Game Mechanics: AC/Mitigation Formula

[Zeelot]

There's been a lot of talk about the AC formula as of late being inaccurate. There's the issue of pets and players landing much higher than they should be on higher level NPCs. There's also the issue of players with very high AC getting consistently hit for similar amounts as having 100+ less ac.

When I have more time I would like to make some logs to test, but I haven't noticed any significant difference 1000 ac and 1150 ac when the difference should be huge.

I know the exact formula on live isn't solidly known, but there are some good parses and theories that have been done on evilgamer.net in the past.

Here's one of them and a link to a pretty comprehensive thread on the topic.

I might, but not any time soon. My main focus right now is to more or less figure out the mitigation formula so that actual mitigations can be determined between various classes and then maybe apply that to atk values. It won't do much for determining usefulness against any particular npc but should be a good genreal indicator of the mitigation improvement any piece of gear would give.

Right now my theory is the formula being ((ItemAC-(SoftcapMod*(Defense+Level)))/OvercapMod)+(SoftcapMod*(Defense+Level))+ShieldAC+ RawBuffAC/6

Actually, that formula is kinda messed up but I'm tired so it will have to do for now. Broken down what it's supposed to be is:
Determine Softcap - This counts at a 1:1 ratio, what is known is that Warriors are on the highest cap, other plate classes are on the second highest cap, Rangers and Berserkers were lower than monks but that semi recent patch to their mitigation increased their softcap to equal monks. Additionally, it's known that Defense skill affects the cap. It is also known from previous dev posts that softcaps were done by armor type (aside from the monk nerf/unnerf and the ranger/berserker change). 1My assumption is that like most combat skills, the success isn't determined by skill but rather by Level+Skill with a modifier. For now I'm using multiplication but it's quite possible it's actually division that is used. The formula I'm working with right now for this *THIS IS NOT CONFIRMED OR EVEN BACKED UP BY MUCH* is that it's Mod*(Defense+Level), for SK's, Paladins, Bards, and Clerics. For the mod for us I have been assuming it's 0.8. The softcaps as far as I know are:
Warrior
Shadow Knight=Bard=Paladin=Cleric
Ranger=Berserker
Monk
Rogue=Shaman
Druid=Beastlord
Enchanter=Magician=Necromancer=Wizard

I'm assuming the levels are 1, 0.8, 0.7, 0.6, 0.4, 0.3, 0.1 but I have very little to back this up.

What this would mean is that a level 70 Warrior with 270 defense would have a softcap of 340, a Paladin would have 272, a Monk would have 204, a Ranger would have 227, etc. After that base softcap is determined you need to look at mitigation aa's (avoidance aa's work the same way for avoidance AC), add up what you have (10 Arch, 2 Class, 5 OoW, 5 GoD, 5 DoD) then modify the softcap by that. If you have all 27% that softcap would increase by 27%. That would change the SK softcap from 272 to 345.

Overcap AC - These values are even harder to figure out since until the softcap is known, you can't figure out what the overcap value is. The assumption right now based on other parses is that overcap AC is worth 1/6 of normal AC for Knights. Whats known for this is that War>Pal=SK>Ranger>Monk>Wizard=Enchanter=Necromance r=Magician. I have no idea where the other classes fit into that or where those classes even fit for values other than having a strong assumption that War is 1/5. If you're an SK with 1500 worn AC not including a shield and max mitigation AA's you would take that 1500-345 and have 1155 AC over the softcap, then take 1/6 of that value and 192 AC.

Shield AC - This has been covered alot, it counts for full value rather than being under the softcap.

Buff AC - Probally the most likely component of my "formula" to be wrong. Going off the assumption that displayed AC is completly meaningless other than to screw with you, I think the real mitigation value on buffs goes off the raw data in Lucy rather than the displayed amount. The raw data multiplies the ac field by 0.3 then truncates it for the displayed amount. For example Form of Defense 3 shows 270 AC which Lucy then parses to 81. Assuming the raw data is actually the value used, I think buff AC takes that and then applies a softcap if applicable, thus an 81 ac buff for us (FoD3) would be 45 mitigation over the cap. Buff AC should be the easiest part of this to prove/disprove (I plan to tomorrow hopefully).

from -http://www.evilgamer.net/forums/show...p?t=155&page=6


The first page of the thread shows a parse that was used to help come up with this theory - http://www.evilgamer.net/forums/showthread.php?t=155

More parses/tests/similar theories from other people from thesteelwarrior: http://www.thesteelwarrior.org/forum...p/t-13652.html

Here's another test regarding AC cap and shields from the steel warrior: http://www.thesteelwarrior.org/forum...?t=7643&page=7

[Kope]

My bad on the other post, this seems to be a more comprehensive post /bump.

[Treats]

Unfortunately the data in that post isn't accurate for this server (Pre-Velious). There was no softcap until late Velious.

There was a hardcap on AC at two different points in the calculation. The first hardcap was on raw item AC depending on level, the second was on total AC depending on class after all calculations were finished.