Spiral Knights - damage increase percentages not exactly right?

damage increase percentages not exactly right?

12 replies [Last post]

Wed, 12/14/2011 - 11:55

Zincknight

it was common knowledge that 1 point of damage increase would give you 7-8% increase in damage however when i was collecting some data for poison resist and poison damage reductions, i found out that these estimates can be way off. heres some of the data i collected below but im not done yet and steam has stopped working for me, ill run more test tommorrow.

Triglav____________hit 1__ hit 2__ hit3__ charge
no bonus__________ 184__ 225__ N/A__ 320
max sword bonus___ 249__ 300__ N/A__ 419
increase %________ 35%__ 33%__ N/A__ 30%

Glacius____________ hit 1__ hit 2__ hit3__ charge
no bonus__________ 150__ 150__ 209__ not collected
max sword bonus___ 235__ 235__ 302__ not collected
increase %________ 56%__ 56%__ 44%__ not collected

BTB_______________ hit 1__ hit 2__ hit3__ charge
no bonus___________ 148__ 148__ 184__ 267
max sword bonus____ 205__ 205__ 249__ 352
increase %_________ 38%__ 38%__ 35%__ 31%

hope this crude table makes sense (thank god i previewed it first ;p)
its late ;p and im tired lol so i may have messed up somewhere in the reasoning but all the data should be right. all these test were run D19 against slimes and zombies wearing no dmg bonus then with full skolver + BTS. all damage is versus neutral mobs (no special dmg bonuses).
not that it really matters... and spiral knights damage calculation is bizzare 0-0 and i expect the devs want it to be that way. just to put out there that with the data so far damage increase can range from 5-9 % per bonus rather than 7%.... and no love for triglav from three rings ;p big suprise lol oh well...more reason to get ASI for slow swords then. would be cool if someone could run more test or even tell me sth i missed ;p. I will see tommorrow gudnight
EDIT : former damage bonus at 7% should be an average value... but i guess its good to know how much it can differ from that value and how little love triglav is getting TT *sniff*

Wed, 12/14/2011 - 12:25

Juances

Notice how the only split

Notice how the only split weapon had the biggest increase, while both one damage type weapons are similar.
If that means something.

Wed, 12/14/2011 - 12:39

Zincknight

... didnt sleep in the end TT haha

yea i noticed that too but it would make more sense if the special damage had a large increase split a medium one and normal a low one or something like that but this is just bizzare... and hurts my head to think about it ;p

Wed, 12/14/2011 - 13:13

Xainthejust

poor Triglav...never had a

poor Triglav...never had a chance. *cries over almost fully heated Jalovec* I wounder if the increased damage increase that the Glacius has is evident in either the DA or GF.

Wed, 12/14/2011 - 14:39

Velcro

It almost looks as if the

It almost looks as if the damage increase is weighted more heavily towards things that deal less damage. And that it treats Glacius as having two separate sources of damage, both being small, both getting a larger effect.

So Glacius has two sources of damage, elemental and normal, both dealing 75. Which receive a larger boost for being small. Or at least that's what my theory would predict.

This could be because of a base increase. So say damage increase works like

New_Damage = (Multiplier)*Old_Damage + Some_Constant

then if Glacius is treated as two separate sources of damage then it gets twice that Some_Constant. I'll fit the data to this model when I get a chance and see what happens.

Note that since your percentages are relative, lower damage values will receive a larger boost from Some_Constant and thus relatively increase more. So that could explain that too.

Thu, 12/15/2011 - 01:43

Zincknight

yea i thought about that ;p

and now that u mentioned that glacius may be considered as 2 sets of damage and the constant may be doubled it makes sense XD could someone test with a DA or GF to double check? since i dont currently have those weapons.

Thu, 12/15/2011 - 06:57

Bopp

another anecdote

A friend and I were wondering about how much increase each level of damage bonus yields. So we ran some tests, using a variety of weapons, and got an answer of about 5.5%. This is pretty different from the 7% or 8% that gets thrown around these forums. Maybe some change has been made, or maybe it depends on which weapons you're testing. Maybe it's time for another comprehensive study by some generous knight.

Thu, 12/15/2011 - 10:10

Velcro

Well as soon as I finish up

Well as soon as I finish up my work for the semester I would be happy to fit any and all data to proposed models and see what we get.

I don't really have a plethora of equipment to test, however.

Thu, 12/15/2011 - 11:02

Ciopo

it might be normalized over

it might be normalized over attack speed to ensure a similar increase over a given amount of time, anyway, before discussing any further, make sure you've read this

Thu, 12/15/2011 - 12:41

Zincknight

hmm

interesting username pupu claims that

Your damage numbers go higher and higher until D23 (about +10 / level), then they start to normalize (+3 / lvl). This is probably due to the nature of 5* weapons and depth scaling.
However, it causes the UV increase to fluctuate between 5% at D19 and 7% at D28. ( i believe all this was acheron based data vs slimes dunno why it would be different from my glaciuses increase) one way to explain this is that since it is doing more damage at first ( because of elemental vs slimes) the constant has a lower effect on the percentage increase which sort of supports velcro's model.

also volebamus test are run with a DA, GF, and DVS, 2 of which have split damage and one of which has very low damage thus an increase of a multiplier + a constant per damage type would give all of these the unaturally high 8.5% he got. which means velcro's model is still plausible considering both of this data, also since 5* weapons are limited in T2... perhaps DA and GF have to be retested in T3. and alot of pure damage vs split damage test should be run.

if it is indeed velcro's model enough test on different depths could maybe allow us to work out what the constant for each depth is XD. and from the data above we can assume the constant increases significantly as you go deeper enough to shift the bonus percentage quite alot.

of course the increase could be a completely different model from velcro's but it makes sense with the data we have so far (atleast to me ;p) and unless we find data that definitely disproves seems like a nice idea to start off with.

apologies for wall of text :p in summary: the old data seems to (in my opinion and reasoning) support velcro's suggested model

Fri, 12/16/2011 - 10:53

#10

Volebamus

Did some retesting myself

Did some retesting myself this morning in Tier 3, and the bonuses don't seem to start as low as 5%. The bonuses seem just short of 6%, but this happens to normal weapons it seems, while the Special Damage swords seem to get a better percentage increase. These numbers are just the first swing of the sword and all against Jellies, but equivalent numbers can appear to similar families like Constructs and Undead wherever weaknesses/neutrality is applicable:

. . . [Base, no DMG boost] . . . . . [Max DMG Boost] . . . FL | Com | Ach | DVS | Lev | XXX | Com | Ach | DVS | Lev | ---------------------------------------------------------- 19 | 150 | 197 | --- | 148 | XXX | 235 | 280 | --- | 205 | 20 | 163 | 215 | --- | 159 | XXX | 256 | 305 | --- | 220 | 21 | --- | --- | --- | --- | XXX | --- | --- | --- | --- | 22 | 189 | 251 | 168*| 180 | XXX | 301 | 358 | 228*| 250 |

*Combined damage swings: 115+53 base, 159+69 max

I am thinking to expand this to the entire tier 3 depth in regards to these 4 swords (and probably more swords, as I am lacking suitable non-weak pierce damage due to the depths I've entered lacking appropriate enemies for it), because some bonuses seem surprising. From depth 19 to depth 22, the bonuses for the swords per DMG unit jump as follows:
Combuster....: 9.4% - 9.9% Acheron......: 7.0% - 7.1% DV Striker...: 5.95% (fl 22 only) Leviathan....: 6.4% - 6.5%

Looks like normal swords are even worse than initially assumed if you ever look into damage boosts UVs/armor/trinkets.

Sat, 12/17/2011 - 11:25

#11

Volebamus

I did a new set for just 1

I did a new set for just 1 floor, but I realized something that might've been obvious to some:

Floor 24, FSC against zombies, first and final swings, (first number set is base, second number set max bonus)
Combuster: (270/356), (390/489)
Final Flourish: (193/241), (271/331)
Leviathan: (193/241), (271/331)

So yeah, against the brandish line has a better damage boost, and FF matches damage with Leviathan. This is actually evident when you check the wiki's damage tables for both, though I was able to at least confirm that the damage boost also parallels for the two.

Sat, 12/17/2011 - 13:18

#12

Velcro

Now if we could find a split

Now if we could find a split damage weapon that deals 193 damage we would be in heaven. Well the data on #11 is consistent with my model. The data from #10 is somewhat inconsistent with my model though. As DVS should have a larger relative change in damage but doesn't.

Another general note is that the increase seems to less on slashes later in combos.

One important thing to figure out though is how exactly damage bonuses are stacked. So far there is some sort of damage multiplier dependent on depth, one dependent on damage increase from equipment, one dependent on weak/neutral/strong, and potentially one from status resists against attacks aligned with that status. Oh and a multiplier from poison.

If I forgot one do remind me. It is important to establish whether these multipliers are applied additively or multiplicatively.

Edit:

Actually I just realized something. We should really do some extensive damage tests with either AP or Sentenza. Potentially both. If they do get changed to pure damage weapons then it would be a golden opportunity to compare split-type weapons with equal strength weapons of pure damage. Assuming they keep the same strength.

Search

damage increase percentages not exactly right?