BigRedKitty: Armor penetration stinks
Daniel Howell contributes BigRedKitty, a column with strategies, tips and tricks for and about the Hunter class, sprinkled with a healthy dose of completely improper, sometimes libelous, personal commentary."Dude, I've got two almost identical pieces of chest-armor, one has more Crit and one has more Armor Penetration. Which of these should I equip: the Gilded Ringmail Hauberk or the Razorstrike Breastplate?"
"Dude, you need the Armor Pen! Grab the Hauberk, the Armor Pen-chest!"
"Seriously, you think Arm Pen is good?'
"Dude! It's so sweet! It reduces the enemy's armor so it totally ramps-up your DPS!"
"Sweet! I'm totally down with the Hauberk."
/vendor Breastplate
This is what it sounds like, when Hunters cry.
An enemy's armor protects him from physical damage. If we can reduce that armor, our physical attacks will inflict more damage. Any problems with that? Sounds pretty simple, yes?
Blizz permits us two mathematical ways to reduce an enemy's armor: we can reduce a flat amount or a percentage.
Warriors use Sunder Armor to reduce a flat amount of armor. If a warrior is able to stack five Sunder Armors, he will reduce the enemy's armor by 3925; a flat amount. If the enemy has 4000 armor before the attack, he'll only have 75 armor to reduce incoming physical damage. Nice.
Hunter use Armor Penetration to reduce a percentage of armor. At level 80, a hunter requires an Armor Penetration rating of 15.38 to reduce an enemy's armor by 1%. If a hunter has an Armor Penetration rating of 154, and he attacks a different enemy that also has 4000 armor, the hunter will reduce the armor by 10% -- 400 -- and leave the enemy 3600 armor to reduce incoming physical damage.
Warrior have it easy; nerf warriors, right? Don't worry, we're going to dislike them even more later.
But let's pretend our enemy has a lot more armor, like 13,000. Now a warrior still reduces that by 3925, leaving 9075. Our hunter faces a different mob whose health armor is also 13,000, so the hunter reduces it by 10% -- 1300 -- leaving 11700.
Notice that previously we said the warrior and hunter were attacking two separate enemies. Well that's not how we raid, is it? The warrior and the hunter are usually attacking the same mob. And this is where a hunter's armor penetration takes it in the keister.
Armor Penetration is calculated using the mob's acting-armor, not his base-armor. Sunder Armor is calculated upon the mob's base-armor, not his acting-armor. What this basically means is that Sunder Armor is calculated first.
Observe.
We grab our mob with 13,000 armor. We know that the warrior's Sunder Armor reduces it by 3925, leaving 9075.
When the hunter attacked his own 13,000-armor mob, the 10% armor penetration reduced the armor by 1300. But when the hunter attacks the mob that the warrior is attacking, the armor penetration is calculated on the post-Sunder Armor value of 9075. Our Armor Penetration is going to reduce that value by 10% -- 907 -- leaving the mob with 8168 armor.
Well that's not bad, but it's certainly not what we wanted. We wanted 3925 from the warrior and 1300 from the hunter. But since Armor Penetration is calculated on the mob's acting-armor, we only reduced the armor by 907, not 1300. Our Armor Penetration rating, which we worked so hard to raise, has had its value reduced from 10% to just 6.98%, (907 / 13000 = 0.06979 or 6.98%).
6.98% instead of 10%? That's more than a 30% decrease in value from our armor penetration gear! And that, frankly, stinks. Even worse, there is another component to this debacle, and that has to do with the squishiness of the enemy. Observe, again.
For a mob with 13,000 armor, our 10% Armor Penetration was reduced to 6.98%. But what if we're fighting a mob with just 6000 armor?
The warrior still reduces that by 3925, leaving 2075. Now the hunter's Armor Penetration is applied, where 10% of 2075 is 207, reducing the enemy's armor to 1868. But reducing 6000 armor by only 207 is a reduction of only 3.45%!! (207 / 6000 = 0.0345 or 3.45%) Geebuz, that's a positively massive decrease from the 10% armor reduction we see on our tooltip.
The bigger the warrior's Sunder Armor, the less effective our Armor Penetration is. The squishier the enemy -- i.e. the less armor he has -- the less armor-reduction-efficiency our Armor Penetration returns. Basically, the more Armor Penetration rating we hunters stack, the less value we get for our efforts. Stinkin' warriors...
So now we return to our choice of armor: the Armor Pen chest or the Crit chest.
Crit is Crit, right? There's nothing anybody in our raid can do can decrease how effective our Crit is. When a hunter crits, he does more damage and generates Focus for his pet, thanks to Go For The Throat. Crit rocks and that's all there is to it. Crit-damage, pet-damage, it's all yummy. Unlike if we chose Armor Pen:
- Armor Pen can, and will, have its effectiveness immediately nerfed due to the warrior in the party throwing Sunder Armor
- The squishier the mob, the less effective our Armor Penetration will be
- Because we chose the Armor Pen-chest, we reduced our Crit by the amount we would've had, had we chosen the Crit-chest
Armor Penetration is not a bad stat, it's just the least-effective one we hunters have when we factor our party members interfering with our Armor Pen's efficiency. Adding Armor Penetration rating isn't a terrible thing, just remember to never sacrifice Agility or Ranged Attack Power or Crit to take Armor Pen.
And now you know why you should always look at your warriors with a crooked glance; they're in ur party, stealin' ur stats...
/shifty eyes
Filed under: Hunter, Analysis / Opinion, (Hunter) Big Red Kitty
Patch 5.2 interview with Dave Kosak
Inside an old alt's vault
The latest patch 5.2 news
All of the latest Mists of Pandaria news





Reader Comments (Page 3 of 5)
Ryan Dec 27th 2008 3:47PM
...that was supposed to be a response to Fandon's comment. No idea why it wasn't posted as such.
Ilnara Dec 27th 2008 4:01PM
The correct answer is : Keep both.
Razorstrike for PvE
Hauberk for PvP
done and done.
Nemos Dec 27th 2008 4:04PM
Great article- I love reading anything like this- give stats, and review why to choose one piece of gear over another. Class specific or not, this is one of the more useful articles I've read- THANKS!
leah Dec 27th 2008 4:04PM
let me first say that I'm amature theorycrafter at best (if at all) and math is ot myu strongest point, so what I'm presenting here is anecdotal evidence at best. but it still makes me wonder.
I had an opportunity to test out to different pairs of pants - crafted and drop. one had about 60 armor penetration and one had about that much crit. both were enchanted with epic ap/crit leg kit. with crafted pant (giantmaim leggins) I gained about 6% armor penetration. with the drop (leggins of stone halls) I gained about 1.5% crit (which brough me up to 27%) our tank was a warrior though I've yet to see any of our tanks have an opportunity to stack all 5 sunders on anything other then tank and spank raid bosses - things die too quickly. I've noticed that my overall dps was higher with pen then it was with crit.
I would need to test it again probably on a boss like patchwerk (otherwse known as "stand back and unleash" boss), but how do you explain my experience so far?
DragonFireKai Dec 27th 2008 4:14PM
@Mensrea
"The point is this: the effectiveness (or lack thereof) of ARP has nothing to do with sunder armor.
BRK spent several hundred words explaining that sunder made ARP worse, which is not really the case at all."
Sunder doesn't make armor piercing worse?
But your l2math post said this.
"The hunter attacking the mob alone sees a 43.4% reduction (2.6% physical DPS increase as compared to the base mob) while the hunter attacking the warrior's sundered target sees a 35% reduction (or 2.4% physical DPS increase as compared to the warrior alone with no armor pen)."
You painstakingly proved that 10% armor peircing provides a .2% lesser increase in DPS when combined with sunder. That's pretty much a textbook definition of Sunder making Armor Piercing worse. It might only be slightly worse, but it's definitely worse.
Andrew Dec 27th 2008 4:30PM
Keep reading.
mensrea Dec 27th 2008 5:45PM
That's not a flat 0.2% decrease across the board, it's a decrease in _one_ particular situation. There are situations where sunder actually increases the effectiveness of ARP by a comparable amount.
Even still, 0.2% IS nothing. You get more variance in your DPS out of the RNG than you do out of that. Hell, LAG probably accounts for a bigger variance in your DPS than that.
So while BRK was all "ZOMG SUNDER ARMOR NERFS ARP!" the truth of the matter is that arp is virtually the same with and without sunder applied.
Amaxe Dec 27th 2008 4:34PM
Yeah, i remember running EnhSim to see what was better for Mongoose (x2) vs Executioner/Mongoose or Executionerx2.
Executioner consistently lowered DPS.
Seems to me that Armor penetration is a broken stat. I have it as an incidental on some items, but not something i would sacrifice another stat for.
Angus Dec 27th 2008 5:12PM
You are a shaman.
ARP is pretty much a garbage stat.
What did EnhSim show your DPS percentages at? I'm guessing with the new system your spell damage was racking up around 20-30% of your damage. None of that cares about armor pen at all.
So right off the bat, 10% armor pen is only helping you do 7% more damage than you would be without it.
Enhance shaman see smaller gains from Haste and armor pen than pretty much any class. BM hunters actually hate haste more thanks to not needing a single bit to have steady shot hit the 1.5 second GCD cap. Hit rating for both Hunters and Shaman is a minor stat as once it is capped (spell cap for shaman) you pretty much don't care for more. It only helps white damage with shaman at that point instead of spell and white. Hunters with 9% for yellow and the normal hit cap on a gun are done after that.
So why does Mail have huge amounts of armor pen, haste and hit rating?
Did the guy in charge of Mail Itemization not get the memo, or do they have a chimpanzee in there pushing buttons?
Amaxe Dec 27th 2008 10:16PM
My stats were focussing entirely on melee (not spells) because I was working on making each component as solid as I could. Being enhancement, autoattack tends to be at the top, with Windfury a close second. The others fall off a good deal more after that.
Cyanea Dec 27th 2008 6:28PM
I always figured Crit was a nobrainer for hunters (especially BM) because it triggers all those lovely things that makes their pets eat my face.
Linross Dec 27th 2008 6:32PM
The topic is informative, however I have an issue with the maths you used to come to your conclusion.
That is true that the less armor an enemy has the less of it will be reduced my penetration, but at lower armor levels each point of armomr counts as more damage reduction. This is because armor sufferes from diminishing returns.
Counting that it might be possible that the amaound of damage that armor penetration increases stays roughly the same no matter the armor of the target. I do not have the numbers on that but it would be nice if someone more knowledgeable on the game formulas then me cleared this up.
Hnetu Dec 27th 2008 7:01PM
In truth Armor Penetration sucks for EVERYONE, not just Hunters.
I've gotten into a ton of arguements in my guild chat just telling people that Armor Penetration is a stat you never want to focus on. If you have two equal pieces of loot and one has armor pen and the other does, take the one that doesn't. It's the only stat that scales down and is therefor worthless in comparison to other stats.
Thanks for writing this, because it makes me feel better that people are being educated about it (oh and that I wasn't wrong all those time, haha.) This coming from someone with a Warrior and a Hunter, so I know it sucks for all of us.
Of course it still has it's uses, but I would never sacrifice other stats for armor penetration. I consider it a nice bonus stat on otherwise good gear, not a stat I look for. Y'know "oh hey that's an awesome ring. And it happens to have armor pen, nice little extra to an already good ring" or something. I dunno..
Chris Dec 27th 2008 7:10PM
Got a little error :P
"But let's pretend our enemy has a lot more ARMOR, like 13,000. Now a warrior still reduces that by 3925, leaving 9075. Our hunter faces a different mob whose HEALTH is also 13,000, so the hunter reduces it by 10% -- 1300 -- leaving 11700."
Shouldn't the HEALTH be armor?
swelt Dec 27th 2008 8:26PM
I think the warrior comments are misplaced. Here's why:
- Blizzard balance PvE encounters on the assumption that some raid member will be applying the Major Armor debuff. That person might be a warrior sundering, it might be a rogue exposing or it might be a hunter pet with the Acid Spit ability. If no-one in the raid is doing that, your DPS will be less, regardless of how much ArP you've got.
- DPS warriors are having the exact same problem with the new percentage based ArP mechanics that other physical DPS classes (including hunters) are complaining about. DPS warriors also may have the hassle of applying sunders to get the major armor debuff.
If you want someone to blame for this, you should be looking at the cloth classes that whined in TBC Arena Season 3 and onwards, when physical DPS classes could effectively make them naked with a combination of ArP talents, abilities and gear. This is why Blizzard changed it, and any fix to ArP will have to address this problem too.
Hydden Dec 27th 2008 9:17PM
I'm now crying because I specifically remember swapping out a crit item for an arp item...
Tuatha Dec 27th 2008 9:23PM
You're not looking at the whole picture, BRK. Armor isn't the end result; damage mitigated is.
Try to calculate how much dps is getting through on that 13k armor boss, then how much your dps increases on a 11,700 armor boss.
Then calculate how much dps is getting through on a 6k boss, and what the increase is between that and a 5,400 armor boss.
I think you'll find the percentage increase of dps in the two scenarios to be shockingly similar.
Note, I'm not saying armor pen is a great stat, I'm just saying it doesn't scale the way you seem to think it does.
Balius Dec 27th 2008 9:40PM
I have to assume that armor penetration is going to be changed during the course of wrath, simply because it IS such an undesireable stat. The resignation you see in people when items with ARP are the drops is palpable, and I can't imagine that Blizzard intended so much of their drop tables to be booby prizes.
In fact, the only real benefit to percentage based armor penetration over the old system is that it keeps clothies in PVP from getting decimated while reducing the effectiveness of plate versus physical attackers. Instead of the rock paper scissors design where a rogue, for instance, could beat a warlock but break his teeth on a paladin, it's a much more neutral system.
billy Dec 27th 2008 9:55PM
Mensrea is right.
This article is horrible. Sunder armor is a given in any PVE situation. If you want to decide if ArP is "garbage" given your current gear, then you need to do a real model with you current gear and determine your personal AEP values for the stats in question.
I can't even fathom how the author's original argument was supposed to work. In my mind it becomes paraphrased thus: "Look at how good armor penetration would be in this imaginary world, and now look at how shitty reality compares... we therefore conclude armor penetration sucks."
Blakkeyez Dec 28th 2008 1:17AM
I'm trying to understand this without knowing much about non-magical attack stats at all (For learning to judge situations that arise as raidleader I suppose or maybe it's only because I'm a nerd that like to read about stats for fun, I'm not sure.). Now I have a simple question that you all probably already know the answer to. Help me out: Does that armor/damage reduction apply only to the hunter's own attacks or is it applied to the mob and then benefit's all players in the raid (that isn't casters)?