5.4 PVP Stat Guide

(Locked)

100 Night Elf Druid
11965
at lower levels, the only difference is how many resil points give you 1% damage reduction.

i am pretty sure that now with 40% base resil added in, resil follows a non-linear return (not just nonlinear, but something with an increasing derivative) at any level.

for example, look at this druid's resil percentage.


resil still works the same way it does where it's less beneficial after a certain point because it has DRs

here's some math i did in literally 30 seconds

your druid: 41.43% damage reduction, 780 resil = .053115% damage reduction per point of resilience

my priest: http://us.battle.net/wow/en/character/sargeras/Snipemasterx/advanced
36.54% damage reduction, 663 resil = .055113% damage reduction per point of resilience

my mage: http://us.battle.net/wow/en/character/sargeras/Anatomyxxmd/advanced
28.10% damage reduction, 481 resil = .058420% damage reduction per point of resilience

it doesn't take a genius to figure out from those numbers alone that resilience gets worse as the number gets higher, hence why stamina is better after a certain point because you'll effectively gain more survivability by having a good mix of the two rather than gemming purely in one or the other

edit: First post I used level 90 numbers.
With zero resil, your EH is 1/(1-.4)=1.6667 times what it would be in PvE
If you have extra 41.43% more resil from 780, then your EH is 1/(1-.4-.4143)=5.11 times the PvE number. 5.11/1.667=3.06, a 206.6% improvement. 206.6/780 = .265% per point of resil.
Using 28.1% and 481 resil. 1/(1-.4-.281)=3.135. 3.135/1.66 = 1.881, a 88.1% improvement. 88.1%/481 = 0.183% of TTL (on average) from each point of resil.

.264 is much larger than .188. Those are called increasing returns.
Edited by Erdluf on 11/27/2012 11:23 AM PST
70 Night Elf Druid
5330


resil still works the same way it does where it's less beneficial after a certain point because it has DRs

here's some math i did in literally 30 seconds

your druid: 41.43% damage reduction, 780 resil = .053115% damage reduction per point of resilience

my priest: http://us.battle.net/wow/en/character/sargeras/Snipemasterx/advanced
36.54% damage reduction, 663 resil = .055113% damage reduction per point of resilience

my mage: http://us.battle.net/wow/en/character/sargeras/Anatomyxxmd/advanced
28.10% damage reduction, 481 resil = .058420% damage reduction per point of resilience

it doesn't take a genius to figure out from those numbers alone that resilience gets worse as the number gets higher, hence why stamina is better after a certain point because you'll effectively gain more survivability by having a good mix of the two rather than gemming purely in one or the other

edit: First post I used level 90 numbers.
With zero resil, your EH is 1/(1-.4)=1.6667 times what it would be in PvE
If you have extra 41.43% more resil from 780, then your EH is 1/(1-.4-.4143)=5.11 times the PvE number. 5.11/1.667=3.06, a 206.6% improvement. 206.6/780 = .265% per point of resil.
Using 28.1% and 481 resil. 1/(1-.4-.281)=3.135. 3.135/1.66 = 1.881, a 88.1% improvement. 88.1%/481 = 0.183% of TTL (on average) from each point of resil.

.264 is much larger than .188. Those are called increasing returns.


this guy knows what he is talking about.
90 Pandaren Monk
7880
There seems to be some flaws in the resilience theory of exponential returns. Maybe I'm wrong but in both instances the damage mitigated is the same amount. 1% damage mitigation = 1 less actual damage taken. The only change is the difference in the percentage between base damage dealt and mitigated damage dealt. This is where the error comes in. In the first instance you are comparing to unmitigated actual base damage but in the second instance you are comparing previous mitigated damage to new mitigated damage. If you use actual base damage both times as you should you'll see the mitigated percentage difference stays the same and matches the actual damage numbers. Resilience scales almost linearly but if anything you could say it has diminishing returns until you hit breakpoints in the incoming damage where the linear scaling produces a better end result. The problem is you can't effectively determine every players incoming damage.

Example:
Getting hit for 100 dam. Reduced by 1% resilience we are now getting hit for 99 damage which equates to 1 actual less damage taken. It now requires the enemy one extra swing to kill us because he will hit us for 99 and the next swing will finish our remaining 1 health. So that 1% bought us 1GCD in time to live.

Now we have 89% resilience and we increase it to 90%, still a 1% increase. Here again we are getting hit for 1 less actual damage for the same 1% increase. If we are getting hit for 11 damage at 89% resilience it will take 10 swings to kill us. 9 swings will get us to 99 and the 10th kills us. Now if we increase that to 90%, still a 1% increase, it STILL takes 10 swings to kill us. We are getting hit for 10 dam per swing, 10 swings to hit us for 100 damage then we die. So, in this example our 1% increase in resilience didn't do anything for us in effective time to live. We still died in the same number of hits.

Now let us move passed a breakpoint in the dam. Now we have 90% resilience. It takes 10 hits to kill us. We increase resilience 1% and are now at 91%. We are still increasing by 1% resilience and receive only 1 less actual damage as in both previous examples. However, we are now increasing time to live by 2 GCD again for the same 1% increase in resilience. Getting hit for 9 damage 10 times gives us 90, the 11th swing gives us 99 dam (a 1GCD increase), and finally the 12th swing kills us for our 2nd GCD increase in time to live. So you say, well that 1% gave us a 2 GCD increase, that's an exponential increase over the first example and you are right but only if you look at this example versus the first one. If you compare all the examples the scaling is linear overall but we had diminishing returns between breakpoints in incoming damage.

We had 3 examples that all increased resilience by 1% which should equate to a 3 GCD increase in time to live.
Example 1 gave us a GCD, Example 2 gave us none, and Example 3 gave us 2 for a total of 3 GCD's as expected.

As I said previously the problem lies in determining the breakpoints in incoming damage. Considering no two people are likely to have the exact same incoming damage, a 1% increase may or may not increase your time to live.

The other issue is that we are only really examining both of the far ends of the damage scale which is why I say I could be wrong. I haven't done the math for the entire scale to determine if each 1% of the scale ultimately adds up to 100 GCD's in time to live for the full 100% resilience. My best guestimate at the moment is that the breakpoints are closer together and more frequent at the ends of scale with more of them being at the higher end, 90%-100% resilience.

If I get time, I'll do the math for the whole scale and update this or if someone can tell me why I'm wrong, have at it. Also, I didn't read the whole thread so this is just in relation to the OP's take on his example of resilience scaling, my thoughts on why it's wrong and not an exact explanation of how resilience scales in game.
Edited by Blåckjåck on 12/10/2012 3:30 AM PST
100 Night Elf Druid
11965
But we don't start at zero resil, we start at 40%, when naked.

If I have no gear, your 100 hits me for 60. If I have 10% resil, your 100 hits me for 50. If I had 600 health, It take 12 hits to kill me instead of just 10, two extra hits, a 20% improvement.

If I add another 10% resil, you hit me for 40. It takes 15 hits to kill me, three more hits, which is a 25% improvement over 12.

Keep going, and each point of resil is worth more (adds more hits-to-kill) than the previous one (until, eventually, I never take damage and live forever).
90 Pandaren Monk
7880
Erdluf
90 Night Elf Druid
Savvy
8210

But we don't start at zero resil, we start at 40%, when naked.

If I have no gear, your 100 hits me for 60. If I have 10% resil, your 100 hits me for 50. If I had 600 health, It take 12 hits to kill me instead of just 10, two extra hits, a 20% improvement.

If I add another 10% resil, you hit me for 40. It takes 15 hits to kill me, three more hits, which is a 25% improvement over 12.

Keep going, and each point of resil is worth more (adds more hits-to-kill) than the previous one (until, eventually, I never take damage and liv


You're right and you're wrong. Each point of resilience is worth more hits-to-kill than the previous point but each point does not always equate to more actual hits-to-kill until you cross a breakpoint in incoming damage.

If you look at your example and then change it to 39 damage, a 1% increase, you now require 16 hits to be killed. Technically, its 15.38 but that .38 equates to an additional swing in order to kill us. If you add another 1% so we are now taking 38 damage we still only require 16 hits to be killed. Technically, 15.79. Again, that .79 equates to 1 additional swing. So the 1% increase 61% to 62% resilience had ZERO increase in time to live. It isn't until you reach the breakpoint in damage received that you will see the scaling increase. To say every point increases time to live is incorrect but yes every point is cumulatively worth more than the previous. This distinction is being over looked from what I can gather.
Edited by Blåckjåck on 12/10/2012 5:38 AM PST
100 Night Elf Druid
11965
For those breakpoints to be meaningful (something you can plan around), you have to know exactly how big incoming hits (and heals) are going to be.

I'm not smart enough to figure that out, so for theorycraft I pretend that incoming damage is somewhat random, at least until I get down to execute range.
90 Orc Warrior
10110
A simple question that I'm too dumb to figure out on my own.

I have a resto shaman at level 90. he has 386K HP unbuffed, and 8173 resilience (+23.28, on top of the baseline 40, for 63.28%).

I have mining, which gives me toughness, which gives me 480 stamina, or about 5K HP; even more when you factor in HP boosting abilities.

Should I drop mining for blacksmithing? The two gem sockets get me 2 resilience gems, or 640 resilience, which is about 1.5% more damage reduction.

Is it worth the switch?
100 Night Elf Druid
11965
5k HP is less than 2% reduced health.

Your current damage taken is <40%. Droping that by another 1.5% is more than a 3.5% improvement.

Seems like a pretty clear gain.
MVP - World of Warcraft
87 Tauren Paladin
10570
12/10/2012 03:01 AMPosted by Blåckjåck
There seems to be some flaws in the resilience theory of exponential returns. Maybe I'm wrong but in both instances the damage mitigated is the same amount. 1% damage mitigation = 1 less actual damage taken. The only change is the difference in the percentage between base damage dealt and mitigated damage dealt. This is where the error comes in. In the first instance you are comparing to unmitigated actual base damage but in the second instance you are comparing previous mitigated damage to new mitigated damage. If you use actual base damage both times as you should you'll see the mitigated percentage difference stays the same and matches the actual damage numbers.

The change in frame of reference was intentional and was the entire purpose of the example, although the absolute value of the damage mitigation is unchanged, the relative value is increased significantly. To take the example even further, if you were already at 98% damage reduction and added 1% it would cut incoming damage in half relative to what it was at 98%, and it would take twice as much damage to kill you as it did before. Where as if you had 0% and you added that same 1% damage reduction it would have a relatively tiny effect on your survivability.

The absolute value isn't changing, that 1% is still mitigating the same amount of absolute damage, but the relative difference in the damage your character is actually taking increases significantly at higher damage reduction percentages. Also, if you go from 99% to 100% you would become immortal, which is clearly better than what all of us are getting from 1% now ;)

Getting hit for 100 dam. Reduced by 1% resilience we are now getting hit for 99 damage which equates to 1 actual less damage taken. It now requires the enemy one extra swing to kill us because he will hit us for 99 and the next swing will finish our remaining 1 health. So that 1% bought us 1GCD in time to live.

As Erdluf mentioned things are never this cut and dry in WoW, if every attack hit for the same amount this would be an effective way to calculate time to live, but attacks in WoW hit for a wide range of damages from the very small to the very large and everything in between. As a result the best way to gauge actual survivability is to simply measure your effective health, because you cant predict what kind of damage you will be receiving and higher effective health is always better.

A simple question that I'm too dumb to figure out on my own.

I have a resto shaman at level 90. he has 386K HP unbuffed, and 8173 resilience (+23.28, on top of the baseline 40, for 63.28%).

I have mining, which gives me toughness, which gives me 480 stamina, or about 5K HP; even more when you factor in HP boosting abilities.

Should I drop mining for blacksmithing? The two gem sockets get me 2 resilience gems, or 640 resilience, which is about 1.5% more damage reduction.

Is it worth the switch?

Yes you should switch if you are willing to invest the money, here are your stats for both cases:

Current: 386k health, 8173 resil, 63.28% reduction. (Effective Health = 1051k)
Proposed: 381k health, 8813 resil, 64.85% reduction. (Effective Health = 1084k)

So you are looking at a 33k increase in effective health if you drop mining, pick up smithing, and throw resil gems in those sockets. I hope this helps you make a more informed decision.
Edited by Eldacar on 12/13/2012 7:56 PM PST
90 Night Elf Monk
4910
Sweet mother of God, what a piece of platinum we have here! Thank you!
70 Orc Hunter
1500
to many big words my @@###@%@
hurts
90 Night Elf Hunter
6250
WOW, my theory of school kids in bgs is sounding about right (now) lol.
100 Tauren Shaman
13280
Best.
Guide.
Ever.

Also, just wanted to point out that we healers also have "PvP Resilience" to cut through; namely: Battle Fatigue. We would need (roughly) 43% PvP Power (11,395 PvP Power Rating) to ignore Battle Fatigue, which is a wildly unattainable amount.
Edited by Gijira on 12/28/2012 9:19 AM PST
100 Night Elf Druid
11965
Best.
Guide.
Ever.

Also, just wanted to point out that we healers also have "PvP Resilience" to cut through; namely: Battle Fatigue. We would need (roughly) 43% PvP Power (11,395 PvP Power Rating) to ignore Battle Fatigue, which is a wildly unattainable amount.

True but that doesn't scale, and is fairly minor compared to Resil. As a Balance Druid in my PvP gear, but in a solo PvE situation, My Rejuv is about 12% stronger than my Moonfire (total heal/damage from a single cast). In a PvP situation against another Moonkin with my gear, Rejuv is about 114% stronger.

PvE MF = 100
PvP MF = 100 * .85 (Moonkin target) * .36 (Resil) * 1.23 (Power) ~= 37

PvE Rj = 112
PvP Rj = 112 * .7 (BF) ~= 79 (more than double 37)

Even if I included a mortal-strike type debuf, and actually had a PvP weapon, Rj scales much better into PvP than my attacks do.
Edited by Erdluf on 12/28/2012 9:11 PM PST
90 Human Priest
15890
Any thoughts on the relative value of intellect and pvp power for healers? I need to find someone who has done the math for it. However, gemming for a stat that I automatically don't get as much out of as a dps play just seems repellent to me.
90 Human Warrior
11790
I love this guide. I am a resil nut now. I do have a question, and apologize if it's been asked; although I see the evidence, sites like "Noxxic.com" state to gem for pvp power. I don't but I had a rather comical death knight the other day who challenged me to a duel who had all pvp power and was a blood dk and said he'd crush me blah blah; well, we talked and he ended up being a nice guy!! I discussed the finer points of resil yet he continued to say he "read all the parses" and "logs" and stated that PvP power was clearly a winner over stacking resil. However; sites like askmrrobot.com state to stack resil, and I tend to agree a bit more with that.

Is there any truth to what this guy was saying about parses and such? I'll admit I can follow the math to a certain point, however my abacus broke and my fingers just don't cut it...

Thanks,

Arës
100 Night Elf Druid
11965
In a balanced fight, resil is stronger. It reduces damage taken by more than power increases damage or healing delivered. This would be true even if all heals were affected by power, but unless you are a resto spec, most heals ignore power.

However there are a lot of unbalanced fights in PvP. When you've got three on two, the three will win faster if they happen to have stacked power. That can be very important, because the advantage is often temporary (reinforcements will arrive at the farm, from LM or from the gy, CC's will expire, others will be placed). Of course in this situation, the defenders still wanted resil.

Likewise, if you usually don't get focused early in the fight, more power could make sense.

Finally, if you stack power, you might lose some fights that you could have won, but win or lose, fights will tend to be over more quickly. That means you can do more fights, and that may increase how quickly you can obtain gear.
This topic is locked.

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]