Armor Specialization feat

mitty · April 2017

I thought that a dedicated post on the heroic feat Armor Specialization would help develop the discussion and perhaps be of use to the GF community. Maybe even pop up in a google search!

As we know, the feat reads: Increases the effectiveness of Armor Class and Defence by 5/10/15%

@rjc9000 states in his Tactician Guide that the feat does not show up on our character sheets, but cites @grimah to confirm its effectiveness. No values for the feat are mentioned.

In this recent thread, @grimah states that Armor Specialization functions as a multiplier on our damage resistance: “15% more. so 60% = 69%”

So, if I am trying to get to 95%, once bondings and camaraderie/friendship proc (ignoring United for this discussion), that would mean we need 95% / 1.15 = 82.61% Damage Resistance to be at the cap.

This is a huge difference – has anyone else confirmed this?

I knew it didn’t show up on the character sheet, but I usually just ran around with 92% or 93% figuring Armor Spec would probably make up the difference, and any shortfall would be negligible.

I’m on PS4 so I am relying (as usual) on the pcmasterrace for data …

Thanks,
Mitty

rjc9000 · April 2017

I am getting some wacky values for my tests. I'll post them in case anybody with better testing/math skills can figure out what's going on.

Human test GF, tested on random mobs in the River District. I had no special debuffs (ie, no Rust Monster), mobs didn't get CA on me, 10 AC (ie, 0 DR from AC) ,and 1030 Def (1000 from Artifact + 30 from ???). 1030/400 = 2.575% DR, which the game rounded up to 2.6%.

Initial Test without Feat:

1 - 0.974 = 0.026 or 2.6% DR, as expected

Test with 3 points in the feat:

1 - 0.946 = 0.054% or 5.4%

If we assume Grimah is correct, we should take our original DR and multiply by 1.15.
(0.02575 * 1.15)= 0.0296125 or roughly 3.00% DR, which is not close to the calculated amount.

However, if I take (DR * 1.15) + DR, then it becomes pretty close.

(0.02575 * 1.15) + 0.02575
(0.0296125) + 0.02575
0.0553625

or, give or take, 5.5% DR, which is closer to the result given.

But it's still off by a small amount... and see the test below for as to why this isn't correct either.

---

Moon Elf test GF, tested on the random wolves in Preview SH. Mobs didn't get CA on me, I had no special debuffs (ex: Rust Monster), 10 AC (ie, 0 DR from AC) and 8000 Def (R10 Stables). 8000/400 = 20% DR.

Test without Feat:

1-0.8 = 0.2, or 20% DR as expected.

Test with 3 points in the feat

1-0.738 = 0.262 or 26.2% DR

If we try Grimah's formula

0.2 * 1.15 = 0.23 or 23%

This is off by 3.2%, according to our calculated result.

Assuming our old formula is right, we get:

(0.2 * 1.15) + 0.2
(0.23) + 0.2
0.43

Which is way off and means our original hypothesis is busted.

So, Grimah's math is off, my tests are off, or Armor Specialization's description makes no sense and is actually a scaling DR formula ala Armor of Bahamut (I tried the Armor of Bahamut formula, except replacing 0.1 with 0.15. Still weird faulty values).
____

@grimah, can you please confirm/deny with proof/screens of your tests?

Also calling in @michela123 and @thefabricant, since the former has experience testing out GF DR buffs, and the latter is a magician who is great at ~~pulling rabbits out of hats~~ using math analysis.

thefabricant · April 2017

I haven't finished testing this feat yet, but I did start giving it a cursory look. For starters, the AC portion of it is working as a multiplier on AC, (so if you would gain 0.5% dr from 1 AC normally instead you gain 0.5*(1+rank*0.05)). The defense portion is NOT functioning the same way however, but on the bright side it appears to be better than the tooltip. I have not yet figured out exactly wtf it is doing with your defense but it is giving you more than an increase of 15%, that is for sure. (So far for every test I have done, it has been a greater increase than 15%.)

hypervoreian · April 2017

Just a litle reminder for al that test this,cause i did that mistake some hours ago:

Some melee attacks of the adds in NW are categorized as AoE attacks even if they are in melee range:
As such there are weird interactions with the buggy AoE resist stat: sometimes it adds up in total dr,sometimes is a separate layer,sometimes it bugs and it works on top of the normal DR.

So test must be done with the same add ,attacking and same type of attack otherwise results will be variant

Edit: also the RI of the add must be known

rjc9000 · April 2017

thefabricant said:
I haven't finished testing this feat yet, but I did start giving it a cursory look. For starters, the AC portion of it is working as a multiplier on AC, (so if you would gain 0.5% dr from 1 AC normally instead you gain 0.5*(1+rank*0.05)). The defense portion is NOT functioning the same way however, but on the bright side it appears to be better than the tooltip. I have not yet figured out exactly wtf it is doing with your defense but it is giving you more than an increase of 15%, that is for sure. (So far for every test I have done, it has been a greater increase than 15%.)

Good to know about how the AC part works. And that the Defense part of the feat is extremely wacky.

Edit: Retroactive DERP. The extra 30 Def came from playing a human, as it just multiplies your Defense stat by 1.03.

hypervoreian said:
Just a litle reminder for al that test this,cause i did that mistake some hours ago:

Some melee attacks of the adds in NW are categorized as AoE attacks even if they are in melee range:
As such there are weird interactions with the buggy AoE resist stat: sometimes it adds up in total dr,sometimes is a separate layer,sometimes it bugs and it works on top of the normal DR.

So test must be done with the same add ,attacking and same type of attack otherwise results will be variant

Edit: also the RI of the add must be known

Awww...****, well, I need to check out my AoE resist stat when I'm able to go back on preview.

grimah · May 2017

Thanks for taking another look into this RJC. I haven't been active much due to a new job I've just moved too, but my tests were done a very long time ago I can not remember what my methodology was. I do remember it was not intensive and would not be surprised if I also had fluctuating results. But I know it did work (in one way or another)

maatmons · May 2017

Well, I just did a very informal test.

I was on Live server (the only one I have access to), I stripped enough gear to get my AC down to 10, took off that pesky Escalating Defense ring, and went to the tower district (where I figured enemies would have little Armor Penetration).

When my Character Sheet said 6% DR, the Axe Thrower dealt “8 (10)” damage to me. When it said 14.8%, he dealt “11 (15)” damage. And when it said 26.8%, he dealt “8 (13)” damage.

That actually looks pretty close to what you'd expect if Armor Specialization just added 15 percentage points to the DR figure on your character sheet.

Does whatever program you're using to calculate your actual DR factor in enemy Armor Penetration? If not, any idea what the wolves have for that stat?

rjc9000 · May 2017

maatmons said:
Well, I just did a very informal test.

I was on Live server (the only one I have access to), I stripped enough gear to get my AC down to 10, took off that pesky Escalating Defense ring, and went to the tower district (where I figured enemies would have little Armor Penetration).

When my Character Sheet said 6% DR, the Axe Thrower dealt “8 (10)” damage to me. When it said 14.8%, he dealt “11 (15)” damage. And when it said 26.8%, he dealt “8 (13)” damage.

That actually looks pretty close to what you'd expect if Armor Specialization just added 15 percentage points to the DR figure on your character sheet.

Does whatever program you're using to calculate your actual DR factor in enemy Armor Penetration? If not, any idea what the wolves have for that stat?

Yeah, ACT factors enemy ArmorPen for testing DR. The easiest way to check the wolf DR would probably to hop on my Tankadin on preview, walk up to the wolves, hold shift, and check effectiveness. I'm pretty lazy and I don't feel like begging Sharp to log on specifically to invite my Paladin to his preview guild.

We still can't explain the mystery of how crazy it is on a level 70: when Sharp tested/calculated more values, he found the feat had a logarithmic curve, which is a bit... weird for a mere 15% feat. Sharp even brought in backup in the form of @micky1p00, who is extremely well versed in math, but even Janne was stumped by what the feat does (or was distracted by DOTA Championships, either way, massive pain to test).

maatmons · May 2017

rjc9000 said:
he found the feat had a logarithmic curve

That's interesting. Would you be willing to post the numbers? I'd like to poke at it a bit. Mostly to check the possibility that it might be a second, multiplicative layer of damage reduction, and to compare to the old MMO standby formula of 1/(1+ax).

micky1p00 · May 2017

rjc9000 said:

maatmons said:
Well, I just did a very informal test.

I was on Live server (the only one I have access to), I stripped enough gear to get my AC down to 10, took off that pesky Escalating Defense ring, and went to the tower district (where I figured enemies would have little Armor Penetration).

When my Character Sheet said 6% DR, the Axe Thrower dealt “8 (10)” damage to me. When it said 14.8%, he dealt “11 (15)” damage. And when it said 26.8%, he dealt “8 (13)” damage.

That actually looks pretty close to what you'd expect if Armor Specialization just added 15 percentage points to the DR figure on your character sheet.

Does whatever program you're using to calculate your actual DR factor in enemy Armor Penetration? If not, any idea what the wolves have for that stat?
Yeah, ACT factors enemy ArmorPen for testing DR. The easiest way to check the wolf DR would probably to hop on my Tankadin on preview, walk up to the wolves, hold shift, and check effectiveness. I'm pretty lazy and I don't feel like begging Sharp to log on specifically to invite my Paladin to his preview guild.

We still can't explain the mystery of how crazy it is on a level 70: when Sharp tested/calculated more values, he found the feat had a logarithmic curve, which is a bit... weird for a mere 15% feat. Sharp even brought in backup in the form of @micky1p00, who is extremely well versed in math, but even Janne was stumped by what the feat does (or was distracted by DOTA Championships, either way, massive pain to test).

I totally forgot. It was the majors, and the group I wanted won !

rjc9000 · May 2017

maatmons said:

rjc9000 said:
he found the feat had a logarithmic curve
That's interesting. Would you be willing to post the numbers? I'd like to poke at it a bit. Mostly to check the possibility that it might be a second, multiplicative layer of damage reduction, and to compare to the old MMO standby formula of 1/(1+ax).

That was Sharp's data, not mine.
Whether he actually has the data or curve around is beyond me.

thefabricant · May 2017

Oh um, @maatmons or anyone else who wants to take a look, here are the numbers:

https://docs.google.com/spreadsheets/d/1MiqPCgV3PjkQ8JDz3h0tbz5rBAHthtGszMcOePqMJWU/edit?usp=sharing

crizpynutz · May 2017

Wow, I will never understand why the character sheet in this game doesn't reflect the actual values.

mitty · May 2017

This discussion is amazing and I am very grateful for all of the testing you guys have done.

I know we don't understand the exact effect of the Armor Spec feat, but in the meantime from a practical standpoint do we think it is giving us at least 15% extra damage reduction? Can we sit at 80% DR on the character sheet and assume we are at 95%? Or maybe add a couple percentage points to be safe?

I have been running around 82% (with camaraderie procs and bondings) and seem to be doing fine. The FBI hill climb gets me occasionally but I think that's due to sloppy play. I'm fine on the turtle and with mSVA. But I hesitate to recommend that to guildmates when I'm not sure.

Edit: Thanks @maatmons I can wrap my head around that. What a huge chunk of stats...

maatmons · May 2017

thefabricant said:
here are the numbers

Thanks for the data!

I had some luck treating it as a (percentage point) bonus to DR calculated as a function of Defense. Specifically, the following equation seems to describe how it works at rank 3.

y = (0.07498045*x)/(1722.055 + x)

Here, y is the amount you add to your DR, and x is your Defense score.

So, for example, if your Defense is 16,000, giving you 40.02042437% DR, rank 3 of the feat should add 6.769458733% extra DR, for 46.789883103% DR in total. That's very close to the observed value of 46.7902602%, and the other 6 data points I checked were also very close. (I checked rank 3 for all the 0 AC test values.)

Edit: I think I got the other ranks too.

Armor Specialization feat

Comments