I may be mistaken but I think this formula is wrong and they said so in that thread. It seems that this formula applies critical damage before the armor (which may be the case with magic attacks but then it lacks magic resistance) and attack after, which would make attack almost useless in virtually every situation.
The formula I've been using from this topic: http://forum.bioware...mbat-mechanics/
final_damage = (base_damage * rand(0.95 to 1.05) - armor * (1 - armor_penetration))
* (ability_multiplier)
* (1 + critical_damage_bonus + flanking_bonus)
* (1 + attack_bonus + damage_multiplier + type_bonus)
* (1 - magic_resistance)
A general note on what is best for stacking (please disregard if it sounds stupid). All stats have diminishing returns: if you have 10% attack item which provides +10 attack doubles that part of the damage. But if you already have 60% attack +10 item only improves it by 17%. That means that no matter how low your critical chance is it is possible to have such a high attack that it will be more beneficial to stack critical damage (although chance would likely be better anyway) and vice versa. It also means that it's possible to determine a mathematical optimum for attack to crit damage to crit chance ratio based on crafting bonuses ratio of 1:2:1 respectively. If I disregard the flanking bonus (mages rarely flank anyway but for rogues it's a huge deal) and damage multiplier (for simplicity let's say that the mage in question doesn't have abilities like Chaotic focus). At 50% critical chance there should be 100% critical damage and 50% attack. If chance and damage are exactly that but attack is lower, it will be better to increase attack and vice versa. If the critical chance is different the equation will be different... I think.
The following thread is very short but it has a very interesting post on this subject written by OrionAnderson: http://forum.bioware...rity-abilities/
I noticed that the formula was old, but also realized it doesn't matter with respect to comparing attack% contributed by attributes versus contribution of crit/crit.
First, remember the commutative law from math class? (a) * b = b * (a). It doesn't matter what order the factors are multiplied. Convince yourself by computing (1.15) * (1.20) and compare to (1.20) * (1.15).
Second, in the new formula, critical_damage_bonus = (critical chance) * (critical damage bonus), and attack_bonus is the sum of all the things that contribute to Attack%, including attributes. So zeroing out everything else that is irrelevant to this thread, we end up with:
(base_damage - armor * (1 - AP)) * (1 + critical_damage_bonus) * (1 + attack_bonus)
Ignoring the incorrect bracketing of the old formula, they end up being the same.
I'm sorry if this is a stupid question but what is this prowler armor for mages? All the prowler armors I've found in schematics list in wiki are for rogues...
You can remove the class restriction by using, respectively, Dales Loden Wool in light armor, Snofleur Skin in medium armor, and Silverite in heavy armor. You want your mage Herald to roam around in Legion of the Dead Armor? Use Silverite.