So how does armor function?

[woofty]

Also, please no comments for the peanut gallery.  Piercing weapons seem to be best against armored targets?  What of bashing weapons?  Piercing damage seems low but how does it fair better against a chopping (slicing) weapon?  I'm just having a hard time arranging priorities and what I want to use versus what is effective.  Can any one enlighten me on this subject, perhaps with a little math as to the function?

 

[dragoon47]

Slashing is good against unarmored targets.

Piercing  is good against armored and unarmored in most cases from what I can tell.

Blunt I think reduces armor values by half so its low damage actually works.

 

[Aeon221]

# You can modify the damage system by editing the following values:
# The first three values determine the amount which will be directly subtracted from damage due to armor.
# The next three values determine the percentage reduction from the damage.

armor_soak_factor_against_cut      = 0.8
armor_soak_factor_against_pierce    = 0.65
armor_soak_factor_against_blunt    = 0.5

armor_reduction_factor_against_cut      = 1.0
armor_reduction_factor_against_pierce    = 0.5
armor_reduction_factor_against_blunt    = 0.75

Like that.

 

[woofty]

# You can modify the damage system by editing the following values:
# The first three values determine the amount which will be directly subtracted from damage due to armor.
# The next three values determine the percentage reduction from the damage.

armor_soak_factor_against_cut      = 0.8
armor_soak_factor_against_pierce    = 0.65
armor_soak_factor_against_blunt    = 0.5

armor_reduction_factor_against_cut      = 1.0
armor_reduction_factor_against_pierce    = 0.5
armor_reduction_factor_against_blunt    = 0.75

Like that.

You're my new hero for the next 15 minutes.  Thanks.

 

[woofty]

Okay, I do have a couple questions.  Are the values presented mutiplicative values or divisitive (Damage/(1+Armor Val)).  Also, is the reduction calulated first of last?  Thanks folks.

 

[Captain_Octavius]

Okay, I do have a couple questions.  Are the values presented mutiplicative values or divisitive (Damage/(1+Armor Val)).  Also, is the reduction calulated first of last?  Thanks folks.

I am pretty sure the order is soak, reduction, speed bonus.

 

[Aeon221]

I've always assumed that the armor formula works like this

(Damage - (Armor Value * Damage Specific Soak Value)/(Damage * Damage Specific Reduction Value)


So if I'm right, then:

IF 50 cut Base Weapon Damage vs AL 30, (f(50) - (30 * 1))/(f(50) * 1)
IF 50 pierce BWD vs AL 30, (f(50) - (30 * .65))/(f(50) * .5)
IF 50 blunt BWD vs AL 30, (f(50) - (30 * .5))/(f(50) * .75)

Where f is the formula for determining total damage including weapon proficiencies, attribute bonuses, skill bonuses, speed bonuses and body location hit bonuses.

But I've got absolutely no proof!

 

[Pode]

Base damage = (50-100% of nominal damage + speed bonus)*skill bonus *att bonus *location bonus

Armor roll = 50-100% of nominal armor value

Damage after soak = base - soak factor * armor roll

Damage after reduction = damage after soak * (1-reduction factor*armor roll / 100)

Multiply this resultant damage by the armor value and integrate this function over the range of armors from 0-60 to get an armor weighted weapon power rating for a given weapon.  Since this result is a linear function of the nominal damage, you can compare the intercepts to determine that 20 blunt = 24.5 pierce = 29 cut.  So the rule of thumb I use is 4 blunt = 5 pierce = 6 cut

 

[darkweaver]

Multiply this resultant damage by the armor value and integrate this function over the range of armors from 0-60 to get an armor weighted weapon power rating for a given weapon.  Since this result is a linear function of the nominal damage, you can compare the intercepts to determine that 20 blunt = 24.5 pierce = 29 cut.  So the rule of thumb I use is 4 blunt = 5 pierce = 6 cut

I appreciate you posting the formulas, but the conclusion is misleading.  A high enough armor negates a range of damage values to 0, with damage becoming non-zero only past that point, so its clearly not linear.  If you mean that the integral over armor from 0-60 is a linear function of the base weapon damage, ok, but that's not very meaningful at all!

For example, with 0 armor, 1 blunt = 1 cut.  On the other hand, with 60 armor, the ratio is considerably more in favor of blunt than 4/5/6...  So you really can't just slap a ratio like that on it, convenient as it might be, since the effectiveness does depend on the armor values in question.

On a practical note, carry a good slashing weapon (for example, Balanced Military Cleaver or Masterwork Heavy Sabre, both 38 damage and good speed/range) for dealing with the riffraff, and a piercing or blunt weapon for dealing with the rarer heavily armored targets - a Balanced Military Pick works well.  At some level of power strike/weapon skill, the pick will reliably one-shot the riffraff too, making the cutting weapon obsolete, except for its slightly better speed and reach...

 

[Pode]

I skipped a lot of the maths in favor of words, which as usual resulted in a lack of clarity.  Take two.

Assume modifers of 1.0 and an average roll of 75% of nominal for both the weapon and the armor.

Compute resultant damage as a function of nominal damage and armor for each of the three damage types.

Multipy these by armor to create a new objective function.  This treats 1 point of damage inflicted on 60 points of armor as being of equal value to 60 points of damage inflicted on 1 point of armor.

Integrate these 3 objective functions over the range of available armors in game (0-60 is what I used).  This gives you the area under the curve, a numerical rating for the performance of the weapon over the whole range of armors.  Which is very meaningful, I would argue that that's the whole point of this discussion: what weapon does the most damage to the most targets, with a preference for damaging armored targets.

You're now left with 3 functions, each of which is a linear function of weapon nominal damage minus some constant.  Blunt has the flattest slope and the lowest intercept at 20, so it most reliably does some positive damage over the whole range of armor.  Piercing doesn't go positive until 25, but the higher slope means that it delivers more punch vs the higher armors.  Cutting has a middling slope and the highest intercept of 30. 
The intercept represents the nominal damage value that gives a net D*A score of zero over the whole range of armor.  The value where the damage to soft targets is just canceled out by the suckage against armored ones.  The lower the better, hence my rule of thumb 4b = 5p = 6c

 

[darkweaver]

I'm pretty sure I get the math, thanks for clearing it up though, especially what you mean by intercepts.

If I understand right,
1) The intercepts aren't meaningful because the actual damage dealt does not go negative, a fact that is ignored by this model.  I think this is effectively selling cutting damage short because it's saddled with a greater baggage of doing "negative damage".
2) Multiplying by armor is an arbitrary thing to do.  It's not clear that this is, as it were, objective, but seems a rather subjective way to boost the importance of armor.  The resultant damage is already a function of armor (and nominal damage) anyway.  Why does it matter that the damage was dealt vs armor or not?  If we're interested in the overall performance of the weapon, let's measure it directly (ie, the damage), instead of multiplying it by armor.

It's possible I'm misinterpreting something here, or just didn't consider it carefully enough...

The other thing is, at any given point, you're not fighting an "average" enemy.  You're most likely to be fighting an unarmored target (as much of the AI armies are low tier, and a lot of high tier troops are helmet-less).  It's arguable whether doing well vs soft targets in say 90% (random guess) of your fights is worth doing poorly vs the few hard targets you run into, but combining those two distinct cases into an average misses them both.


What *would* be really interesting is at which point the resultant damage is equal for each type vs a high armor target.  It's clear that lots of damage is good vs soft targets (which usually means cutting), but when is it worth it to pull out, say, a military pick instead of a heavy sabre, is to me the meaningful question.  Do you do that if you're going to attack a rhodok sharpshooter, or does it only become worthwhile for a swadian knight/sergeant etc.