Hitpoints versus PRR: Which helps your character more?

[sirgog]

I've run some numbers using the forums Eladrin gave. Here's a very quick rule of thumb you can use when planning gearsets.

Not going to go into the maths here, but you can highlight the black paragraph below for info on my method.

[warning: calculus]Basically I'm considering the number of 1 damage hits required to kill a player (assuming no rounding, so with 1 PRR each hit does 0.9935 damage) as a two-variable function F of PRR and HP, and differentiating this function with respect to PRR and HP (to get dF/dPRR and dF/dHP) and looking at the ratio of these two at various plausible values. I'm then finding a linear approximation to segments of the function Y=(dF/dPRR)/(dF/dHP) to make guidelines that can be easily understood.
[/calculus]


600hp character (what I predict will be the standard on a 'squishy' melee, e.g. Rogue, at 25)

Your first 70 points of PRR are worth 3.5hp each. Your next 70 points are worth 2.5hp each, and after that additional PRR is worth 1.5 hp each.

So for raw survivability, a 600hp 50 PRR rogue should be very close to a 635hp 40 PRR rogue. And a 600hp 150 PRR rogue will be close to a 640hp 130 PRR rogue.


800hp character

As above, but add 1hp in 'value' to each PRR.


1200hp tank

The first 100 points of PRR are worth about 6.5hp each. From 100 to 150, each point of PRR is worth about 4hp, and from 150 to 200, 3hp per point.


Whilst these are rough rules of thumb, they should prove useful in eyeball comparisons of possible gearsets and enhancements.

[Osma77]

Too tired to check math, but thanks for working that out.

[DarkSable]

[warning: calculus]Basically I'm considering the number of 1 damage hits required to kill a player (assuming no rounding, so with 1 PRR each hit does 0.9935 damage) as a two-variable function F of PRR and HP, and differentiating this function with respect to PRR and HP (to get dF/dPRR and dF/dHP) and looking at the ratio of these two at various plausible values. I'm then finding a linear approximation to segments of the function Y=(dF/dPRR)/(dF/dHP) to make guidelines that can be easily understood.[/calculus]

Looking at your math, it seems solid, but here's the thing. The optimal solution is not to max one or the other (as is so often the case in this game.) For this, the optimal solution is to find the point where both stats are maximized, relative to each other. (It's late and I can't remember the term, but for the layman, it's the reason that four times six is less than five times five. Or, in the min-max example, 2 times 25 is WAY less than 13 times 15.)

600hp character (what I predict will be the standard on a 'squishy' melee, e.g. Rogue, at 25)

Hmm... I think you're vastly overestimating how much hp those Epic levels are going to give us - because remember, you have to have a 25 base con to get Epic Toughness; something that only tanks are likely to go for, especially with all the other goodies.

[Xyfiel]

How does it factor in all incoming damage and effects?

Is PRR going to save me from 600+ point disintegrates?

It is only one tool in the toolbox. Hps is a more universal tool.

One should not need a refresher on Calculus to understand if getting more AC, DR, PRR, Dodge, or Hps is the best option.

[Scraap]

How does it factor in all incoming damage and effects?

Tested that. Full rider damage. Purely first number from the weapon mitigation.

Dodge on the other hand: pure removal.

[Malshier]

One should not need a refresher on Calculus to understand if getting more AC, DR, PRR, Dodge, or Hps is the best option.

x1000

[sirgog]

Looking at your math, it seems solid, but here's the thing. The optimal solution is not to max one or the other (as is so often the case in this game.) For this, the optimal solution is to find the point where both stats are maximized, relative to each other. (It's late and I can't remember the term, but for the layman, it's the reason that four times six is less than five times five. Or, in the min-max example, 2 times 25 is WAY less than 13 times 15.)


Hmm... I think you're vastly overestimating how much hp those Epic levels are going to give us - because remember, you have to have a 25 base con to get Epic Toughness; something that only tanks are likely to go for, especially with all the other goodies.

I'm not posting all the actual numbers. I didn't go as far as to hit it with Lagrange Multipliers or anything like that though - I'm after approximations, not precision here.

On the epic levels: It's not just levels (which I think do increase the amount of HP Con gives you but would need to check). It's also new gear. Superior False Life. +8 Con. Etc etc etc. But even if you top out at 540 on your Rogue or 670 on your Monk, the 600 approximation won't be too far off.


Finally on 'this system is too complex!' comments - has anyone seen how complex Vanshilar's calculations were to determine proc rates and damage dice on various Shroud proc effects? You don't need to understand them in their entirity to be able to use his summaries either. You don't need to understand standard deviations to use the information that Greater Incineration is better than Disintegration, which is roughly equivalent to Lightning Strike and Corrosive Salt, which are better than Incineration, Cacophony or Magma Surge, all of which are better than Steam or Elemental Mastery. (Hell I don't understand all of his statistics).

[Veriden]

Hmm... I think you're vastly overestimating how much hp those Epic levels are going to give us - because remember, you have to have a 25 base con to get Epic Toughness; something that only tanks are likely to go for, especially with all the other goodies.

Not in the least, keep in mind the rogue is no longer getting 6hp per level + con. He's getting 10hp per level + con and epic destinies gives a nice boost to hp should such a destiny be selected. Fury of the wild is a flat +100 hp at lvl 3. so thats 150hp right there should they choose to work for such.

[donblas]

Looking at your math, it seems solid, but here's the thing. The optimal solution is not to max one or the other (as is so often the case in this game.) For this, the optimal solution is to find the point where both stats are maximized, relative to each other. .

That's not totally true, in that HPs are useful against everything that deals damage, while PRR has a much more limited effect. If the effect on melee damage is similar, then HPs are far superior due to their universal application.

Now dodge type mitigation might be another matter as it prevents the blow (and any special effects) from landing and in some circumstances that will be more useful than more HPs.

[sirgog]

That's not totally true, in that HPs are useful against everything that deals damage, while PRR has a much more limited effect. If the effect on melee damage is similar, then HPs are far superior due to their universal application.

That's true to some extent. However, if you have 600hp and 73 PRR (which is about as resilient to physical attacks as 900hp and 0 PRR), you will need a lot less healing than the 900hp/0 PRR character. So PRR has some advantages over HP too.


(Note my 'rule of thumb' rules break down a little here because they were designed for comparing only small to moderate changes in PRR. Just trust me that 600/73 is close to 900/0).

[ltlFox]

I've run some numbers using the forums Eladrin gave. Here's a very quick rule of thumb you can use when planning gearsets.

Not going to go into the maths here, but you can highlight the black paragraph below for info on my method.

[warning: calculus]Basically I'm considering the number of 1 damage hits required to kill a player (assuming no rounding, so with 1 PRR each hit does 0.9935 damage) as a two-variable function F of PRR and HP, and differentiating this function with respect to PRR and HP (to get dF/dPRR and dF/dHP) and looking at the ratio of these two at various plausible values. I'm then finding a linear approximation to segments of the function Y=(dF/dPRR)/(dF/dHP) to make guidelines that can be easily understood.
[/calculus]


600hp character (what I predict will be the standard on a 'squishy' melee, e.g. Rogue, at 25)

Your first 70 points of PRR are worth 3.5hp each. Your next 70 points are worth 2.5hp each, and after that additional PRR is worth 1.5 hp each.

So for raw survivability, a 600hp 50 PRR rogue should be very close to a 635hp 40 PRR rogue. And a 600hp 150 PRR rogue will be close to a 640hp 130 PRR rogue.


800hp character

As above, but add 1hp in 'value' to each PRR.


1200hp tank

The first 100 points of PRR are worth about 6.5hp each. From 100 to 150, each point of PRR is worth about 4hp, and from 150 to 200, 3hp per point.


Whilst these are rough rules of thumb, they should prove useful in eyeball comparisons of possible gearsets and enhancements.

your results look incorrect
you implement that more PRR - less effective - that is wrong
also PRR can't excess 100

first of all formal itself simpler to understand & use than diveriatives
it is HP/(1-PRR) , ok i could be wrong about what is easier for whom

so when ever some one need to decide wich gear better for him - he could put his pairs of numbers & compare results , & over all it will be more accurate calculation

same for dodge HP/(1-PRR)/(1-DODGE)

[merentha]

I have to get my brother in on this thread...he'd have a ball.

<- This guy hates the maths beyond figuring out the paystub.

But I do appreciate that some of you do and are figuring out PRR for all of us.

[Recared]

That's true to some extent. However, if you have 600hp and 73 PRR (which is about as resilient to physical attacks as 900hp and 0 PRR), you will need a lot less healing than the 900hp/0 PRR character. So PRR has some advantages over HP too.


(Note my 'rule of thumb' rules break down a little here because they were designed for comparing only small to moderate changes in PRR. Just trust me that 600/73 is close to 900/0).

Interesting insight both. Thank you.

[sirgog]

your results look incorrect
you implement that more PRR - less effective - that is wrong
also PRR can't excess 100

first of all formal itself simpler to understand & use than diveriatives
it is HP/(1-PRR) , ok i could be wrong about what is easier for whom

so when ever some one need to decide wich gear better for him - he could put his pairs of numbers & compare results , & over all it will be more accurate calculation

same for dodge HP/(1-PRR)/(1-DODGE)

PRR's formula was confirmed by Eladrin to be:

Damage suffered through PRR of x:
0.35 + 0.65 (0.99^x)

Or in more simple terms, 35% of damage is 'immune' to PRR, and then each point of PRR drops the other damage by 1% (stacking multiplicatively).

[Nyvn]

Looking at your math, it seems solid, but here's the thing. The optimal solution is not to max one or the other (as is so often the case in this game.) For this, the optimal solution is to find the point where both stats are maximized, relative to each other. (It's late and I can't remember the term, but for the layman, it's the reason that four times six is less than five times five. Or, in the min-max example, 2 times 25 is WAY less than 13 times 15.)


Hmm... I think you're vastly overestimating how much hp those Epic levels are going to give us - because remember, you have to have a 25 base con to get Epic Toughness; something that only tanks are likely to go for, especially with all the other goodies.

I agree with you about the opportunity cost, get the best bang for your buck. That's pretty true of the new combat system as a whole though. It holds true for AC, To-Hit, Dodge, PRR, and HP.

[Wraith_Sarevok]

You don't need any math to figure this out... it's common sense.

Every point of damage you don't take is a point of damage that you don't have to heal. So it's like virtual HP.

Obviously, in a game where raid bosses deal upwards of 160+ damage and your average endgame monster deals 80+ damage, a %-based damage resistance is going to be worth MUCH more than measly static HP boosts. I mean why do you think people keep constantly telling me that 9/- DR is worthless?

So high PRR comes 1st. High HP comes 2nd. You want both to be a great tank.

Seriously... calculus? Way to confuse most of the player base. Give the Average Joe some credit. They can understand the idea that percentages are greater than static numbers. I got that idea on the first day that I ever played Diablo 2.

[Angelus_dead]

To an extent you can treat PPR as virtually increasing your hp total: that only applies for the purpose of how much you can survive before needing heals. To correctly decide how valuable they are isn't simple math question of understanding how each curve progresses... it depends on how they act differently against different threats, and how often the game situations involve those threats.

PPR has the advantage of lowering the amount of incoming healing you'll need. Hp's advantages are applying to non-physical damage, and helping you absorb all of a heal without wasting it.

Remember that at level 25, individual heal spells will probably hit for a lot more than they do now (unless the Spell Power nerfed something with Empower Healing... not sure about how that adds up). So having high hp will remain valuable to avoid wasting some of level 25 Mass Heal spells with all the new bonuses.

[Shade]

Or if you actually ran any difficulty new content youd realise:

The only challenge content in the new stuff is places where casters crush you with spells. Melee damage is fairly minimal.
EG: Ring of Fire CR:30 - Ultra easy (99% melee monsters)
Fight to the Finish cr30: - Very hard (many many casters)

PRR has no affect on spells.
HP does.

HP wins as it always has. /Thread.

[ltlFox]

PRR's formula was confirmed by Eladrin to be:

Damage suffered through PRR of x:
0.35 + 0.65 (0.99^x)

Or in more simple terms, 35% of damage is 'immune' to PRR, and then each point of PRR drops the other damage by 1% (stacking multiplicatively).

ok , i'm sorry , i've obviously was incompetent in question

from this formula your results looks right

but still i think it's simpler to put numbers & formula in exel or other software to compare any given numbers , or lets say online stats calculator if some one make one..

[sirgog]

Or if you actually ran any difficulty new content youd realise:

The only challenge content in the new stuff is places where casters crush you with spells. Melee damage is fairly minimal.
EG: Ring of Fire CR:30 - Ultra easy (99% melee monsters)
Fight to the Finish cr30: - Very hard (many many casters)

PRR has no affect on spells.
HP does.

HP wins as it always has. /Thread.

Saves trump HP by a long, long way against spell damage.