Hunter thoughts on the Patch 3.2.2 Armor Penetration nerf
by Eddie Carrington 
Sep 9th 2009 at 9:00AM
I'll be honest with you. I'm not sure why this is turning into such a big deal is to everyone. I'm not saying I don't understand the mechanics of the change. And I do see the impact to us Hunters, especially endgame raiding Marksman Hunters. But really, let's take a step back and look at the reasoning behind the change.
Ghostcrawler said that they implemented this change after noticing everyone was picking up Armor Pen at the expense of their core stats. Really? And somehow we're surprised that when a secondary stat becomes more important than a class' primary stat and it forced a nerf? It seems more of a correction in my opinion.
If you're a raiding Marks or Survival Hunter you should be focusing on picking up as much Agility (Attack Power if you're a Beast Master) as possible. Hence why in Cataclysm we're seeing things like Armor Penetration removed from the itemization.
Think about it this way. Imagine if someone out there found out that if you stacked Crit over Agility you could find 5% gain in DPS. Would that really make sense as a class mechanic? Sure, we'd all probably switch over for that gain, but could we really blame Blizzard for a change because we found a loophole in the game mechanics?
Now we can all debate if a 15% reduction is too much or not. (Finding that sweet spot is always difficult.) We do know that when Armor values were at 100% everyone laughed at stacking Armor Penetration. Even today it's not considered a stat worth considering until you're raiding the high end raids like the hard modes of Trial of the Crusader or at least Ulduar 25 man. But I think we can all agree that when stacking secondary stats outweigh the benefits of stacking our primary ones, then it's a problem that probably should be addressed.
Filed under: Hunter, Patches, Analysis / Opinion, (Hunter) Scattered Shots
Reader Comments (Page 1 of 5)
fernando Sep 9th 2009 9:06AM
i drop armor thinking survival was all about agility... may be i did the wrong thing, because other hunters with Grim Toll make more dps
lasfoola Sep 9th 2009 9:26AM
Ummm.. What?
Knyle2 Sep 9th 2009 11:20AM
I think it makes sense for a "secondary" stat to have more value.
A primary stat ( i dont like that term at all) affects multiple things, therefore is more generalized. so of course a direct stat mod like ArP would yeild higher effect.
Are they going to Nerf mail spell power now and give us more int? or make int a heavier stat because we all stack a "secondary" stat.
P.S.
Agil, stam, con, spirit, str, int, Are base stats.
ap, sp, crit, haste, hit, mp5, are modifiers!
callie510 Sep 14th 2009 4:16PM
>> Are they going to Nerf mail spell power now
>> and give us more int? or make int a heavier
>> stat because we all stack a "secondary" stat.
Poor example. They plan to do precisely this - spellpower is going away entirely in Cataclysm.
They're just balancing. Quit the QQ and change your gems a little.
paul Sep 9th 2009 9:10AM
This is one of the things that always Get me: People complain when things get nerfed (changed) to their class, but if you take a step back then most of the time theres a valed reason and its for the good of the game.
Guess some peple just like to QQ about everything though, even if it does make the game better.
Dzur Sep 9th 2009 11:17AM
Why are you buffing someone who can't spell VALID? This person also fails to make a VALID point.
paul Sep 9th 2009 11:31AM
Well, Im sorry If my spelling isnt purrfect. I deepli apolagise for not looking at my wow.com comments with the same attitude that I would for my D`gree SA's.
And yoor totally write, having one letter in my post makes my whole pount not valed anymore.
Xigageshi Sep 9th 2009 12:36PM
Well I was gonna comment to say i'm right there with ya paul, but after that response to the nitpicker, I simply had to kudo's you! though I suppose, I should have said, "Im write thar wit yous paul" :D
people take the game way too seriously, I think having alts is a good way to prevent this kind of nerf thinking, because whenever blizzard has nerfed anything ever, its to bring it in line with other classes, so you can practically see it as a buff to everyone else, making both terms completely meaningless.
this is blizzard doing what they've always done, balancing. and I must say I have enormous respect for how well they manage it, considering how much of a behemoth this game has become.
Aylia Sep 9th 2009 9:10AM
The "primary" stat is the stat you gain the most benefit from. It doesn't matter if it's ArP or agility.
Nerfing ArP only forces people to waste gold on regemming their gear again (and losing some DPS to boot).
The ArP nerf is just making us all waste gold and perform worse in raids. Hooray.
Gareth Sep 9th 2009 9:40AM
Yep, this is another pointless change, what he's actually saying is he screwed up and managed to make it so that armour penetration was more useful then the stat he wanted people to use. If armour pen was overpowered in game terms then nerf it, but if he just wants people to pick a different stat then why not wait for the next expansion?
Even then I question the logic, at some point armour pen should be hitting diminishing returns, if it isn't (I couldn't be bothered to check through the formulae) then its a total screw up.
So what he is basically saying is, I want everyone to follow the same gearing rules and not to deviate from his simple paint by numbers idea of how to play the game.
Watching the total screwup that was the deathknight class (started overpowered as I thought it would, but needed three patches of nerfs to get it remotely balanced, no wonder they don't want to release another class!) I am seeing where the problems are, gear resets mid expansion? Another great idea...
Knob Sep 9th 2009 10:06AM
The ArP nerf doesn't hit hunters badly in PvE at all. It's at the most a 100-200 DPS loss with the current theoretical maximum being in excess of 10k. Nothing to cry about so just don't.
Texicles Sep 9th 2009 11:45AM
@ Gareth
Actually, ArP does not have diminishing returns (until you hit the hard cap, at which point, you get NO returns). That is part of the reason that ArP is going away for cataclysm. It's hard to balance around because of the math involved.
Armor has diminishing returns. This means that your first point of armor will reduce your incoming damage by a much greater percentage than your 15,000th point of armor. When represented graphically, armor's mitigation is logarithmic.
ArP, is functionally the removal of armor from a target mob. Just as their 15,000th point of armor didn't give them much mitigation (relative to their first point of armor), a small amount of ArP won't give you much gain. However, as you stack more and more ArP, you force the mob closer to the point of 0 armor. As you ignore the last of their armor, you are ignoring that quantum of armor that gives them the most benefit. ArP's value represented graphically, as it is the inverse of armor, is exponential.
This nerf, rebalance, correction or whatever else you wish to call it is not the end of the world... of warcraft, or even the end of ArP stacking. The theorycrafting community at Elitistjerks has already determined the new hard cap for ArP and it is, with trinket procs, still attainable or very nearly so. Has ArP diminished in it's value as a stat? Yes. Will people stop stacking it? Some. Is it still possible to stack it and reap the rewards? Yes.
LilBanshee Sep 9th 2009 2:26PM
The way armor scales is often misunderstood... Sure, the amount it reduces a single hit by diminishes at higher values, but what does not diminish is the effective health increases it gives you. Consider the information at http://www.wowwiki.com/Armor under the heading "The following is the same formula for a lvl 80 tank against a level 83 mob". The way you presented the information, and indeed the way WoWWiki presents it, makes it seem like each successive amove of armor helps you out less than the last. This is a skewed interpretatino of otherwise accurate data.
The problem is that both you and they are comparing the wrong things. What matters is your effective health total scales from armor. For example, at 50k armor you only take 25% of the initial damage, so a tank with 40,000 health therefore has an effective health of 160,000 unmitigates points of physical damage. The formula for that is: MaxHealth * (1 / PercentageDamageTaken) = EffectiveHealth
So, taking the percentages given on WoWWiki, and sticking with my 40,000 health tank example:
at 0 armor: 40000 * (1 / 1.00) = 40000 EffectiveHealth
at 5000 armor: 40000 * (1 / 0.769) = 52000 EffectiveHealth
at 10000 armor: 40000 * (1 / 0.625) = 64000 EffectiveHealth
at 15000 armor: 40000 * (1 / 0.526) = 76000 EffectiveHealth
at 20000 armor: 40000 * (1 / 0.454) = 88000 EffectiveHealth
at 25000 armor: 40000 * (1 / 0.4) = 100000 EffectiveHealth
at 30000 armor: 40000 * (1 / 0.357) = 112000 EffectiveHealth
at 35000 armor: 40000 * (1 / 0.322) = 124000 EffectiveHealth
at 40000 armor: 40000 * (1 / 0.294) = 136000 EffectiveHealth
at 45000 armor: 40000 * (1 / 0.27) = 148000 EffectiveHealth
at 50000 armor: 40000 * (1 / 0.25) = 160000 EffectiveHealth
From the above we can derive this: X armor = 40000 + (2.4 * X) EffectiveHealth. Therefore, armor does NOT have diminishing returns, every point increases your effective health by the exact same amount as the last point (specifically, a linear 2.4 effective health per point of armor), just like how stamina also increases effective health linearly. Also note that the '2.4' value scales up as your stamina scales up. For example X armor = 50000 + (3 * X) if you have 50,000 health. So if you increase both stamina and armor you get geometric growth rates, quite the opposite of diminishing returns!
LilBanshee Sep 9th 2009 3:01PM
And now on to clarify how armor penetration really works... since armor linearly impacts effective health, and armor penetration decreases armor by a linear percentage, armor penetration in effect reduces the opponent's effective health at a linear rate.
Let's stick with our example of a 40,000 health opponent with their armor capped (75% damage reduction). This person has 160,000 effective health.
If you have 400 armor penetration you reduce their armor by 32.5%, so they have 33750 armor now, and 121000 effective health.
If you have 800 armor penetration you reduce their armor by 65%, so they have 17500 armor now, and 82,000 effective health.
The first 400 armor penetration reduced their effective health by 39,000. The second 400 armor penetration also reduced their effective health by the exact same number, 39,000. Linear. No different than how armor values scale linearly, no different from how the majority of stats in the game scale linearly.
What makes armor penetration powerful isn't that it follows some magical exponential growth rate or defies the conventional mechanics of the game by being linear when other stats have diminishing returns. What makes it powerful is simply that the linear rate at which it decreases opponents' effective health is a HUGE linear rate relative to other stats. At the armor pen cap of 1230, an armor-capped opponent's effective health drops to a mere 25% of their original value, thus essentially doing quadruple damage (vs a cloth-wearer at 5,000 armor the difference is much smaller, maybe a quarter more damage). So its extremely powerful against heavily-armored opponents. Compare this to critical strike rating which, at its cap (which, by the way, takes WAY more than 1230 crit rating making it much harder to attain) you at best might get double damage, maybe more with an ideal talent spec like fire mage.
Moral of the story: Armor pen does not scale in some obscure way, the first point and the last point are equally effective. It's just extremely powerful against heavily-armored opponents (like bosses or plate-wearers)
Random Sep 10th 2009 4:30AM
LilBanshee, while your math demonstrates an excellent understanding of the simple functions of armor and ArPen, you've not considered the relative value of each increase.
For instance, in your example, adding 5k armor gives you an additional 12k effective HP. So the first 5k armor you get raises your effective HP from 40k to 52k. This is an increase of roughly 33%. The second 5k armor you get raises your effective HP from 52k to 64k, which is an increase of roughly 23%. The next 5k gives you an increase in effective HP of roughly 18%. And so on, until that final boost from 148k to 160k is only a relative increase of 8%. This is why armor scales logarithmically.
Armor penetration scales the same way, only in reverse. This will be complicated, so stay with me. Let's say that your hypothetical target with 40kHP and 50k is being attacked by a dps that does only physical damage and deals exactly 1k dps. With no ArPen, it would take this dps 160 seconds to kill the target. Add enough ArPen to negate 5k armor, and the target is reduced to 148k effective HP. With the dps still dealing 1k damage, it will now only take them 148 seconds to kill the target. Had the dps needed to achieve a 148 second kill without ArPen, the only option would be to increase dps to 1082. So negating 5% of the target's armor was equivalent to an 8% increase in dps.
Now let's say the dps adds enough ArPen to negate another 5k. The target drops to 136k effective HP, and the dps now needs only 136 seconds to kill it. Had the dps needed to raise dps instead of negate armor, the new dps would have to be 1177, which is an 17.5% increase from base. This means that negating that second 5k of armor, instead of yielding an 8% increase in dps like the first, yields a 9.5% increase, relative to the original dps. The third 5k of armor negated would yield an 11% increase, and so on. By the time ArPen was stacked to 100% so that the target only had 40k effective HP, that last 5k negated would be worth a 92.31% increase in dps... quite a far cry from the measly 8% yielded by that first 5k negated.
So the more ArPen you stack, the better it gets. Blizzard's challenge now is to make ArPen cost enough ilvl points that it's reasonable to build other stats as well. Unfortunately, this will be virtually impossible to do without making ArPen worthless, because once you have a decent amount of it more will always be better, and if you don't have a decent amount you'll always be better off avoiding it as much as possible. The best Blizzard can hope for is to force physical dps to a situation where they maintain a high ArPen set for bosses with a ton of armor to negate and use a set with as little ArPen as they can manage for the rest.
Dragon Sep 10th 2009 11:16AM
see everything above my comment? this is why Armor pen spell power etc are being removed... my head hurts after reading this. Cataclysm can't come soon enough in my opinion.
LilBanshee Sep 10th 2009 1:23PM
It turns out that the rules for armor penetration make an abrupt change at armor value 8317. It stops being a flat percent, and instead a big chunk of the armor lingers on even with armor penetration capped. So the example armor penetration data I gave at 50k armor was dead wrong, it doesn't go anywhere near dropping the armor of such an opponent to 0 (which would cause 300% increased damage). It appears that the maximum amount armor penetration can increase your DPS by is about 54.5%, which it hits at 8317 and then sort of lingers near into higher armor values. See http://www.wowwiki.com/Armor_Penetration for the charts that show this behavior.
@Random
You are right about that actually, in terms of damage per hit the damage does scale up at a greater-than-linear rate, it's only total time to make a kill that shrinks at a linear rate. So I was indeed looking at it a little wrong. Still, even before the growth curve comes into play the stat is already head and shoulders above competing stats. Let's compare the damage-increasing effects of ArPen side by side with crit rating, since they share the same item costs (i.e. an epic gem of either type is +20 rating):
Rating values of __ = __ damage
0 = 100% (initial damage with no ArPen or crit rating applied)
200 = 106% if points invested in ArPen, or 104% average damage if points invested in crit rating
400 = 113% if points invested in ArPen, or 109% average damage if points invested in crit rating
600 = 121% if points invested in ArPen, or 113% average damage if points invested in crit rating
800 = 130% if points invested in ArPen, or 117% average damage if points invested in crit rating
1000 = 140% if points invested in ArPen, or 122% average damage if points invested in crit rating
1230 = 154% if points invested in ArPen, or 127% average damage if points invested in crit rating
Note that even at low ratings, ArPen is about 50% better than crit rating on a point-for-point basis. By 1000 rating it becomes more like 75% better. If you were to cap it at 1230, it becomes 100% better. It's not just the scaling that makes it overpowered, it is already overpowered even at the first few points.
Darkharmony Sep 9th 2009 9:10AM
If they are so worried about people taking up secondary stats over your primary then logic would suggest that doesn't the same go for hit and expertise? I don't see what the big deal is, I mean armor pen was pretty much a junk stat before and now it's actually worth taking. I may not be a hunter but the same goes for other classes. To me especially end game raiding, armor pen only makes sense. Doesn't it make sense that as the bosses armor keeps growing that we also need to keep growing our penetration so that we are able to get past such heavy armor? I mean sure as a DK I could just keep growing my strength like I had but that only goes so far and it does nothing for me if I'm unable to get past such heavy armor.
LilBanshee Sep 9th 2009 9:19AM
The problem isn't at all that X stat is better than Y stat. The problem is more that by using armor penetration to its fullest, several classes were doing more damage than they were intended to. Blizzard estimates how much damage everyone should be doing and tweaks their abilities and talents to try to keep them within a few percentage points of one another. When people figure out a way to do much more damage than the blizzard simulations said they should be capable of, they gain an unfair advantage over all of the other classes. Blizzard is just bringing physical DPS classes back in line with the damage of caster DPS classes.
Snarkman Sep 9th 2009 9:45AM
@ LilBanshee
After this nerf, I will lose about 200 dps (and I normally pull about 6500-7000 dps). So, to think that this change will really affect the relative dps of physical vs. caster classes is completely wrong.
This change is about PvP, not PVE. Plain and simple.