Official Documentation Request - Mana cost of skills / Mana efficiency / Consistency of Terminology

Feedback and Suggestions
1

I can’t find any official in-game explanations on how the mana cost of a skill is actually determined. In fact searching the game guide for “mana” actually yields no results whatsoever.

Likewise, there doesn’t seem to be an officially published formula for the mana cost of a skill.

Based on the dozens of forum/reddit posts I’ve seen from people asking about mana efficiency, I would take it to mean:

“All mana costs are multiplied by (1 / 1 + %Efficiency)”

i.e. 50% Efficiency = 1 / (1+0.5) = 2/3 = ~66% mana cost multiplier.

Problem 1: There are several different verbiages in regards to the scaling of mana costs, and it’s unclear how they are applied (less, more, +efficiency, -# cost, -% cost, etc.)

There are also several different forms of “reduced mana cost” and we have no way of knowing whether they’re all the same mechanic (efficiency) just with inconsistent verbiage, or a different mechanic altogether, and when/where/how they’re being applied.

Using the Acolyte spell “Sacrifice” as an example we have:

  1. MINUS Flat Mana Cost i.e. a flat amount is subtracted from the cost somewhere. Does this occur before or after efficiency calculations are applied? Order of operations would say after but I have a feeling that’s not the case?

  2. MINUS Percent Mana Cost i.e. a percentage of the total cost is… subtracted? I have a feeling this is actually a multiplier, but the way the text reads it’s a subtraction. When/how does this cost reduction occur in the calculations? Is this just “mana efficiency” but with different verbiage?

  3. Costs LESS / MORE Mana - Is “costs more/less mana” just equivalent to mana efficiency? Is it subtracted from your total positive mana efficiency, or multiplied, or what?

Problem 2: There are numerous ways of using mana (spend, expend, channel, cast, drain, consume, “damage dealt to mana”, etc.) and Mana Efficiency doesn’t seem to universally apply to all of them.

Using the Acolyte skill Bone Curse as an example:

“Signet of Agony” : “Bone Curse is converted to an aura that drains 15 mana per second”

There’s a similar node for Aura of Decay:

“Mana Blight drains +2 mana per second per point”

In this case mana efficiency only seems to affect the cost of turning on the aura (or off? why does it even cost mana to turn an aura off?) not the “mana drain” from having the aura enabled.

Right now it feels like there’s some secret code that we aren’t in-on. Like internally EHG have all agreed that there are different mana consumption mechanics specifically so they can balance around mana efficiency, but that info was never passed along in the descriptions.

Is this actually intentional? It would be nice to have an official clarification (preferably in-game) on how mana costs are determined and where efficiency is/isn’t applied as right now it feels very opaque at times.

I know this would be a tremendous amount of work, but I would tentatively propose going through all the instances where “mana cost” appears in a skill tree (even when it’s costs “more” or “less”) and convert all the existing values to corresponding mana efficiency values.

That would at least make more straightforward to understand how the total cost of a skill is being affected e.g. if you have 15% mana efficiency on an amulet, and a skill says “___ has -20% to mana efficiency” (instead of "costs MORE mana) it’s very clear how the effects interact.

3 Likes
2

Flat modifiers (to mana costs & damage) are always applied before % modifiers.

For mana efficiency, new cost = old cost / (1 + efficiency %) as you said. Since both mana efficiency & more/less mana costs are both multipliers, their order is irrelevant

Mana drains aren’t the same as mana costs, nor are channelled costs (channelled costs certainly have their own % modifiers as the Sentinel has a % reduced channelled cost affix).

The Rogue also has a “consume” mana modifier (Coated Blades & Detonating Arrow’s Infused Discharge).

3

Since both mana efficiency & more/less mana costs are both multipliers

True, it shouldn’t matter what order the multipliers are applied in, the question was mostly: what is the interaction between the less/more/efficiency multipliers (are they additive with each other or multiplicative) and are all of these actually multipliers? e.g. “Sacrifice costs -50% mana.” Is -50% mana the same as +50% mana efficiency? If so, why isn’t it worded that way? Mana Efficiency has severely diminishing returns, but a blanket “this skill costs HALF when triggered” is dramatically different (do you get what I mean?)

Mana drains aren’t the same as mana costs, nor are channeled costs (channeled costs certainly have their own % modifiers as the Sentinel has a % reduced channeled cost affix)

I know they’re treated differently in most cases, but “global mana efficiency” mods i.e. the ones that aren’t coming from a skill tree do seem to universally apply to the channeling cost of skills, so why are “drains mana” and “consumes mana” different? Are they literally different mechanics? If so why is there zero documentation to explain that?

All-in-all I am sure we as a community are smart enough to piece together the most likely formula / interaction and read between the lines, I’m just of the opinion that the current mana system is very opaque and inconsistent, and as someone that’s trying to make a build that’s currently heavily limited by mana costs, trying to minimize mana costs has been deeply frustrating at times (I’m the one who posted about drain life mana cost the other day, which turned out to be a bug, but still).

It would just be nice to get some official clarification (and ideally documentation) from the devs before things get too complicated.

4

They’re all multiplicative with each other (though I can’t remember off the top of my head whether a less modifier & a more modifier would be added then applied or applied separately).

No, -50% mana cost is not the same as +50% efficiency.

I’d have to have a look at it but not at 11:18pm…

The testing wouldn’t be difficult with a high level character to swap skill points easily & see what happens to the mana costs.

1 Like
5

For efficiency Less/more think of efficiency as something actually applied to your mana pool (but not displayed this way because it would be infinitely more confusing).

At 100% increased mana efficiency your mana is worth twice as much as other peoples mana. So to pay for a spell which costs 200 mana you only need spend 100 mana, because your mana is just that good. The game displays this as a reduction in cost from 200 to 100 because that’s easier to see, but that’s the actuality of what’s really happening in the math.

A Less/more Cost multiplier acts on the cost. So that same 200 cost spell with 25% less mana cost would cost 150 mana to cast for some pleb with baseline mana, but for you with your super cool Efficient mana it will only cost 75. Thus they interact multiplicatively in no particular order. (200*.75)/2 = (200/2)*.75 = 75 mana from 200.

Efficiency boosts the coolness of your mana, cost reduction reduces the cost, and that’s how they stack. Technically Efficiency is a multiplier on your mana pool and doesn’t REALLY have diminishing returns in the way you think it does. 100% efficiency will always get you another mana pools worth of spells.

Say you have 100 mana (for ease of calculations) 100% inc effiency basically gives you a mana pool of 200, 200% gives you a mana pool of 300. Etc.
So if you have a spell that costs 10 mana, you can cast it 10 times with 0 inc efficiency, 20 times with 100, 30 times with 200, and so on. Every 100% = 10 more casts. They’re not diminishing returns they’re constant. That said, the trade offs you get vs other stats are where you may see diminishing returns in the optimization. As there’s likely a multivariable equation for spell cast maximization but that involves multivariate calculus to find the limits and I don’t wanna do that.

6

I mean I get what you’re saying, but it feels like you have just reworded things to eliminate the whole idea of “diminishing returns”.

Like yeah, the extra 100% efficiency always gets you another one of your base mana pools as casts, but the difference between going from 10 → 20 casts is proportionally way bigger of a deal than going from 30 → 40 casts.

If there were no diminishing return then every 100% efficiency increase would double the amount of casts you get. That’s not how it is though, only the first 100% doubles your mana. Every additional percentage of efficiency you get has less and less of an impact overall.

Your whole ELI5 explanation of mana interactions was very good, and a great way to think about things, like yes, the game is just printing you the new mana cost for your convenience, it could just as easily be reverse and show you the original mana cost, but subtract less than that from your mana pool.

But overall I’m not sure that helps explain how the various other forms of mana modifiers interact with each other.

1 Like
7

I know that these games and their communities love to throw around diminishing returns, but that’s not the correct use of the term. For the same input you receive the same output. That’s not diminishing returns. That’s actually constant returns. You can tell because there’s no exponent in the function. You need an exponent to have diminishing returns as there’s no curve in the line without it. This is linear scaling there are no diminishing returns, You can check as well by taking the first derivative to check your marginal returns.

Essentially with diminishing returns there’s some asymptote to your function that you’re approaching and so your line curves but never hits it.

For example let’s say I’m making a pizza place. I have 1 oven and 1 employee. Each oven can handle 30 pizzas per hour and each employee makes 10 pizzas per hour. I add a 2nd employee, now we’re producing 20 pizzas per hour, I add another employee, but this employee only produces 7 pizzas per hour. Why’s that? I should be getting 30 right? an oven can handle 30. Well there’s no enough space in the kitchen for 3 people running around making pizzas. so I can’t get maximum efficiency. Those are diminishing returns. I applied the same inputs, but I got less output.

1 Like
8

That makes 100% sense and I’m sure mathematically you’re 100% correct, but it just feels like semantics at this point. The OED definition of diminishing returns is:

proportionally smaller profits or benefits derived from something as more money or energy is invested in it

And this kind of situation fits that definition… the first 100% you invest gets you a 100% return. But the next 100% you invest only gets a 50% return.

There is an asymptote here, it’s just horizontal. If the formula is 1/1+(x/100) it’s just gonna approach 0 forever. The question is what happens when the mana cost goes below 1? Does it always round up? Does it ever round down? In the context of mana cost having to be a whole number there are diminishing returns eventually if no matter how much you pump into efficiency you’ll either bottom out at 1 or 0 depending on how their code works.

1 Like
9

Look, do you want to have this argument or do you want to call it a semantic debate? because you called it semantics, then immediately said it wasn’t semantics and that I was wrong.

You have, $100, you gain 100% dollar efficiency, you have $200 that’s 100% profit, you gain another 100% dollar efficiency you have $300, that’s 200% profit on the same investment. You do not start your investment over and suddenly take investment from your new base. Every 100% dollar efficiency is another 100% return on your investment of $100. There are no diminishing returns. you’re putting the same amount in you’re getting the same amount out, there is no proportional difference.
You’re basically trying to calculate this like elemental damage resistence in poe, Where you are at 75% so you now invest more and further reduce the damage you take. Elemental resist actually has increasing returns on investment each point is worth more than the previous point (in positive values, at negative values it actually has diminishing returns up til 0, so after kitava 2, the 31st point of lightning res is worth less than the 30th.) This is not diminishing returns this is constant returns. Diminishing returns would be say if at 100% inc mana efficiency you got +100 mana, at 200% inc mana efficiency you got +190 mana, at 300% you got +270 mana. etc. Those would be diminishing returns, you’d put in the same 100% investment but get 90/80/70… back.

Ok so, As to your final point. Technically this is correct ISH in the sense that you can have ability specific mana efficiency and generic, and things do have a fixed cost. However That is simply a breakpoint in the formula. Thinking of it as an asymptote is… teeeechnically correct but it’s misleading when you say that. It’s more like the 0 point in elemental resist in poe, where the 2nd derivative changes its sign and the returns increase at an increasing rate, excepts in this case instead of flipping from diminishing to increasing returns we hit a hard wall where we have 0 returns, but up til that breakpoint we were getting constant returns, each point of efficiency was as valuable as the point before it. Essentially what you’re doing at that point is you’ve reduced the cost to 0 and hit infinity, thus no matter how much further you invest you’re just still at infinity. Though they might hard code it at 1. (still however you’d actually still be getting constant returns as each % would give you the same return at that point 0, just different returns than before so it’s more of a break point).

Anyways I’m gonna leave it at this, Mana efficiency is basically always worth as much as the previous point. It can only have diminishing returns when compared to other mods, when you get down to getting the spell to 2 mana, is it worth the investment to get it to 1 mana for double the number of casts, or could you use your resources (mod slots talents etc) to increase the damage by more than double (or sufficiently to increase TTK to the point where double casts becomes irrelevant). Which is a different discussion.

1 Like
10

Every 100% efficiency does give you an additional amount of casts. Going from 100% to 200% reduces the cost from 1/2 to 1/3 meaning you go from being able to do two full sets of casts to three sets of casts. And how much is three more than two? Yup, one full set.

11

I understand that you generally always get 1 full additional “mana bar” worth of casts per 100% efficiency. In that regard there are no diminishing returns. I’m just saying the difference between going from 10 → 20 casts is proportionally a way bigger deal than going from 90 → 100.

It’s clear that as you get more and more Mana Efficiency you would always just be approaching but never reaching 0 mana cost. And assuming Mana is always rounded to a whole number, the graph is a step function. Like once you get to the point where a spell costs 1 mana you’re at the point of severely diminishing returns. Adding another 100% efficiency is not going to give you another mana bar of casts at that point. The cost will probably get rounded back up to 1 until the point where you’ve put in several thousand % more efficiency and it rounds down to 0 (if their code even allows that) at which point you’ve broken it and your casts go up to infinity.

But this is all very pedantic at this point. The original presenting complaint wasn’t “I don’t understand mana efficiency” it was mainly “we have no concrete information on how all the inconsistently-worded mana modifiers in the game interact with each other.”

And I’m sure we could easily guess the formula for the mana cost of a skill, it’s probably the same as the formula for damage. The problem is that they have all sorts of documentation for damage numbers but not mana. Not only is the damage formula published, literally every node in the skill tree that says “MORE” damage also has a subtext that says “multiplicative with other damage modifiers” but there’s no documentation whatsoever for mana mods.

We can assume but it would be nice to not have to.

And running into a situation like “mana drain” which is apparently its own thing that’s completely unaffected by mana efficiency for some reason, makes it clear that we don’t have certain (probably important) information.

12

It’s not.

It does though, see my comment above. Mana costs are displayed as round numbers but they aren’t used that way. But you generally can’t get that much efficiency so it’s a moot point.

You say that like it’s a bad thing.

“We” mostly do.

I respectfully point you towards my comment here, though I’ll tweak it a bit to be more encompassing, though I’m assuming that all more & less modifiers are multiplicative rather than additive but I also don’t think there’s many (any?) skills that have multiple more/less mana cost modifiers.

13

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.