CIPHER - Soul Annihilation lists and inflicts incorrect damage

[dam]

It would seem the Soulblade's Soul Annihilation lists (or deals) incorrect damage.

 

 

I had 72 focus.

I have 17 might which is +21% damage.

I also get +10% from cipher power levels.

 

The ability listed it would do 100 (I'm rounding, it was 101, or 100.7, or 103 whatever) damage.

 

I hit a beetle and dealt some weapon damage, + 209 raw damage.

 

 

This is consistent with the earlier report mentioning the 5 focus drain instead of whole pool, which also states there may be a problem wrt damage.

 

 

 

 

Edit: solution + math below, the damage listed in the ability is off by 5, the ability needs to state it benefits from damage multipliers excluding might which is already factored in the initial display.

Edited November 25, 2017 by dam

[dam]

Yeah nope, something is not right.

 

Listed damage :

 

Dealt damage :

[dam]

New try, this time on a flanked target, which means Sneak Attack qualifies (if it is applied to raw damage at all).

 

85 focus, 116.2 damage (for reference, an earlier look at 51 focus listed 75 raw damage)

 

Regular hit lists +245.1 raw damage.

 

Log detail lists the following modifiers :

+21% might (stat at 17) (which iirc is multiplicative after all other bonii)

+20% fine saber (sharp)

+15% fine saber (fine)

+50% sneak attack

+40% biting whip

 

Okay so we're supposed to get what...

(116.2 * ( 1 + 0.2 + 0.15 + 0.5 + 0.4)) * 1.21 (might is already accounted for in the ability's raw damage display) =

116.2 * 2.25 = 261.45 , yeah that's not it.

 

Let's try without stacking Saber bonii, you never know.

116.2 * 2.10 = 244.02 , that is not it either.

 

 

The thing is, our first try at 261.45 could be it, as a matter of fact if for some reason the game thought we had a bit less focus available.

Soul Annihilation listed 116.2 but that was at 85 focus.

Let us assume the engine removes the 5 focus activation cost first, then applies multipliers to our remaining damage.

We're looking at 245.1 / 2.25 = 108.9333 damage from the ability.

 

The number looks realistic at 80 usable focus.

If we remove our 21% bonus from might like so :

108.9333 / 1.21 , we get 90.0275 which starts resembling very much two well-known variables.

 

90.0275 = 10 (base ability damage) + 80 (usable focus after paying the 5 focus activation cost) + 0.0275 (get the fck out of my math you're ruining everything)

 

 

The number also looks realistic at 75 listed damage for 51 focus.

75 / 1.21 = 61.9834 = 10 base + 51 focus + 0.9834 (I'm thinking the game shows 51 focus, but I'm actually much closer to having 52 and the display is rounded)

 

 

 

 

 

 

 

Yeah tell you what, that's got to be it.

 

Soul Annihilation's formula is ((10 + (current_focus - 5) * (1 + might_bonus%)) raw damage

or : (current focus + 5) * (1 + might mod)

This number then gets multiplied by all your other bonii excluding might.

 

It does not however, benefit from Pen >> Armor , for one or both of two reasons :

1/ it is considered a lash-like effect (lightning strikes from monks for example), which are currently bugged (unless it's intentional for them as well) in this respect.

2/ it is raw damage and is excluded from penetration maths from the get go.

 

 

 

 

Okay, all things said and done, I submit that :

 

1/ the damage displayed in the ability tooltip is incorrect by an offset of +5

2/ the ability description should be updated to state that this damage benefits from all applicable multipliers (sneak, assassinate, biting whip, weapon quality...)

Edited November 25, 2017 by dam

[dam]

Although... the tooltip does say : 10 + 10% cipher power level + 21% might

 

Where the hell's my 10% here...

 

 

45 focus lists 67.8 raw damage = 1.5066 damage per focus

51 focus lists 75 raw damage = 1.4705 damage per focus

85 focus listed 116.2 raw damage = 1.3670 damage per focus

 

 

 

Scenario 1

10 base damage is flat and no multipliers are applied

Bonus damage before focus = 10

 

Let's discount our alleged 10 base free damage before focus is factored in, because that one skews results.

45 focus lists 57.8 raw damage (+10 base) = 1.2844 damage per focus

51 focus lists 65 raw damage (+10 base) = 1.2745 damage per focus

85 focus lists 106.2 raw damage (+10 base) = 1.2494 damage per focus

 

 

Yeah, nope, shoulda gotten equal damage per focus, and we don't.

Standard deviation is 0.018 (give or take), which is insufficient.

Samples fall within the 2 sigma 95% range, but no better.

This is insufficient and scientific proof is not established.

 

This bit of math is incorrect (even a larger sample pool wouldn't help much at only 95%)

 

 

 

 

 

Scenario 2

10 base damage gets bonii, bonii are additive : 1 + (0.21 + 0.1) = 1.31x multiplier

Bonus damage before focus = 13.1

 

Or maybe the 10 flat damage gets multiplied by 1.31 as well ?

 

Let's redo, and shave 13.1 from our initial numbers.

45 focus = 54.7 damage = 1.21555 per focus

51 focus = 61.9 damage = 1.21372 per focus

85 focus = 103.1 damage = 1.21294 per focus

 

std dev is 0.001339 this is a huge improvement.

This formula shows 99.33% confidence on this small sample size, aka 0.67% error rate, aka 1 error every 149 events.

This is not quite 3 sigma (99.7) but we're getting closer.

This means using this formula we need to make 149 tries before the game engine delivers a result which doesn't match.

 

C'mon I'm so close -.-

 

 

 

 

Scenario 3

10 base damage gets bonii, and bonii are multiplicative : (1 + 0.21 ) * ( 1 + 0.1) = 1.331x mult

Bonus damage before focus = 13.31

 

Or maybe the 10 flat damage gets multiplied by 1.21 first (might) then 1.1 (cipher levels) ?

 

Let's redo, and shave 13.31 from our initial numbers.

45 focus = 54.49 damage = 1.20960 per focus

51 focus = 61.69 damage = 1.21047 per focus

85 focus = 102.89 damage = 1.21088 per focus

 

std dev is (rounded down) at 0.000653.

This formula also shows 99.33% confidence, again on this small sample size, aka 0.67 error rate, aka 1 error every 149 events.

Well, technically it's 0.6694 error rate, so 1 error every 149.39 (rounded up) events ^^

 

 

 

 

 

 

Sadly it's not that either.

 

However, we're not working with floats (as we should for confidence calculations), we're working with what the engine gave us.

It seems, from empirical evidence, that the game engine rounds its floats, at the very least display-wise.

It wouldn't make much sense to show users damage numbers like 13.4894394 .

We may, therefore, assume the samples aren't perfect and account for the margin of error in one of the formulae (scenarios 2 and 3).

 

 

Given the rounding, given 99.33% hit rate on both formulae, I'm gonna stand by them and claim one of them works to calculate Soul Annihilation's damage.

 

That means the tooltip is still off ^^

Edited November 25, 2017 by dam

[AndreaColombo]

Subbing.

[Cdiaz]

Hey dam,

 

Massive kudos for the quick maths! Ive added your info to our systems to aid our devs if necessary. Thanks again!

 

You def the best,

 

-Caleb