Resilience Scaling <Math included>

85 Tauren Warrior
1960
If a class is fine at a certain amount of resilience (and I am stating that this point will exist for every class if damage is to be brought to a reasonable level) and the same percentage of a higher Ilevel item's budget is used for resilience, eventually replacing some of the PVP gear with PVE gear will become the clearly superior choice. Unless that same chunk of resilience has a higher value at this point, high enough to outweigh the extra secondary stats that you will always need (extra damage, healing, regen) versus the survivability you do not necessarily need.


Higher iLevel would mean that incoming damage is higher and subsequently the point would be raised. This would necessitate the extra resilience on gear. Sure, it might become optimal for stomping teams with weaker gear to sacrifice some resilience but that is hardly the situation of concern.
Reply Quote
85 Goblin Warlock
6425
01/14/2011 11:46 PMPosted by Estrada


Armor doesn't, it simply doesn't have the increasing returns that resilience has. Armor is perfectly linear, just like strength is.


"Armor" isn't a function, it's only one variable. Graph armor rating vs damage reduction and what you get is a graph of diminishing returns, a logarithmic function. Graph armor rating vs life expectancy and you get an evenly sloping line. There is no need for you to "correct" what I wrote, because it was correct to begin with. Armor does have diminishing returns. Resilience needs to have the same or similar behavior.


I still believe if Resilience had a DR function like Armor then PvE gear would surpase PvP gear relatively fast..
Reply Quote
85 Goblin Warlock
6425


I think a Hyperbola is a function of y=x^2 etc.... at least thats' what I remember of it. I am probably slightly wrong it has been quite a long time since I had that sort of math, and I was only a business major not a "super math dood"

A hyperbola usually has the form 1/x. or generally a/(b+cx) where a,b,c are constants.


Yeah I might be thinking of a parabola
Reply Quote
85 Tauren Warrior
1960



Armor doesn't, it simply doesn't have the increasing returns that resilience has. Armor is perfectly linear, just like strength is.


"Armor" isn't a function, it's only one variable. Graph armor rating vs damage reduction and what you get is a graph of diminishing returns, a logarithmic function. Graph armor rating vs life expectancy and you get an evenly sloping line. There is no need for you to "correct" what I wrote, because it was correct to begin with. Armor does have diminishing returns. Resilience needs to have the same or similar behavior.


The returns don't diminish though. If you go the route of saying armor has DR than all linear stats have DR, 1 Str provides less damage percentage at 20000 Str than it does at 2000 Str. The idea of diminishing returns is that you get less benefit as you increase how much you have. You get the same benefit from armor regardless of your current armor.
Edited by Bredman on 1/15/2011 12:01 AM PST
Reply Quote
85 Worgen Druid
3515


"Armor" isn't a function, it's only one variable. Graph armor rating vs damage reduction and what you get is a graph of diminishing returns, a logarithmic function. Graph armor rating vs life expectancy and you get an evenly sloping line. There is no need for you to "correct" what I wrote, because it was correct to begin with. Armor does have diminishing returns. Resilience needs to have the same or similar behavior.


I still believe if Resilience had a DR function like Armor then PvE gear would surpase PvP gear relatively fast..


It could, but that's not a given. The location of the cap and the relative value between resilience and the best offensive stat that's not available on PvP gear is going to determine in the long run whether and when that might occur. it's entire possible for PvP to remain superior forever IF the cap is difficult to reach and IF one offensive stat is better than the others (and that stat is available on PvP gear).
Reply Quote
85 Worgen Druid
3515



"Armor" isn't a function, it's only one variable. Graph armor rating vs damage reduction and what you get is a graph of diminishing returns, a logarithmic function. Graph armor rating vs life expectancy and you get an evenly sloping line. There is no need for you to "correct" what I wrote, because it was correct to begin with. Armor does have diminishing returns. Resilience needs to have the same or similar behavior.


The returns don't diminish though. If you go the route of saying armor has DR than all linear stats have DR, 1 Str provides less damage percentage at 20000 Str than it does at 2000 Str. The idea of diminishing returns is that you get less benefit as you increase how much you have. You get the same benefit from armor regardless of your current armor.


No, you're mixing together relative and absolute terms.

Strength gives the same absolute benefit at all levels of strength, assuming all other things are equal. If one point of strength is going to provide 10 DPS, then two points of strength will provide 20 DPS, and so on until infinity. The slope will remain constant (in this case, 10) forever.

Armor value literally provides less absolute benefit as it increases. Graph armor vs damage reduction, and it won't be a line with a constant slope. The slope will actually decrease as X approaches infinity, possibly becoming zero at some finite value of X.

What you're thinking of (I hope) is some thing more like Armor value vs Survival time in seconds. In that case, armor provides a linear benefit. So it is possible to say that armor provides a linear benefit (it depends what you define as the intended benefit) but it ALSO can be said to provide diminishing benefits (again, depending on what you define as the intended benefit).

Hence, your statement that "armor is linear" can be true, but my statement that "armor is dimishing" is not false. AND armor is not linear in the same sense that Strength is linear. Armor is linear in a relative sense and diminishing in an absolute sense, Strength is linear in an absolute sense and diminishing in a relative sense. Resilience is linear in an absolute sense, but geometic in a relative sense.

So no matter what, if you do an apples to apples comparison (relative vs relative or absolute vs absolute) all three of those functions will have a different shape from the others.

The bottom line, once again, is that resilience needs to have the same general behavior that armor rating has, for the same reason that armor rating was designed to work the way that it does: Otherwise you either hit a cap where the value of additional rating is zero, or it scales to the point where players become impossible to kill.
Edited by Estrada on 1/15/2011 12:17 AM PST
Reply Quote
85 Tauren Warrior
1960
The diminishing benefit of armor for you is a number on your character sheet. They could make a number out of strength, as I did, and put it on your character sheet to have diminishing returns. The realized benefits of both Armor and Strength are non-diminishing.

Your use of relative and absolute doesn't quite make sense. The benefit of Strength is constant relative to all your other DPS stats in the same way that the benefit of Armor is constant relative to your other defensive stats. What defines absolute gain?
Edited by Bredman on 1/15/2011 12:30 AM PST
Reply Quote
1 Gnome Warlock
0
Since when were the forums constructive this thread needs more trolling.
Reply Quote
90 Goblin Priest
11275
If I could get 30% damage reduction with a reasonable amount of resil.. then it started to DR .. and past 40% it was even worse .. then I would drop pvp gear left and right past 30% and just stack PvE gear.

At some point resilience won't be providing any real value per point, and then PVE gear will always be optimal in it's place because you can use it to "dodge" some resilience.

I still believe if Resilience had a DR function like Armor then PvE gear would surpase PvP gear relatively fast..


All of these quotes represent a misunderstanding of the meaning of "diminishing returns"

Diminishing returns does NOT mean that "at some point, adding 100 resilience rating will not yield any return"

Diminishing returns means "adding 100 resilience rating will always provide the same return, whether you're going from 0 to 100, or 1500 to 1600 rating"

Without diminishing returns, the statement would be "adding 100 resilience rating will provide better returns, as your baseline amount of resilience increases"
Reply Quote
100 Human Paladin
14550
Does this mean resilience is inherently more valuable to tanks with our 10% damage reduction built in for being a tank? Is this calculated multiplicatively or additively with resilience?
Reply Quote
90 Goblin Priest
11275
Does this mean resilience is inherently more valuable to tanks with our 10% damage reduction built in for being a tank? Is this calculated multiplicatively or additively with resilience?


Damage reduction effects (and most other numbers in WoW) stack multiplicatively.

I'm not sure how to answer your question of resilience being more valuable to a tank, though - what are you comparing it with? I mean, if resilience worked against PvE bosses, tanks would literally wear nothing else, at the expense of everything else.
Reply Quote
100 Human Paladin
14550
01/15/2011 2:35 AMPosted by Mystine
I'm not sure how to answer your question of resilience being more valuable to a tank, though - what are you comparing it with? I mean, if resilience worked against PvE bosses, tanks would literally wear nothing else, at the expense of everything else.


Thanks. I'm comparing it in PvP.

Basically I want to last really long in battlegrounds and be a harrasser to the other team. I already do that somewhat with 1390 resilience, but I want to wear as much as I can get I think. I remember being a key in winning that new BG, I forgot its name, when I just would NOT die, preventing the horde from capping for long enough that we passed them in points just before the end. I was getting some heals, but I'd always last until the healers rezzed and healed me to full again. I didn't kill anyone but I didn't let them cap either.
Reply Quote
90 Goblin Priest
11275
In terms of extending your Time To Live, Resilience is absolutely the best stat to do it, until you hit the cap.

EDIT: Relative to anything else you can get, such as STA or Prot Mastery or avoidance.
Edited by Mystine on 1/15/2011 9:42 AM PST
Reply Quote
85 Tauren Paladin
4725
01/14/2011 7:43 PMPosted by Bredman
I've already mentioned that Resilience is exponential and they need to reduce it to a linear stat.

Armor is linearized, the formula in the OP is the armor formula from Warcraft 3; replace the 0.06 with 1/26070 and you get WoW's level 85 armor formula.

There is no diminishing of armor, it's linear up to 300% survivability increase (75% damage reduction) and then caps.


The only problem is you can't make a stat like Resilience (-%damage taken) linear, because it will always scale to -100% damage taken, and the close you get to that point the more each point of reslience is worth due to it affecting the remaining portion of damage taken. IE if you have -50% damage through resilience, each point is worth double whats it's worth a zero, or at -75% damage reduction each point is worth quadruple what it's worth at zero.

The only way to get around that would be to make resilience reduce a flat amount of damage (IE -1800 per attack), but that would unfairly punish fast-attacking small damage classes like Affliction Warlocks and Rogues, and be meaningless to burst classes like Mages and Destro Locks.
Reply Quote
90 Goblin Priest
11275
01/15/2011 9:51 AMPosted by Gloriousmoo
The only problem is you can't make a stat like Resilience (-%damage taken) linear, because it will always scale to -100% damage taken, and the close you get to that point the more each point of reslience is worth due to it affecting the remaining portion of damage taken. IE if you have -50% damage through resilience, each point is worth double whats it's worth a zero, or at -75% damage reduction each point is worth quadruple what it's worth at zero.


What the hell? That's precisely how resilience already works, and what you're describing is exponential scaling, not linear scaling.
Reply Quote
85 Orc Warrior
2290
I'm really not sure why resilience scaling is a problem for anyone.

Everyone can stack resilience, as long as the stat can benefit everyone what is the problem with it scaling well?

If it gets to the point were no one dies because resilience is to high, i'm sure it'll need to be adjusted. In fact i'm sure it will need to get some kind of diminishing returns on it eventually.

My point however is, having a stat that scales to well that only benefits certain classes (like ArPen did) it's understandable why it would cause some upset.

Resilience effects everyone, it is THE PvP stat. It is meant to keep PvE gear out of PvP. It HAS to been good, it HAS to be better than damaging stats, because when it isn't as good as damaging stats, people wear PvE gear.

They're taking such efforts to keep PvP gear out of PvE content, it only makes good sence to do the same in reverse.
Reply Quote
01/15/2011 10:05 AMPosted by Mystine
What the hell? That's precisely how resilience already works, and what you're describing is exponential scaling, not linear scaling.

I think this matter is not as clear as people keep saying it is.

In terms of absolute damage taken versus absolute resilience level, the scaling is clearly linear. If an attack does X damage against 0 resilience, it does X*(1-R) = X - XR damage against R% resilience. As a matter of basic and unarguable 8th grade algebra, this is a linear equation in all variables. It's linear in damage. It's linear in resilience %. It's linear in resilience rating, since % is in turn linear against rating. There is no exponential term of any kind. To say that absolute damage scales exponentially with absolute resilience is simply wrong.

In terms of relative damage taken versus relative resilience values, the scaling is also linear. If I am taking Y(R) damage at at R% resilience, and I move to R+dR, then dD = Y(R) - Y(R+dR) = Y(R) - [ Y(R) * 1/(1-R) * (1-(R+dR)) ], which after some algebra is equal to [Y(R)/(1-R)] * dR, which is linear in dR. So dD is linear in dR; relative resilience is also linear with relative damage.

The only time you get nonlinear scaling is when you compare absolute damage to relative resilience, or relative damage to absolute resilience. But neither of these comparisons is valid. In fact, it is only absolute damage versus absolute resilience that matters in PvP, because only absolutely reducing your life to zero can kill you, and healers must heal absolute damage done, not relative damage done.
Edited by Venificus on 1/15/2011 10:24 AM PST
Reply Quote
85 Tauren Paladin
4725
The only problem is you can't make a stat like Resilience (-%damage taken) linear, because it will always scale to -100% damage taken, and the close you get to that point the more each point of reslience is worth due to it affecting the remaining portion of damage taken. IE if you have -50% damage through resilience, each point is worth double whats it's worth a zero, or at -75% damage reduction each point is worth quadruple what it's worth at zero.


What the hell? That's precisely how resilience already works, and what you're describing is exponential scaling, not linear scaling.


Edited by Gloriousmoo on 1/15/2011 10:28 AM PST
Reply Quote

Please report any Code of Conduct violations, including:

Threats of violence. We take these seriously and will alert the proper authorities.

Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.

Harassing or discriminatory language. This will not be tolerated.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)

Reported!

[Close]