Pierce vs Blunt vs Cut Damage

[Mephisto]

This topic realy bugs ma mind. I hope anyone could give me a bigger overview of which damage type is better for different situations.
So far i have noticed that cut damage is best for smacking down doors, and barricades. Cut damage kills unnarmed peasants like through butter and blunt damage can knock other people out.

I've made out a test with Bec de Corbin. Attacked an ordinary player with both of it's ends. Apparantly the pierce did more damage as it only requiered two hits at the enemy. It took blunt 4 hits.

Well, anyone enlighten me?

[Odion]

pierce>blunt>cut

but length of weapon is greater than all, then speed is second, and if you look at the list of weapons cut has the lenght and speed, that is the only reason anyone uses cut weapons.

[CrazyCracka420]

http://infinitum.dyndns.org/crpg/calc.htm

Go to melee weapon damage calculator and play around.  In my opinion, cut works best versus people using little to no armor, up to about light mail armors.  When you start getting up to medium and heavy mail armors and higher, you should really start using pierce damage (think of it like "ignoring" the armor, basically the armor doesn't do nearly as much good when you use pierce).  Blunt damage generally works about the same as pierce.  I haven't run the numbers through the calculator, but a couple clan-mates of mine said that at high level armors, blunt and pierce damage basically do the same amount of damage.

I would go with a blunt weapon vs pierce if you are going up against a lightly armored person, just because a lot of blunt weapons have the chance to knock someone down (whereas pierce doesn't).  Versus heavy armor I've always liked my pierce weapons.

Generally "cut" damage has a higher listed damage value because when you factor in armor, it usually does a lot less than what it says.  Also, the enemies total health + armor is factored into the equation, as well as your strength and power strike.  Higher power strike allows you to still use cut damage versus more heavily armored opponents (but pierce would almost certainly still be better).

[CaptainQuantum]

This is a rather interesting part of the warband engine, so I will share what I knowledge I have.First of all let me make some definitions for ease of explanations, the weapon damage is split into two parts. The first part is the base magnitude of the damage on a naked character which I shall call n, and then there is the type of damage which is either b (blunt), p (pierce) or c(cut).

If we have weapons each of the same damage, but have different damage types; for example 30b, 30p, 30c, the blunt(b) will always do more damage than pierce and the cut and the pierce (p) will always do more than the cut (for none naked targets). However in general this is not the case since weapons are balanced for use in different situations.  So n will be smaller on blunt than pierce, smaller on pierce than cut (typically for weapons with all other stats the same). This is the order in native, but I have just noticed pierce has a lower reduction factor so there is an area where pierce>blunt. Another thing to note; on an unarmoured character all damage types do the same damage.

Now for the difference between the 3 damage types; each damage type decays differently with increasing armour. They depend on soak factors and reduction factors which are listed in your module.ini, I shall also list them below.

armor_soak_factor_against_cut       = 0.65
armor_soak_factor_against_pierce    = 0.5
armor_soak_factor_against_blunt     = 0.4

armor_reduction_factor_against_cut       = 1.6
armor_reduction_factor_against_pierce    = 1.1
armor_reduction_factor_against_blunt     = 1.3

The relation for how these behave in the engine is below:
effective_damage = (ran.*0.1+0.9)*RD - round((1.0-1.0/math.exp((ran.*0.55+0.45)*RD*0.014))*((ran.*0.1+0.9)*RD-(ran.*0.55+0.45)*SF)+(ran.*0.55+0.45)*SF)
RD=raw damage  RF=reduction factor*armour  SF=soak factor*armour  ran.=random.random () which is a structure call (I guess of random numbers)

There is also something for the extra penetration flag which archery uses, it essentially just multplies the RF and SF values by another value. The values are stated below:
extra_penetration_factor_soak = 1.2
extra_penetration_factor_reduction = 0.6

Raw damage is another thing entirely, and is different depending on your class so I won't post this.

Sorry for the wall of text, but this is pretty much everything on armours against damage type so I hope some of it is useful.

[CrazyCracka420]

I was told there wouldn't be any maths...can you break it down for the layman (or extrapolate the data for a couple examples)?

[CaptainQuantum]

I was told there wouldn't be any maths...can you break it down for the layman (or extrapolate the data for a couple examples)?

When you are dealing with this kind of thing, there will always be maths, since it's the way the universe works and also the only thing a computer understands.

I would break it down if I had time, but I want to play cRPG, I have spent a lot of my day studying Lagrangian dynamics. Doing much more simplistic maths may be a relief but I suspect cRPG will be a greater one. I just left a full description of the way damage behaves for armour in it's completeness, you can just read in to the simpler stuff if you are not a mathematically oriented person.

For the behaviour I will make 2 simple analogies:

• The soak factor can be interpreted as a factor which subtracts from the damage.
• The reduction factor can be thought of as the reducing a percentage of what is left over after the soak factor is applied.

Also the equation has random (ran.) functions in it, these are normalised probabilities and hence vary from 0 to 1. As armour value increases the amount that the random part contributes also increases.

This is as far as I will go for now.

Edit: Urist pointed out my error, I have corrected it now, so thanks to Urist.

[CrazyCracka420]

No wonder I always wanted to pull my hair out when forced to do any scripting or programming.  I'll just keep playing with layman understanding of the way pierce, blunt and cut work.  I don't need to know the exact algorithm. 

But it was nice of you to post what you know.

[Kekn]

Then can I at least say thanks for the formula; as a mathematician/analyst/programmer I DO appreciate something like this 

[Elmokki]

The relation for how these behave in the engine is below:
Damage after armour reduction = (ran.*0.1+0.9)*RD-[round{(1.0-1/((ran.*0.55+0.45)*RF*0.014)*[(ran.*0.1+0.9)*RD-(ran.*0.55+0.45)*SF]+(ran.*0.55+0.45)*SF}]
RD=raw damage  RF=reduction factor*armour  SF=soak factor*armour

I was gonna simplify it a bit but then I realized it's missing a ). Not sure where that would come or where you should remove a ( from.

(1.0-1/((ran.*0.55+0.45)*RF*0.014)*[(ran.*0.1+0.9)*RD-(ran.*0.55+0.45)*SF]+(ran.*0.55+0.45)*SF

One more ( in that part than ).

[Christo]

(1.0-1/((ran.*0.55+0.45)*RF*0.014)*[(ran.*0.1+0.9)*RD-(ran.*0.55+0.45)*SF]+(ran.*0.55+0.45)*SF

U-huh.

*scratches head*

[CaptainQuantum]

I was gonna simplify it a bit but then I realized it's missing a ). Not sure where that would come or where you should remove a ( from.

(1.0-1/((ran.*0.55+0.45)*RF*0.014)*[(ran.*0.1+0.9)*RD-(ran.*0.55+0.45)*SF]+(ran.*0.55+0.45)*SF

One more ( in that part than ).

Yeah sorry about that the way I had to do this was by following the code down and reading through the if statements to see which one I needed to apply, so it was quite likely I had left a bracket out Thanks for telling me anyway. I will figure this out and repost.

Edit: I have recompiled the equation directly from the code, it should be fine now. The extra bracket could of came from pretty much anywhere, since the ran. is actually renamed from random.random () - seems Taleworlds have a lot of originality in naming their structs. I could of forgot to delete the brackets. On a side note, if I get time I will post a guide for the behaviour of how the raw damage is calculated and what I have posted here, I currently do not know how speed bonus is calculated though (I could ask Urist).

[RamsesXXIIX]

I don't use the maths, but here's a simpler solution for you:

Cut has the highest damage of all weapons, but also both the greatest reduction and soak factor. As such, it deals the lowest damage against armor, relatively speaking.

Pierce has the lowest reduction factor and a mediocre soak factor. This means that piercing weapons penetrate armor easily, but still suffer from quite a lot of soak factor. As such, you will glance the least often with piercing weapons, giving you a rather consistent damage output.

Blunt has the lowest soak factor and a mediocre reduction factor. This means that blunt weapons will glance more often on armor than pierce (Not as much as cut though), but once it "breaks" through the armor, it deals a lot of damage. This is invented to make some blunt weapons effective against armor, and others not.

For example: Staffs and mauls both have blunt damage type. One would then think both would be good against armor. This is not the case.

The staff deals so little damage that in most cases it will glance off, never "breaking" the reduction or just barely.
The maul, on the other hand, is great against armor. Its damage is so high it has no troubles getting through the reduction, and once it "breaks" through, it has a very low soak factor. 

TL:DR: Both pierce and blunt are good against armor, as long as you are not using a low-damage blunt weapon.

[Paul]

Pierce has the lowest reduction factor and a mediocre soak factor. This means that piercing weapons penetrate armor easily, but still suffer from quite a lot of soak factor. As such, you will glance the least often with piercing weapons, giving you a rather consistent damage output.

Blunt has the lowest soak factor and a mediocre reduction factor. This means that blunt weapons will glance more often on armor than pierce (Not as much as cut though), but once it "breaks" through the armor, it deals a lot of damage. This is invented to make some blunt weapons effective against armor, and others not.

wat?
blunt ignores soak the most, thus glances the least. pierce ignores reduction the most, so it does the most damage on strong hits against heavy armor.

[Elmokki]

There's something just plain wrong with the equation still.

={(ran.*0.1 + 0.9)*RD} - round[ {(1.0 - 1.0 / ((ran.*0.55 + 0.45)*RF*0.014))*(ran.*0.1 + 0.9)*RD - (ran.* 0.55 + 0.45)*SF} + {(ran.*0.55 + 0.45)*SF}]

ran. on average is 0.5.

So:

0.95*RD

- 1.0
- 1.0 / ((0.725)*RF*0.014))
* 0.95*RD
- (0.725)*SF
+ 0.725*SF

So basically the 0.725*SF things remove each other removing the SF from equation completely.

((0.725)*RF*0.014) is 0.0105 * RF. With 80 armor and cut it's 1,344 and with 1 armor and pierce (0 makes it NaN) it's 0,0115. That means the multiplier for 0.95*RD will range from about 86 to about 0.75. With raw damage of 50 0.95*RD is 47.5. It's no real surprise that 47.5 * even 10 goes way way way over what the function is supposed to do. Though I might've made a mistake in both here and in Excel.

={(ran.*0.1 + 0.9)*RD} - round[ {(1.0 - 1.0 / ((ran.*0.55 + 0.45)*RF*0.014))*(ran.*0.1 + 0.9)*RD - (ran.* 0.55 + 0.45)*SF} + {(ran.*0.55 + 0.45)*SF}]

I bet there should be () so that the first SF is related to the multiplying of 1/(modified RF)*RD

[Mr. Hannibal]

There's something just plain wrong with the equation still.

={(ran.*0.1 + 0.9)*RD} - round[ {(1.0 - 1.0 / ((ran.*0.55 + 0.45)*RF*0.014))*(ran.*0.1 + 0.9)*RD - (ran.* 0.55 + 0.45)*SF} + {(ran.*0.55 + 0.45)*SF}]

ran. on average is 0.5.

So:

0.95*RD

- 1.0
- 1.0 / ((0.725)*RF*0.014))
* 0.95*RD
- (0.725)*SF
+ 0.725*SF

So basically the 0.725*SF things remove each other removing the SF from equation completely.

((0.725)*RF*0.014) is 0.0105 * RF. With 80 armor and cut it's 1,344 and with 1 armor and pierce (0 makes it NaN) it's 0,0115. That means the multiplier for 0.95*RD will range from about 86 to about 0.75. With raw damage of 50 0.95*RD is 47.5. It's no real surprise that 47.5 * even 10 goes way way way over what the function is supposed to do. Though I might've made a mistake in both here and in Excel.

={(ran.*0.1 + 0.9)*RD} - round[ {(1.0 - 1.0 / ((ran.*0.55 + 0.45)*RF*0.014))*(ran.*0.1 + 0.9)*RD - (ran.* 0.55 + 0.45)*SF} + {(ran.*0.55 + 0.45)*SF}]

I bet there should be () so that the first SF is related to the multiplying of 1/(modified RF)*RD

visitors can't see pics , please register or login