Spiral Knights - Damage Vs Armor and Shields Analysis

Damage Vs Armor and Shields Analysis

18 replies [Last post]

Tue, 12/10/2013 - 09:26

Skepticraven

This Thread's Damage Spreadsheet

Purpose:

Identify Armor + Shield resistances at various depths.

Identify how depth affects armors (How 5* acts at D1).

Identify Shield HP.

Identify the heating curve of armor.

Methods:
This study utilizes a large quantity of various health boosts and changing the health boosts to measure the exact damage delt from a monster attack.
This study started with utilizing Shufflebots in guild hall trianing, but may in the future include traps from the advanced training.
By using multiple ranges and understanding the mechanics of what exactly the health percentage is displaying, damage ranges can be minimized down to an exact value. (Further Description Here)
Additionally, once damage has been delt, one can reduce the current health in intervals of 40 (one pip).
For additional ease of shield values, an extra attack can be endured with a known damage for the current armor being worn.

The current equation for identifying damage range is:
A = Health (In Pips, from the profile menu)
B = Remaining Health % Indicator (In decimal form)
C = Healed HP (In Pips, via hearts collected or capsules used)
D = Number of Damage Dealing Attacks

High Damage Range = (40A-RoundUp(40A*B)+40C)/D
Low Damage Range = (40A-RoundUp(40A*(B+0.01)-1)+40C)/D

Conclusions:

Sprite Perk is equal to UVs (Discussed/known in other threads)

Shield breaking damage can be obtained (Example: Proto shield absorbs 121 of the 235 damage before breaking at D23. HP can be determined separately by identifying via "Pure Damage".)

Armor Bonus is additive at low values (Damage reduced from Spiral Sallet + UV Med = Damage reduced from Spiral Sallet + Damage reduced from UV Med) Note: There is rounding (example: D23 combined UVs for Med+High [12+18≠29]).

Advanced Training is Depth 1.

Ghost blocks, crystal blocks, and respawning shufflebots deal 1 point of damage at D1, and 50 damage at D25-28 (S6).

Tested with deathmarks and drake scale against a layoafer (ice devilite) in save winterfest, armorless types of armor have 0 armor and there is no inherent armor in PvE.

UVs, Shields, and Armor have tiers (different reduction) for the different tiers.

Shield HP starts at 50, and increases each depth until the stratum is the shield star count +2. (Omega line seems to have 25% more hp than all other shields. This has not yet been identified completely.)

Notes:
If the armor image currently in my arsenal look visually the same, I assume they act the same. If you wish for me to add in a datapoint with a specific armor set, please request it.

Current list of identical looking armors that I have (Normal Resist ONLY):
Helm of set = Armor of set
5* Arcane Salamander = Heavenly Iron = Shadowsun = Snarbolax = Vog Cub = Perfect Seerus Mask =Mercerial Demo
5* Chaos = Mercerial (Non Demo) = Ice Queen
0* Fencing Jacket = Spiral (Proto Gear) = Vitasuit = Cyclops Cap
3* Quicksilver = Scary Skelly

Wed, 12/11/2013 - 08:14

Skepticraven

Shields Question

When calculating the hp of a shield, is it a good assumption to measure the equivalent "pips" and use that as the HP?

I know we can probably test this way down the line breaking a shield on non-normal pure damage (Say a proto shield vs Seerus Bombs).
I'm aware of this study for shields, but wonder how accurate it is with rounding errors.
Right now on the shields sheet in my data, I'm assuming the shield HP is the equivalent number of health (right above it on the UI).

*Edit: Comparing the damage numbers vs the equivalent pips of health, it is obviously incorrect (unless shields receive double the damage or something).

Wed, 12/11/2013 - 09:07

Zeddy

During my talks with Glacies, it seems like shield HP scales by depth or something? Shield damage at depth 19 seems identical to depth 24, even though the enemy most decidely inflicts more damage at depth 24.

Wed, 12/11/2013 - 12:55

Skepticraven

@Zeddy

For a Proto Shield:
D1 broke in 9 attacks. (14*9=126 damage)
D13 broke in 2 attacks. (92*2=184 damage)
D24 broke in 1 attack. (235*1=235 damage)

D1 is probably unreliable data to use because I had to chain 9 consecutive attacks without my shield charging.
It most definitely changes with depth, but not at the same rate damage does (probably matches closely at high depths - or may even match for heavier shields)

Edit: I guess I should craft a drake scale shield and test that for finding the HP.

Thu, 12/12/2013 - 11:50

Zeddy

That's three different tiers. Try different depths within the same tier.

Wed, 12/25/2013 - 19:34

Skepticraven

Interesting discovery

We can determine depth 1 shield hp through blocking ghost (or crystal) block spawns. Each ghost block deals 1 normal damage upon spawning (never reduced damage). Although shield recharges too fast to stand on one, there is a place to walk between 3.

That being said, one part of the equation for shields is removed (once it is discovered how they change with depth). Then the same analysis with reduction can be applied.

Additionally, status resistances can also be observed in the advanced training area. (although the shield will recharge, so there may be some estimate through pixel counting).

Edit:
Shufflebot spawns do the same 1 damage on all 3 guild training depths. This is extremely useful. Shield hp can be seen on depths 1, 13, and 23. Interestingly enough, it seems all shields have 50 hp at depth 1 (don't believe me? try it out yourself).

Thu, 12/26/2013 - 19:48

Glacies

Decided to try that thing with a multitude of shields and you're almost correct:

Everything - 50HP
Omega Shell - 62HP

Tested it with the Crystal Blocks in the Advanced Training Hall and I'd be willing to try it on the other depths but as Zeddy noted, I didn't see any difference between a Wolver Bite on Depth 19 as I did on Depth 23 along with other attacks. Even a Kat didn't see a damage difference.

Thu, 12/26/2013 - 23:14

Skepticraven

@Glacies

I dont have Omega Shell yet, but that seems like a special shield that always has the most HP.

I may take a hack at some other respawning blocks to find their damage to gather depths. A couple that I can think of offhand are ones in the jigsaw valley (ghost + a couple hard blocks) and winterfest maps (plenty of ice, and even the sharp ones after the boss). The jigsaw valley might be easy to record with the one gate that is active at the moment.

I'd have to check the gate again to see if there are some low depth candlestick keeps to look for the areas with respawning crystal blocks. That is probably the best option to keep the HP counting low and measure the change between depths.

Mon, 01/06/2014 - 08:53

Skepticraven

Added some data for shields in Stratum 6

Drake Scale Shield Values are as follows:
D25: 425
D26: 425
D28: 425

Shield HP is definitely NOT increasing with depth (At least for Drake Scale - I'll have to at some point craft a Dragon Scale or test the Drake Scale shield at a lower depth - I think it may change closer to Stratum 3). However, it definitely increases with Stratum changes (S5 drake scale shield hp is 425, S6 is 475).

Edit: I realized the previous data is wrong, because I used chaos as opposed to drake scale for some data. New data updated.
Edit#2: Stupid math error. Added damage instead of subtracting. The shields have 425, not 475 hp. Total HP = (Number of attacks)*(Pure Damage of attacks) - (Extra Pure Damage measured on Armor)

Sat, 12/28/2013 - 19:47

Glacies

So unless Attacks are increasing based on Stratas too by the exact same amount, 5* Shields must be hitting the highest possible amount just from hitting Tier 3.

Here's two pictures to compare from, both from the Chromalisk Family which both do pure-piercing damage (49.88% exactly of the Shields HP)

http://i.imgur.com/QRyq3lg.jpg - Depth 19
http://i.imgur.com/hGUOGpJ.jpg - Depth 25

The Depth 24 one was a non-statused attack which does the same damage as a Chromalisk against the shield (No Piercing UVs on either). There's absolutely no pixel difference which a 50HP difference should have at least shown a tiny change within the bar.

So unless the damage increases at the same exact rate as HP (which if this was the case it'd be 50 more damage exactly) then I don't see it increasing. I will try again later with a 1* - 4* Shield though to attempt at the damage difference see if it's there.

Sat, 12/28/2013 - 22:37

#10

Skepticraven

Interesting points...

I have a feeling a lot of this "increase per depth" type stuff is very much related.
I have a feeling I should also be testing Zeddy's 4% value soon.

As I was looking over the values for the proto shield on D1, I thought it might be equivalent to the Mining Hat values (Pictorially, the shield looks equivalent to the mining hat in the player menu - half what is seen in the inventory for helmets/armor). The math works out - The reduced damage that goes towards the shield HP after 9 attacks would be 6*9=54. I measured 4 extra damage after the shield was broken. The unfortunate part is that this same doesn't match completely for D13 (as it is off by 1). This could be a rounding error.

I may have to add in a chromalisk tab to my data to measure damages with divine. I was completely oblivious that they deal pure piercing damage. This will help greatly with my analysis. It will be slightly easier (being pure pierce) than finding all the difference block spawns. (And possibly change the name of this thread to the "Armor Analysis Megathread")

Now into your images...
I'm not a fan of measuring pixels since this game can be resized. If the images were exported as SVG's, my opinion would be different. However if we assume that the 5* shield has the same increase as I measured with the 3* shield, that would infer a 50+/-1%Total Shield HP difference between them. I've also read elsewhere that monsters with status deal about 20% less than those without. I don't disagree that it looks to be the same % damage being delt, I just think there are too many variables being changed at once to make complete sense of the data. Things like...
What is the chromalisk damage at those depths? (I've read elsewhere status-inflicting ones are 20% weaker)
What are the HP of the shields at those depths?

With the block tests I've been doing, I have been identifying the pure damage being delt by the block (using the minimization in post #1's link), and then breaking the shield (followed by measuring damage past the broken shield again through minimization).
I don't mind taking all the data, but don't have too wide a variety of shields.

I also realized something that may be of interest to you, Glacies.
We can test a "pure status" damage against shields.
How? I was playing around with a friend in the advanced training and he ended up tossing pots at me to try and burn me. It broke the proto shield I was testing at the time. I swapped to my Crest of Almire (which has good fire resist). I noticed something interesting: I got 0 shield hp damage (but had invincitinks showing damage was attempting to be placed). It is only depth 1, but it may be able to be reproducible with poison oil on the ground (Perhaps more to test when I find time to do so during C42).

Soo much to test, so much time it will take to test all the things... Maybe I'll collect the 100kcrn bounty Krakob put up for deciphering armor. :P

Sun, 12/29/2013 - 23:30

#11

Glacies

I should note that the entire Beast Family inflicts pure-piecing damage only on Elite to Shields, Zeddy confirmed for me that they do a majority of Piercing and Less Normal to normal damage but Divine Mantle will work just as well for that.

These are basically the damage values against a 5-Star Shield:

222px (61.66%) - Weak Damage (Dust Bunny Family)
180px (50.00%) - Moderate Damage (Wolver/Chroma Family)
144px (40.00%) - Strong Damage (Alpha Wolver Family)

Against a 3-Star Shield:

91px (37.916%) - Moderate Damage

Against a 2-Star Shield:

13px (7.22%) - Moderate Damage

As far as I know, that was just a theory for Shields which I disproved. I tested it during my initial testing and a Kat had the exact same damage values as a Hurrkat (Poison Variant) on the non-status attacks so there's absolutely no difference there.

Just to note on the "I'm not a fan of measuring pixels since this game can be resized." The game in no way re-sizes any of the HUD elements so the Shield Bars will always be the exact same length. The current ones I know are:

120 Pixels long - 1-Star*
180 Pixels long - 2-Star
240 Pixels long - 3-Star
300 Pixels long - 4-Star*
360 Pixels long - 5-Star

The ones marked with * mean that I'm just taking a guess on them since it just looks like the shield increases by 60px per rank.

Comparing pixels isn't a bad idea since it'll give you a general feel as how much HP the Shield has.

Though the 2%-per bar thing still generally holds up since it applies no matter where you go. It's just 2% off of the entire base HP I guess. (in which case, on S1, each unit of defense = -1 Damage.

EDIT:

I just realized, if all 3* Shields share the same HP at S5, this means that a Wolver on Elite does 264 Piercing Damage (equal to 6.51 Pips of HP) which if that's the case it means a Five-Star Shield has 528HP give-or-take by doubling the damage on Stratum 5 which is equal to about 13.2 Pips of Damage.

Mon, 12/30/2013 - 11:36

#12

Skepticraven

@Glacies

It is likely that since you cannot apply the status to the shield, it will absorb the status as 20% more damage (to be equal to non-status) unless it is resistant to that status. Once I get back to my desktop (in a bit less than a week), I'll be back to testing and fill find out the HP values for 4* 5* shields via using the dragon scale craft path.

As for the HP corresponding to length, as we see at D1 values this may be a bad idea. All the different lengths have the same hp. I can only assume that with the exception of the omega shell line, we could make a good comparison (as you have already) in T3 areas.

Mon, 01/06/2014 - 14:02

#13

Skepticraven

Boo Hiss

I started to do some chromalisk data, only to realize the T1 area (Snarby mission) they deal normal+pierce. I've been attempting to add the curve for shield HP and cannot get reliable data for 0-2* shields in T1 areas. I don't fully trust this data just yet, since I have some with tundralisks. I am aware that when they spit the blob, they have 2 damaging ticks. The first is the one I'm trying to measure and the other seems to be a "freeze" tink - which deals damage to a shield or status to armor.

More data will still be needed, but I may switch gears back to doing armor for a bit.

Mon, 01/13/2014 - 12:40

#14

Skepticraven

You know what would really help right now?

Finding missions that have some form of shankle on every depth. I started doing pure damage measurements to find shankle damage and it seems like the optimal method for discovering how things change with depth.

So far I know of shankle spawns in the following locations:
D4 (Faith in Armor) [Normal,Gold Normal]
D9 (Rescues and Recycling) [Element, Normal, Gold Normal]
D11 (Chilled to the Bone) [Shadow]
D15 (Sewar Stash) [Element]
D20 (The Great Escape) [Element]
D22 (The Great Escape) [Shadow]

I'll have to peel through the missions a bit more, but I should be able to find more missions with shankles for damage analysis.
I see the obvious pattern already for shield HP.
It increases per depth, until a certain point (depending upon their star count). After that point, they stay the same.

I'm likely going to scrap the entire shufflebot damage aspect in favor of using shankle damage. This allots for many more depths to be analyzed for pure damage on most types (T1 I've only found normal shankles). Also, an interesting note: Golden shankles deal less damage than killable shankles.

Edit: Adding more missions.

Sun, 01/12/2014 - 11:20

#15

Zeddy

Pretty much any depth with a clockwork passage should have shankles.

Sun, 01/12/2014 - 12:04

#16

Skepticraven

@Zeddy

I'm aware. It just takes a long time to cycle through all of the missions to find the clockwork tunnels themed depths. The wiki helps a lot for this, but is missing about half of the mission depth descriptions (And almost none list shankles as monsters within that depth).

Edit: I will also note that compounds also have a chance, but at a lower rate than the clockwork tunnels.

Sat, 01/25/2014 - 13:58

#17

Skepticraven

Update

I believe I have finished the majority of the shield HP values. (Conclusion updated in first post, and all calculations appear somewhere in the "Pure Damage" sheet.)
The main portion missing is the comparison between 4* and 5* shields in S6. I only have 2 datapoints in this area so far (1 4* on D24 and 1 5* on D28)
The odd part is that even with those 2 points, it seems that the images are accurate in showing that 4* and 5* shields have the same HP (and break the stratum rule). This makes it almost worthless to upgrade shields to 5*, pending additional tests in shadow lairs.

Additionally, I believe that shields function with similar resistances in comparison to one piece of armor. It is likely that when I figure out the model for that, shields can be matched and fit to that equation.

Fri, 09/19/2014 - 11:19

#18

Skepticraven

↓

A long standing update.

Glacies noted that difficulty affects barrier damage.
I immediately responded that I wasnt surprised and suspected that a baseline armor existed for varying difficulty (elite base armor = 0, normal = something non-zero).

I then went to test.
D4 is one of my favorite depths to test.
D4 Normal difficulty shankle damage against full drake (armorless vs normal) = 40 damage.
D4 Elite difficulty shankle damage against full drake (armorless vs normal) = 57 damage.

Next step. Ensure it is armor, not damage.
D4 Normal vs lost souls +4 nova driver = 55.
D4 Elite vs lost souls +4 nova driver = 55.

D4 Normal shankle vs +4 nova driver = 35.
D4 Elite shankle vs +4 nova driver = 35.

Next step. Make sure it is base armor, not added armor.
D4 Normal shankle damage vs full chaos (19 pips) = 97, 95, 93, 91, 89 (percent remaining each hit after 5 hits)
D4 Elite shankle damage vs full chaos (19 pips) = 96, 93, 90, 86, 83 (percent remaining each hit after 5 hits)

Maybe chaos is too weak, even for normal.
D4 Normal shankle damage vs full proto (17 pips) = 97, 95, 92, 90, 87 (percent remaining each hit after 5 hits)
D4 Elite shankle damage vs full proto (17 pips) = 96, 92, 88, 84, 80 (percent remaining each hit after 5 hits)

I don't know just yet, but obviously I can conclude that difficulty has a hand in armor but not damage.

Search

Damage Vs Armor and Shields Analysis