Important Notice: I have come to information (wyrd's test logs and artemis' test in particular) that indicates
that the following information is at least partially incorrect. According to these logs from live, the variance max is around
141% of the base figure that originates from the damage calculation. Until I find concrete and up to date information, use
the following data with caution.
Overview
http://daoc.catacombs.com/class.cfm?ThreadKey=10852&DefMessage=324516&classid=41
To calculate damage, the game takes the WeapSkill, WeapDam, and the target's AF stats to determine how much damage you do.
Being higher level than an opponent grants increased damage and vice versa. Your character will hit very hard against grey/green
opponents, and hit for very little against red/purple opponents. The mechanics for this are two derived stats known as WeapDam
and weapSkill. WeapSkill is basically compared to AF to find out how much of a multiplicative bonus/penalty you receive for
hitting something different from your level (it appears to govern the grey to purple damage reduction we see). WeapDam is
the effective DPS of your weapon capped by your level. It is theorized that variations in discernable damage comes from the
defender's af cap at their level. Lower level creatures and players would have lower caps on af therefore making the attacker
do more damage by adding a higher multiplier to the attack. Higher level opponents would have higher af values and would therefor
cause lower damage to be done. At the same time mythic has stated changes to critical shot damage which appears to be level
based damage modifiers where the damage done is based off of a multiplier that takes specifically into account the attacker's
spec, and the target's level to determine the bonus/penalty to the damage inflicted on top of normal damage calculations.
It is unconfirmed if this also applies to all other damage as well.
It is important to note that with the 1.44 patch, WeapSkill was not updated to display correctly. Although higher is better,
the way it is calculated for your character sheet is no longer the same as how the game uses it. Therefore, it is not a completely
accurate measurement anymore, and it is a now an approximation (so far, it is accurate within 5%). Before the actual theory
is presented, a few more complex terms need to be defined:
WeapDam (i.e. your effective DPS) = capped base DPS for your level x quality x condition
EDPS = effective dps of the weapon
For base damage this is (clamped factor and weapon condition/quality)
For cap damage this is (clamped factor only)
Condition and quality of the weapon will affect effective damage in the following way :
Effective dam = Clamped Damage * Quality(%) * Condition(%)
The clamped damage is your maximimum allowed damage at your current lvl:
Clamped Damage = 1.2 + LVL * 0.3
Base Damage (per hit) = WeapDam (as a DPS) x unmodified weapon speed
Damage Multiplier = (WeapSkill / target's AF)
slow weap bonus = (1 + ( ( SPD  2 ) * .03) ) )
2h weapon bonus = (1.1 + (0.005 x 2hand Spec) )
Max Damage (per hit, 100% mark in damage variance equations) =
EDPS * (your WS/target AF) * (1absorb) * slow weap bonus * SPD * 2h weapon bonus
Damage Cap (per hit) = This is the absolute amount of damage that can be passed on to a player from a single hit. It is a
function of DPS, unmodified weapon speed, slow weapon bonus, or two hand weapon bonus if applicable.
As you can see, the maximum damage per hit is really just the Base Damage multiplied by the Damage Multipliter. This is the
100% mark referred by the weapon specialization section. If you are completely unspecced, then the damage variance of your
weapon will fall between 25125% of this 100% mark. If you are fully specced (or overspecced), then the damage variance will
be 100150% of this mark. As you can see, "Max Damage" is really a misnomer, as you can still go over this. The damage variance
is not something that appears on your character sheet, it is an intangible benefit to speccing in weapons. The actual variance
in damage itself (not the range you fall within from speccing) is not linear.
There are several other unknown factors involved in this equation, but their effects are small enough that it does not ruin
the approximation. These include: condition of the armor worn on the location hit, the CON of the target, and potentially
other factors.
To calculate the effects of relics, base character resists and realm abilitys that add resists you would perform the following
calculation:
Effective DPS * (your WS/target AF) * (relic bonus) * (1absorb) * (1base resists) * (1RA resists) * slow weap bonus
* SPD
Example, an attacker with 2000 weaponskill, 65 composite weapon specialization, using a 16.0 effective dps 4.0 speed one handed
weapon attacks an opponent with 635 af wearing chain armor.
EDPS * (your WS/target AF) * (1absorb) * slow weap bonus * SPD * 2h weapon bonus
16.0 * (2000 / 635 ) * 0.73 * 1.06 * 4 = 155.9785827
To calculate how resists factor into this, the following equation is used:
EDPS * (your WS/target AF) * (1absorb) * (1resist) * (1RA Resist) slow weap bonus * SPD * 2h weapon bonus
Example, an attacker with 2000 weaponskill, 65 composite weapon specialization, using a 16.0 effective dps 4.0 speed one handed
weapon attacks an opponent with 635 af wearing chain armor, and who also has 20% character resist along with 20% realm ability
resist.
16.0 * (2000 / 635 ) * 0.73 * 0.80 * 0.80 * 1.06 * 4 =
1.980928 = 99.826291
Further Reading
http://www.critshot.com/forums/archive/index.php?t9533.html
I believe the following entries, act as supporting evidence that this is the general formula for how mythic calculates base
damage.
Testing parameters:
Attacker: WS 1973, no relics, 63 bow spec, 5.6 speed bow with 16.5 effective dps
Taret: AF 648, 27% absorb, 16% effective resists.
Formula to be tested:
Effective DPS * (your WS/target AF) * (relic bonus) * (1absorb) * (1base resists) * (1RA resists) * slow weap bonus * SPD
* arrow bonus * toa bonus
Where EDPS is the effective dps on the weapon being used
SPD is the listed weapno speed
2h weapon bonus = 1.1 + (0.005 x spec)
SLow Weapon bonus = 1 + ( (spd  2) x 0.03)
Absorb is the absorption figure of the armor as a percentage (.27 for scale)
Where ffb arrow bonus = multiplier of the type of arrow used (ex. barbed flight footed broadhead arrow is 1.25)
GSV bonus is the archery damage multiplier from a TOA bow.
This results in predicted min value
1.415 2h bonus
16.5 dps
3.045 WS/AF
0.73 absorb
0.84 resists
1.108 slow weapon bonus
5.6 speed
1.25 FFB bonus
1.04 GSV/toa bonus
====================
351 minimum damage (predicted min with 16% resist)
Actually tested values came to be
Min: 320
Maximum was 459
Without absorb factored in, the calculation is 573 damage. With 27% absorb (from the armor) this brings the figure down to
418. A character with 16% resists would bring this down to 351 min. If the character had 26% resist (as i theorize) then
this figure comes to 309.7 (310 truncated). If the figure is 310 then the base variance range would be 310  465 which would
contain the reported min and max values here as well.
I believe the formula is sound and is accurate within an acceptable margin of error, although I think what happened here was
that wrw didnt notice he was using neutral arrows to the armor type he was testing. He apparently thought it was vulnerable
so he only calculated 16% resist.
Thus his 320 and 459 max works (with some rounding involved) if the character resist was 26%
310 x 1.04 gsv bow bonus = 320/321 . 310 is obtained by 441 (damage before abs is factored in) x 1.25 x 1.04 (gsv bonus).
573 x .73 (abs ) = 418. 418 x .74 (char resist) == 310.
I think that wrw also didnt have a large enough sample size to get the absolute min/max. I theorize that the min was 310 and
the max was around 465. Another possibility is that this difference between predicted damage and actually logged damage is
the mythic admission that the weaponskill values listed on the character sheets are some what incorrect in their display as
compared to how they are used in game post 1.44 patch. This could leave open for a margin of error in the final predicted
results.
