to be pretty stable. Some things are going to throw it off slightly, like free DoT on fire attacks, setwide discounts for Claws, or the odd bug (which would include Snap Shot and Aimed Shot, by the way: I'm pretty sure I've brought this up before, but that's very obviously a case of Castle forgetting to change the end cost when he changed the damage and recharge of the blaster version only in I11; sooner or later somebody on the powers team is going to look at that and go "Hey, that's funny...").
So EPA is also pretty useless for your purposes: like SP said, the attacks that conserve the most endurance are, generally, the ones with the worst DPA.
So what do you do? Well, here's the good news: because recharge increases at a faster rate than damage (and consequently endurance cost), I can give you the Grand Unified Rule of Thumb of Endurance Usage:
Over Time, the Endurance Cost of a Given Power is Directly Proportional to How Often You Use That Power (Assuming You're Not Using AoEs on Single Targets or Something Silly Like That).
What attack do you use most often? That attack is costing you the most endurance! It really is that simple.
Now, that's endurance use over an extended period. When it comes to "burst" end use, the situation is a little different, but still relatively simple. Any time you can begin and end a fight in a short enough time span that the recharge of your attacks is mostly irrelevant, you'll get the most bang for your endurance buck by slotting the powers that use the most endurance in a single use. An example: You walk up to three dudes, Blaze, Fire Blast, Bonesmasher, Total Focus, THE END. In the case of short, discrete bursts of activity like this, with recovery periods in between, your biggest end burners are the ones with the highest endurance cost per individual use, because you are only using each power once, maybe twice.
Unfortunately, at some point on the continuum between short-duration curbstomping and protracted firefights, things get really messy and confused; however, unless you regularly find yourself ending fights really, really, really fast you will be hard pressed to go wrong with the extended activity model.
Frankly, Sam, those numbers are telling you something between jack and squat. They look pretty but they don't mean much.
|
Unfortunately, at some point on the continuum between short-duration curbstomping and protracted firefights, things get really messy and confused; however, unless you regularly find yourself ending fights really, really, really fast you will be hard pressed to go wrong with the extended activity model. |
As far as EPA being basically the same thing as DPA, I guess you're right, but does that mean I shouldn't calculate it just so I can look at it? Why not? I have basic stats written down in the spreadsheet, so the calculations are done automatically. Doesn't really hurt to have an extra column to just so I don't have assume.
But, basically what you're saying, is that powers rarely conform to their expected DPS and EPS. OK, I can work with that. In fact, it's something I've been saying for a while, that most powers will be off their cycle almost all the time. However, smaller powers with a higher uptime will tend to be off their cycle BY MORE than big powers, which are both important enough and take up little enough time that people will WANT to fire them as soon as they are up. However, that doesn't really help with calculations any.
On the other hand, what EPS tells me is that "If I choose to spam this power, this is how much it will cost me over time." And when I look at something like Ring of Fire, which looks like a good attack I'll want to spam, and find out it has easily 50% higher EPS than, say, Fire Blast... I will think twice about whether I want to use it for damage.
By the way, the other side of this, which is EPA, isn't just about exceptions. It's about what's an attack and what isn't. For instance, Chilblain has a lower DPA than Frozen Fists, but it has a higher EPA, for the simple fact that... It's not really an attack. It's a status effect with damage. I can use it as an attack, but the numbers tell me that I'll be doing that at a steep cost. Steeper, in fact, than I would be if I used something else, instead. Something balanced as a pure attack, rather than the half-way point between attack and control. This actually extends to something like Shocking Grasp, the power I got kicked in the face over. Yeah, it deals a lot of damage, but at what cost?
I guess you're still right, in that I can just look at DPE, realise it's doing less damage for its cost than pure attacks and assert from that, but having it written down for me doesn't hurt.
And then there's the other thing I use EPA for. If we view my endurance consumption as a tap, then the EPA of each power the flow that goes through it. If I continually use powers that keep the tap wide open, I'll run out fast. Obviously, using powers that keep the tap more closed will result in my doing less damage, but identifying what's keeping the tap the widest open for the longest time tells me what I want to slot for... Tap reduction...
---
Basically, here's my question - if I took the powers with the highest EPA and slotted those for the most reduction, would I be cheating myself of endurance reduction which I could have better used elsewhere?
Samuel_Tow is the only poster that makes me want to punch him in the head more often when I'm agreeing with him than when I'm disagreeing with him.
|
Possibly. Imagine you use a low EPA (say, 5) attack 10 times in the 30 seconds your average fight lasts. Now imagine you have another attack, this one with high EPA (say 20) and you use this attack exactly once.
Slotting to reduce end costs by 50% in the 1st attack will save you 5x.5x10=25 end over those 30 seconds. Reducing the second's End cost by 50% will save you 20x0.5x1=10 endurance over that 30 second period.
So don't bother with EPA. Estimate the number of times you use a given attack per combat. Multiply that number by it's endurance cost. Do this for all attacks. The attack with the largest number is costing you the most end overall. Slot it for more endred.
Pardon the shoddy title, but I honestly couldn't think of a better one without overflowing the title text box. This is basically a question that's been bouncing around in my head like a piece of flubber for a long time now, and it seems like I'm just not smart enough to figure it out. Here's me hoping someone can out-and-out tell me the answer and see if that helps clear my head.
I'm talking about the ever-present question - which is better? Damage Per Second (of cast cycle) or Damage Per (second of) Activation? And which one has a greater impact? Endurance Per Second (of cast cycle) or Endurance per (second of) Activation?
Arcana once said, and I may be getting the numbers wrong, that as long as you have a combined power uptime of over 75% (that is to say, you spend over 75% of your time animating powers), then "per activation" statistics become more important, whereas if you have less than that, then "per cycle" statistics are more important. Now, it's mathematically provable that if you have a full attack chain and you replace a lower DPS, higher DPA attack with a lower DPA, higher DPS one, your attack chain will actually deal less damage over time. But the problem is that this is only really good in deciding what to take, what to use and how to use it, and even then only in regard to absolute outgoing damage.
As I remember it, the basic algorithm for "deciding" on attack chains is to take the power with the highest DPA, add it in, add in its DPS and its uptime, then add the next one over and the next one over and so on until your uptime exceeded 100% (i.e. you needed to spend more time animating than was actually possible). This gives us powers selected for their DPA, but contributing their DPS.
While that's all fine and good, here's my question: How much COST do those powers contribute? Suppose I have an attack chain figured out and it costs too much. I need to know which power in it costs the most, so how do I figure that out? Do I take EPA or EPS?
One the one hand it seems obvious - EPS gives me drain over time, so that's probably what I should be using. However, it's quite possible to self-drain yourself to almost zero so fast that power recharge stats don't come into play, such as opening up with Dark Regeneration and following that with Fireball. Or, you know, summoning and upgrading your henchmen. What that means is that sometimes it's not a question of how much endurance a power consumes over its whole cycle, but rather how much it costs short-term. As in, RIGHT NOW. If I fire off three powers and they kill my whole end, I won't worry about their recharge. I'll worry that they fire too fast and/or cost too much.
But then on the other hand, I'm reminded of how attacks are selected based on DPA, but contribute their DPS, so what am I really looking at? Am I worried about how much the power will cost me RIGHT NOW even if it recharges slowly, or will smaller attacks cost me more because they recharge too fast? Which one do I look at?