Finding the weapon damage ranges

[POCKET951]

How do you find the damage range of quality 6 weapons?

I could find the max armor ranges in the XML's but I didn't find anything about the dmg range for compound crossbows or desert vultures, or any Q6 weapon for that matter, just curious because I want to be able to easily tell if something is well rolled or not.

like yeah I can generally tell from knowledge and experience what's a high and a low roll, but I don't know what is the highest possible roll or the lowest possible one

I have a Q5 compound crossbow with 117.5 Dmg..which seems like a super high roll

[BFT2020]

        <passive_effect name="EntityDamage" operation="base_add" value="13.5" tags="perkArchery"/> <!-- damage offset -->

        <passive_effect name="EntityDamage" operation="perc_add" value="-.15,.15" tags="perkArchery"/> <!-- random EntityDmg -->
        <passive_effect name="EntityDamage" operation="perc_add" value=".1,.5" tier="2,6" tags="perkArchery"/> <!-- tier bonus -->

 

All of these determine the damage value once you factor in the type of ammo you are loading

Edited January 17, 2022 by BFT2020
Corrected from bow to crossbow (see edit history)

[theFlu]

Hmm... items.xml should give the answers, but the damage setting is spread between the weapon and the ammo types, and not exactly knowing how they interact isn't going to let me give a great answer.

 

The 'basic' bolts have damages 50, 59, 63 (stone, iron, steel).

The comp crossbow (items.xml:5428 gunBowT3CompoundCrossbow ) has a damage of..:

        <passive_effect name="EntityDamage" operation="base_add" value="13.5" tags="perkArchery"/> <!-- damage offset -->

        <passive_effect name="EntityDamage" operation="perc_add" value="-.15,.15" tags="perkArchery"/> <!-- random EntityDmg -->

        <passive_effect name="EntityDamage" operation="perc_add" value=".1,.5" tier="2,6" tags="perkArchery"/> <!-- tier bonus -->

 

So supposedly, 13.5 * 1.5 +/- 15% for a quality 6. That's 17.21 to 23.28 damage for the bow.

If you're sporting steel arrows, adding 63 wouldn't get us past 86 .. so there's some way to calculate all that, but you'll need to test quite a bit to come up with the actual formula.

You can probably see why I just want to shrug now ...

 

EDIT: I seem to recall some testing that lead me to believe that the effect of skills was shown on the tooltip, but the effect only altered the damage of the ammo (was testing guns, but it looks like bolts are set up similarly).

Edited January 17, 2022 by theFlu (see edit history)

[BFT2020]

 

Also

 

Don't forget that perks add to the damage (though I am not sure if that is displayed in the UI or not)

 

From Archery perk

 

           <passive_effect name="EntityDamage" operation="perc_add" level="1,5" value=".1,.5" tags="perkArchery"/>

 

Edited January 17, 2022 by BFT2020 (see edit history)

[theFlu]

Hah, we even did basically the same edit

[BFT2020]

We need @Boidster to run a bunch of tests to see how the numbers fall out  😉

[theFlu]

Assuming I got the range for the weapon right earlier, going with steel bolts and 5/5 archery, and my fuzzy recollection of that test; I could suggest the bolt damage simply being 63*1.5 = 94.5

From 117.5 backwards that would give 117.5 - 94.5 = 23 damage for the crossbow itself, which is two points above the range of a quality 5... something is adding something extra there. If it is the (50%) archery skill applied to the bow as well, the total 117.5 would be below the range which makes equally little sense ..

 

Some 10% night time buff or some such? I don't think these would show up on the UI, but who knows..

[POCKET951]

When I get home from work I am going to get some more numbers to see if we can reverse engineer this😅😂



With 3 mods 140.4 Weapon damage
Made a quick spreadsheet with the character sheet dmg , I do have have 4/5 points in Archery and 8/10 Agility

Steel bolts: 63 Dmg
Iron bolts: 59 dmg
Stone bolts: 50 dmg

could also not tell you why the mods alter the item tooltip dmg, but not the chracter sheet dmg

Edited January 18, 2022 by POCKET951 (see edit history)

[theFlu]

4 hours ago, POCKET951 said:

I do have have 4/5 points in Archery a

Well, there you have:

x-bow q5: 13.5 * 1.4 +/- 15% = 16.07 to 21.74

steel bolts with 4/5 arch: 63*1.4 = 88.2

the sum doesn't fit yet: 104.3 to 109.9

 

Giving the range of of the bow a benefit from archery would;

((13.5 * 1.4 +/- 15%)+63)*1.4 would give a range of 110.7 to 118.6

 

That only goes against my recollection of the old test; could well be the case, this would make more sense. Assuming you're using steel bolts.

[POCKET951]

On 1/18/2022 at 9:35 AM, theFlu said:

Well, there you have:

x-bow q5: 13.5 * 1.4 +/- 15% = 16.07 to 21.74

steel bolts with 4/5 arch: 63*1.4 = 88.2

the sum doesn't fit yet: 104.3 to 109.9

 

Giving the range of of the bow a benefit from archery would;

((13.5 * 1.4 +/- 15%)+63)*1.4 would give a range of 110.7 to 118.6

 

That only goes against my recollection of the old test; could well be the case, this would make more sense. Assuming you're using steel bolts.

we are struggling with this alot. surely someone at TFP understands how the method to this madness works. TFP please help or share your secrets
 

[BFT2020]

10 minutes ago, POCKET951 said:

we are struggling with this alot. surely someone at TFP understands how the method to this madness works. TFP please help or share your secrets

 

Let me give it a shot this week.  I want to create a test world and just have everything in it give me a specific weapon at the beginning and in large quantities.  Once I got my sample size to my liking, I can check my internal equations to see how they play out.

 

This will give me a break from working on my mod as I hit a mental wall with it (I got a lot of changes I want to make to it, but can't motivate myself right now to start them up) as it will be easy to do and allow me to do some testing / modeling of equations.

[BFT2020]

@theFlu @POCKET951

 

So, this is what I saw with a sample size of 144 for each quality level (2 storage crates per quality) - Compound bows

 

Explanation of matrix:

• Top row (greyed) for each ammo type is worst weapon found for that quality level
• Bottom row is best weapon found for that quality level
• Left column in each quality is the weapon found
• Right column is my equation applied
• So top row, my equation has to be equal or less to be true; bottom row is equal to or greater than

My equation

• (Ammo Entity Damage + Entity Damage offset for Compound Bow) * (random factor) *bonus based on Quality level (0.1 to 0.5 starting at Q2)
• so
  • (38+9.15)*(0.85 to 1.15) = Quality 1 Damage Range for iron arrows, then multiply that by the factor of 0.1 depending on quality level = Iron Arrows
  • (42+9.15)*(0.85 to 1.15) = Quality 1 Damage Range for steel arrows, then multiply that by the factor of 0.1 depending on quality level = Steel Arrows

 

 

I didn't see anything that violated my equation, but I might need to go further on my sample size (though out of 864 total compound bows, not a single one violated the equation)

Edited January 22, 2022 by BFT2020 (see edit history)

[POCKET951]

19 minutes ago, BFT2020 said:

@theFlu @POCKET951

 

So, this is what I saw with a sample size of 144 for each quality level (2 storage crates per quality) - Compound bows

 

Explanation of matrix:

• Top row (greyed) for each ammo type is worst weapon found for that quality level
• Bottom row is best weapon found for that quality level
• Left column in each quality is the weapon found
• Right column is my equation applied
• So top row, my equation has to be equal or less to be true; bottom row is equal to or greater than

My equation

• (Ammo Entity Damage + Entity Damage offset for Compound Bow) * (random factor) *bonus based on Quality level (0.1 to 0.5 starting at Q2)
• so
  • (38+9.15)*(0.85 to 1.15) = Quality 1 Damage Range for iron arrows, then multiply that by the factor of 0.1 depending on quality level = Iron Arrows
  • (42+9.15)*(0.85 to 1.15) = Quality 1 Damage Range for steel arrows, then multiply that by the factor of 0.1 depending on quality level = Steel Arrows

 

 

I didn't see anything that violated my equation, but I might need to go further on my sample size (though out of 864 total compound bows, not a single one violated the equation)

seems like a reasonable enough sample size for the compound bows.

844 compound crossbows theoretically should yield similar results though?

[theFlu]

28 minutes ago, BFT2020 said:

I didn't see anything that violated my equation, but I might need to go further on my sample size

The sample size seems good, and the equation is definitely in the usable range; but.. (sorry )

 

It looks your prediction might "overvalue" quality somewhat, like 0.5 points per quality. In both directions.

In your quality 1 you have samples landing spot on on the prediction, but the higher the quality, the further the samples "miss" the limits.

 

This may well be just a rounding issue somewhere in the programming, or just bad luck with the sample - but I can't spot any "as exact" hits in the higher qualities as the first, I'd assume at least _one_ closer hit.

[theFlu]

I wasn't sure if I'm just being nitpicky, or if there "must" be something at play here, so I thought about it a bit..

 

Your predicted range with Q6-Steel is 65.22 - 88.23. That is 23.01 'wide'.

Your results land between 69.7 and 84.1. This is 14.4 wide.

14.4 / 23.01 = 0.6258

The area where your 144 samples land, only cover 63% of the prediction. The likelihood for that sounds.. quite low.

Edited January 22, 2022 by theFlu (see edit history)

[BFT2020]

1 hour ago, theFlu said:

I wasn't sure if I'm just being nitpicky, or if there "must" be something at play here, so I thought about it a bit..

 

Your predicted range with Q6-Steel is 65.22 - 88.23. That is 23.01 'wide'.

Your results land between 69.7 and 84.1. This is 14.4 wide.

14.4 / 23.01 = 0.6258

The area where your 144 samples land, only cover 63% of the prediction. The likelihood for that sounds.. quite low.

probably because the perc_add's should be additive first - the game looks to add both the quality bonus and random variation at the same time before multiplying:

 

Equation:  

• (Ammo Entity Damage + Entity Damage offset for Compound Bow) * (1+random factor + quality adjustment)
• Low - weapon damage * (1-0.15+(0.1 to 0.5))
• High - weapon damage*(1+0.15+(0.1 to 0.5))

Table of these results

 

This puts it a lot closer to the distribution which is what we would expect with the sample size.  Plus, none of the displayed values violate the equation.

 

So next, how do perks come into play?  More data crunching to follow.

[theFlu]

13 minutes ago, BFT2020 said:

Equation:  

• (Ammo Entity Damage + Entity Damage offset for Compound Bow) * (1+random factor + quality adjustment)

<snip>

This puts it a lot closer to the distribution

Yup, that tightened it up real nicely. A bit surprising - I tried a couple combinations, but didn't think variation wouldn't multiply with quality ...

[theFlu]

Since we're still in @POCKET951 's thread;

 

Applying that formula to the OP:
min (63 + 13.5) * (1 - 0.15 + 0.4) = 95.625
max (63 + 13.5) * (1 + 0.15 + 0.4) = 118.575

 

Soo, yup, you should be close to the max roll on that. But that would not be showing the effect of skill points then; maybe it isn't? Let's see what BFT figures out about that ...

[POCKET951]

12 hours ago, theFlu said:

Since we're still in @POCKET951 's thread;

 

Applying that formula to the OP:
min (63 + 13.5) * (1 - 0.15 + 0.4) = 95.625
max (63 + 13.5) * (1 + 0.15 + 0.4) = 118.575

 

Soo, yup, you should be close to the max roll on that. But that would not be showing the effect of skill points then; maybe it isn't? Let's see what BFT figures out about that ...

Well it is quality 5 which is 0.4,if it was Qual 6 it would be 0.5, so that checks out kind of

 

It would also appear that skillpoints are not factored into the weapons sheet dmg, which would make sense

 

Also if we can correctly calculate the damage ranges of all weapons,we could make a spread sheet for the min-maxing community to look at and reference. 

Edited January 23, 2022 by POCKET951 (see edit history)

[POCKET951]

@BFT2020 are you still working on this or has it been shelved?

[BFT2020]

9 hours ago, POCKET951 said:

@BFT2020 are you still working on this or has it been shelved?

Sorry shelved.  I couldn't get consistent results with the perks added since they don't show up as UI.  Using zombies as targets, my damage was all over the place based on where I was hitting them.

 

Not patient enough to do data crunching if I can't get repeatable results with minimal effort  😁

[POCKET951]

Fair. I don't blame you.