I looked at the numbers using your current formula. I have no open account atm (not enough time) so I can only refer to experience. My Skald stated before does around 350-400 against decent casters. 370-450 against decent stealther (bonus damage with hammer vs. leather) around 350-400 against hib reinforced (again bonus) but only 270-300 against studded (reistant), 240-280 against chain/scale (resistant) and only 200 against plate (bonus). So somewhere in the formula absorb must be connected with absorb. If Plate only gives you 34% absorb and nothing more I should do more damage on Plate than on chain - which is not true. Plate: (1-21% melee resists) * (1-34%) vs chain: (1-36%)*(1-27%)

note: Plate is Armsman not Pala, Stealther without WS debuff but casters might have pd.

was thinking so too in the first place but then i got the test results. for example:
- plate: 106 - 160
- leather: 145 - 219
- cloth: 162 - 243 (with base af shield to get to 102)

now we do math:
162 (cloth) * 0.66 (plate) = 106.92
243 (cloth) * 0.66 (plate) = 160.38
162 (cloth) * 0.90 (leather) = 145.8
243 (cloth) * 0.90 (leather) = 218.7

so atleast for 102 item armor factor there is no difference other then absorb.

now we use the same testing set but with +25%BB spec af
- cloth: 139 - 208
- plate: 91 - 137

now we do math:
139 (cloth) * 0.66 (plate) = 91.74
208 (cloth) * 0.66 (plate) = 137.28

so the absorb has also no extra impact if you have a spec af buff. so the reason why you think you deal lesser damage to plate may be the equipment or the fact that an armsman has more defense tools then any chain class. only thing i have not tested out yet is toa af bonus, as i can not image that this is different then spec af. and even if it would be it would not make the plate in your example stronger then the chain.

a cloth af 51 + base af caster shield = 102 af. compared with a heretic without armor items with only a spec af (heretic shield buff is spec a spec af buff, thats special about this class) that gives you 517 af / 5 => 103.4 af. the result is that the heretic does not get any use from that spec af buff. i am guessing that: AF = AF * (item level) / (max item level). so if your item level = 0 you get zero bonus from spec and base af.

if i compare a cloth caster with base af and a spec af buff with a heretic with the same cloth armor + his own spec af with the amount of buffbonus so that you get equal af. the damage values match each other. so there is no difference between the af from a base or spec af buff.

and finally some testing on toa af bonus that is really looking deadly to me... test subject is a rr13 heretic with 25% buff bonus and 50 speced in enhancements. using pd9 and 26% item res + 5% res from the shield +3% class resist + 6% mythirian = 40% primary + 30% sec res => damage * 0.42. the values i use for this test are the resistances, that's why they are so high. the lvl 75 hits me slash mob hits me for 511 (- x) and the x i am using here (because its not caped and gives me big numbers to see small differences ). i only post the min damage values here:
naked + sitting: 3315
naked: 2000
naked + 50 toaaf + baseaf + specaf: 1333
naked + baseaf + specaf: 2000
naked + sitting + 50 toaaf: 2210
naked + 40 toaaf: 1428
naked + 25 toaaf: 1600

1) as allready said, if level 0 armor = naked => baseaf and specaf have no influence
2) sitting gets extra damage
3) toa af must be calculated total different then baseaf and spec af because it has a influence on damage in both cases, naked and naked + sitting.
4) doing math on this toa af effects the damage with: "DAMAGE = DAMAGE / (1 + 0.01 * TOA_AF)". so that 50 toa af are reducing the damage you get from 100% to 66.66....% or you can also say its like a plate armor...

sure i need to do more testing on toa af and how it works if you have armor equiped but still this is a very huge impact.

At this point it looks like ToA Af is being added to each individual piece. Not divided among the armor.


Example: 100 AF vs 100 AF + 50ToA AF

100/150 = 0.666667


Naturally these tests were done naked but it seems to indicate ToA AF is not divided, unlike base AF and possibly spec AF.

 

-----signature-----

have done a bit more tests. same heretic:
cloth + base + spec: 286
cloth + base + spec + toa: 273
cloth + sit + base + spec: 475
cloth + sit + base + spec + toa: 452

"naked" / ("naked" / "cloth + base + spec" + TOA_AF * 0.01) = "cloth + base + spec + toa"

so the effect of the toa af armor reduce is getting more weak the more your other armor (base + item + spec) reduces the damage.


@StanleyM84
and no toa af is not added on items itself it's added on the effect from the item.

well okay it's going to be even more complex -.-

1) if you sit down you take ~ "DAMAGE = DAMAGE * 1 / 0.6" in compare to standing

2) being naked and having 0 af is different from being equiped and having 0 af too. so it looks being naked is calculated total different with a different fixed factor maybe. was trying necromancers af debuff for this. i am getting 2000 damage on a naked target and ~2800 target with lvl 51 cloth armor - necromancer af. also looks like you cant be debuffed below 0 af.

3) so maybe af debuff works different. so i tested cloth armor vs cloth armor + base af - af debuff (that is more or less equal to the base af) and they were ~equal. so af debuff is working as expected.

4) so next try. i used 5 pieces of rpg armor and naked arms (there is no rpg armor for arms). and now things get really really strange... the damage on the naked arms was exactly the same as when being complete naked (this was expected). but the damage on the rpg armor is strange. the damage values where nearly the same as on the naked armor piece. but did not exactly matched (what they should do if they are so close together...). the result of this is that the game switches armor between two categories "real armor" and "other or no armor". i hoped that the game puts rpg armor as "lvl 0" cloth armor but it doesn't. but that is still nothing special. but if i now look at the damage values... the arms got exactly 2000 as min damage. the torso with rpg armor got 2002 (a naked torso would also get 2000 damage) so why the hell the rpg armor results in +2 damage? on the rpg head i got 2001 damage (again 2000 on a naked head). so why the hell i get more damage when using rpg instead of being naked and why the hell different armor slots get a different damage value? in all other test that i have loged the damage values of all slots did match when being equiped with the same armor. well there was one short and dirty test with af debuff (that should debuff to below 0 af) that did not. i got two values that where to close together but don't logged that test and don't know if it where different armor slots. very very strange stuff


if anyone got some ideas on special test cases and maybe special items to test with. would look for something like a "af 0" armor or so. i think i saw such an item that displayed 0 af but i don't remember it. this teleport quest helm or something like that anyone a idea? okay tested it.

EDIT: the helmet really got "af 0" but would need other items as well. 10% chance to hit the helmet makes testing sick not to say that doing the quest on a testing char is sick too ^^

EDIT2: done a quick test with the helmet. same result as with the rpg armor. damage nearly equal to naked. but a few points higher... so "af 0" items are all handled the same way?

run a longer test on the rpg armor. and the result is that the rpg armor has not only 50 possible damage values. it is just a real random between min and max with every value possible. in the test i did not match the min value from the naked armor but i got bigger values then the max value from the naked tests. so the rpg armor still takes more damage then someone being naked.

A guess of mine, might be way off but...


Maybe they were worried about a divided by zero error. So they made RPG armor have a armor level of negitive 1 or something very small. Or maybe these minor statistics anomolies are inherent only to RPG armor.


Just a guess.

 

-----signature-----

found some error in my calculations too... naked = 100% quality and armor is 70% quality. while rpg armor has no quality so it's 100% also. have to look at my numbers which other side effect is effected by this "missing quality" in my calcs

EDIT: okay i think i got it now : easy if you do not run calculations with wrong values ^^ "1 / (20 + AF)" is it. doing some more tests but with this one i should be able to finish the melee damage formula!

I lost the context, what exactly is "1 / (20 + AF)" expressing? The damage mitigation from armor?