Ah, I see. Interesting.
Analysis of Shield Defenses (and determining where the Grand Tortoise stands)
Sat, 09/12/2015 - 20:27
#52
Mon, 09/14/2015 - 08:45
#53
.....give me a cookie
Mon, 09/14/2015 - 13:40
#54
cookies but 1
(melody 'It's raining men)
singing 'It's raining cookies, hallelulia, it's raining cookies amen'
Oh but I unfortunally have to bring in one sad message though... the shield bar at the D27 Black Trojan is flawed...
But nevertheless thanx for your contribution <3 <3 <3
singing 'It's raining cookies, hallelulia, it's raining cookies amen'
@Hex:
Time to update the opening ;)
Mon, 09/14/2015 - 17:04
#55
Omega Shell + Grand Tortise
I've got both if you need.
Tue, 09/15/2015 - 04:04
#56
@Fading-Dusk
Yes, the Data of both shields are still missing. In best case both are at heat Level 10 or Level 5.
But even if this is not the case These will be still valuable Data for comparision.
Tue, 09/15/2015 - 15:08
#57
↓
5* heating curve may be useful if you try and find resistances at various heat levels. Keep in mind, 4* likely has a different heat curve than 5* (I've only measured 0*, 1*, and 5*).
Fri, 09/18/2015 - 03:18
#58
@Holy-Nightmare
Have these images been scaled? I don't remember a 960 x 720 resolution option.
Sat, 09/19/2015 - 07:46
#59
@Hex
But the length of his full shield bar has the same amount of pixels, correct? In this case it wouldn't matter.
Tue, 09/22/2015 - 10:55
#60
stoic shell and grand tortoise
hmmm... i do wonder about a thing....
I finally got hands of a plate shield with a shock uv and thus I started to upgrade it to the way to the GT. I did finish my testing run with the heat 1, 3-star version of the Grand Tortoise and wondered if I should make a whole recording of its whole way by testing all the stuff until 5 stars, heat 10 - which would equal at least 30 vana-shield-test-runs. But on the other side it would offer enough data to compare each single heat level of the GT series of all 3 torto shield star levels.
@Hex:
But on the other side, it seems that you do not have that much interest anymore since you stopped updating the opening with the recent data I added into this topic.
So my question:
Is this topic still active? Is someone else interested into analysing all shields to eachother and bring in equivalents between these shields?
In the end it would feel kinda dumb if I keep adding data which is not wanted after all. This is also the reason why I did not post my recent 3 star results yet...
Tue, 09/22/2015 - 12:36
#61
@ Hex
Cropped yes, scaled no
On a side note the IPS tanked just as good as my VPS (except against fire and shadow)
Thu, 09/24/2015 - 16:11
#62
But the length of his full shield bar has the same amount of pixels, correct? In this case it wouldn't matter.
It DOES matter. The GUI for the shield health bar is always 360 pixels long. When the Grand Tortoise takes an attack, it has 191 pixels left, so that means it has 191/360 of its maximum health remaining.
The shield health bar size in those screenshots is 279 pixels long for some reason. Can you explain this?
This means that in that first image of the set, his shield has approximately 143 pixels remaining (can't tell exactly due to it being a scaled jpg instead of a png) but it's not 143/360 of his shield health remaining, it's 143/279. When you scale it back up to 360ths to match the other shield's denominator, you lose precision because there's no such thing as a fraction of a pixel.
My guess is that imgur scaled down his image size, it tends to do that when your uploaded image is too large.
Thu, 09/24/2015 - 20:38
#63
hmmmm......
Try this instead: http://www.filedropper.com/skdata
stupid imgur....
Sun, 10/04/2015 - 07:10
#64
DATA! Marvellous, shiney DATA! -quoosh
@Holy: Your link requires a login.
@Everyone, but in special to Hex:
Data extraction service. It was surprisingly easy to get this done. Yet, given the amount of screens that needed to be extracted, it is still a time intense work. So I hope that this does motivate the topic opener Hex to update his opening :o. Here we go:
================================================================
Trojan's overhead smashes - remaining shield bar pixels:
================================================================
Smash vs Ancient Plate Shield (Heat 10):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 260 pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 258 pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 255 pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 252 pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, _88 pix
Smash vs Barbarous Thorn Shield (Heat 10):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _30 pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _30 pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _31 pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _31 pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, _±0 pix
Smash vs Crest of Almire (Heat 10):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _30 pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _30 pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _31 pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, _31 pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, _43 pix
Smash vs Dragon Scale Shield (Heat 8):
D24(Black): ???, ???, ???, ???, ???, ???, ???, _±0, ???, ??? pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, _±0, ???, ??? pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, _±0, ???, ??? pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, _±0, ???, ??? pix
D27(White): ???, ???, ???, ???, ???, ???, ???, -2*, ???, ??? pix
*Value in % of the knight's max health - aka the leftover damage the player recives in % by using Wyvern Scale Mail, Dragon Scale Helm and no normal defense 'at all' after the Trojan's smash.
Smash vs Grand Tortoise (Heat 5, 10):
D24(Black): ???, ???, ???, ???, 191, ???, ???, ???, ???, ??? pix
D25(Black): ???, ???, ???, ???, 191, ???, ???, ???, ???, ??? pix
D26(Black): ???, ???, ???, ???, 191, ???, ???, ???, ???, ??? pix
D27(Black): ???, ???, ???, ???, 191, ???, ???, ???, ???, ??? pix
D27(White): ???, ???, ???, ???, _75, ???, ???, ???, ???, _78 pix
Smash vs Iron Plate Shield (Heat 4):
D24(Black): ???, ???, ???, 192*,???, ???, ???, ???, ???, ??? pix
D25(Black): ???, ???, ???, 190*,???, ???, ???, ???, ???, ??? pix
D26(Black): ???, ???, ???, 191*,???, ???, ???, ???, ???, ??? pix
D27(Black): ???, ???, ???, 190*,???, ???, ???, ???, ???, ??? pix
D27(White): ???, ???, ???, _75*,???, ???, ???, ???, ???, ??? pix
*the accuracy is about ±3 pixels wrong due to screenshot compression
Smash vs Mighty Shell (4* ver. of GT) (Heat 1,3):
D24(Black): 156, ???, 156, ???, ???, ???, ???, ???, ???, ??? pix
D25(Black): 154, ???, 155, ???, ???, ???, ???, ???, ???, ??? pix
D26(Black): 153, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(Black): 151, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(White): _57, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
================================================================
Trojan's forward bash move - remaining shield bar pixels:
================================================================
Bash vs Ancient Plate Shield (Heat 10):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 132 pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 131 pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 131 pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 130 pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, 130 pix
Bash vs Barbarous Thorn Shield (Heat 10):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 112 pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 112 pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 112 pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 111 pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, 111 pix
Bash vs Crest of Almire (Heat 10):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 172 pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 172 pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 172 pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, 173 pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, 173 pix
Bash vs Dragon Scale Shield (Heat 8):
D24(Black): ???, ???, ???, ???, ???, ???, ???, _50, ???, ??? pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, _49, ???, ??? pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, _48, ???, ??? pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, _47, ???, ??? pix
D27(White): ???, ???, ???, ???, ???, ???, ???, _47, ???, ??? pix
Bash vs Grand Tortoise (no data):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
Bash vs Iron Plate Shield (no data):
D24(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D25(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D26(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(Black): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(White): ???, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
Bash vs Mighty Shell (4* ver. of GT) (Heat 1,3):
D24(Black): 102, ???, 102, ???, ???, ???, ???, ???, ???, ??? pix
D25(Black): 101, ???, 101, ???, ???, ???, ???, ???, ???, ??? pix
D26(Black): 100, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(Black): _99, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
D27(White): _99, ???, ???, ???, ???, ???, ???, ???, ???, ??? pix
Phew... it still took longer than expected. I also consider to create a chart on a personal wiki page. Actually I had a wiki page in mind for some time, now I got something to start with :D
Also a side note:
Mighty Shell got an unexpected double level, so 3*, heat 2 will stay clear - though, as far as it looks now, it does contain the same value as heat 1 and heat 3. From this moment on, I will use the smallest heating to make double levelling impossible.
Sat, 10/03/2015 - 08:48
#65
in case you were wondering
In case you were wondering whether anyone was still reading this thread: Yes. I await your conclusions with interest. Keep up the good work.
Sun, 10/04/2015 - 07:07
#66
Adding data
@Bopp
Why do you not just help adding some data? I am very sure that you have any shields in backpack that are able to bear 1 Trojan's attack.
@All:
Added the wiki page. I am just a starter in wiki editing, but finally I am almost satisfied with the outcome. Here the link:
http://wiki.spiralknights.com/User:JenovaSPIRIT/Shield_damage_tests
If you see yourselfe able to fill in a number or two, just contribute :D
Sun, 10/04/2015 - 13:45
#67
Need any help with Volcanic Plate Shield?
If so, I can help you with that.
(Or do Volcanic and Ironmight have the same normal defenses?)
Sun, 10/04/2015 - 18:18
#68
testings
@Vanaduke-Destroyer:
This still needs to be verified yet. Actually this is exactly one of the reasons why we do these tests: Which shields have same defenses and how much are these defenses? Since the normal defense bar exceeds the actual displayed area, we can only guess where it ends. And even then it could be slightly flawed.
It would be great if you can contribute with some data about this, exspecially, since the Volcanic one belongs to the group that Hex wanted to test initially.
Edit:
Here my examined data of the Dragon Scale Shield vs the weak version of Trojan's Sword Slamm:
DSS, Heat 08:
D24 (black): 108
D25 (black): 107
D26 (black): 108
D27 (black): 107
These are some veeery akward numbers, but I double prove it with 2 screenshorts for each number. Does anyone got an explanation about this?
Edit: Edit:
It did just bother me that much, that I just now epquipped my Crest of Almire and tried to reproduce that *weak* Sword Slamm (smash) with another Shield aside DSS - I succeeded:
Here are 3 screenshots that show my heat 10 CoA actually taking only slight damage by Trojan's Sword Slamms in Depth 24:
http://fs5.directupload.net/images/151005/wxlri6yz.png
http://fs5.directupload.net/images/151005/9mbqyyrs.png
http://fs5.directupload.net/images/151005/cthf4ibe.png
Messured remaining pixels: 167
Meanwhile I started to believe that that flawed number is somehow connected to the Attack Buff. I got that feeling that after the Trojan used the buff and Charged into a direction and after that attacking me with a Sword Slamm, I was getting that weak punch.
But yet again, I failed when I tried to reproduce this on purpose, but it could be that I just mights missed the timing when his Attack Buff ended.
So... maybe it appears in the second after that Attack Buff ends?
Sun, 10/04/2015 - 19:19
#69
what's your procedure?
After 60-some posts, I've lost track of exactly what your testing procedure is. Please explain it, and I will consider executing it, if I can find some time.
Sun, 10/04/2015 - 22:19
#70
@Bopp
The main purpose is - like described in the opening - to test various shields vs trojans overhead sword slamms. So the most important tests would be exactly those Sword Slamm attacks done by the black trojans.
I personally want to add some more data, so I keep testing the Trojan's forward Charge Attack, too, but this is optional to you.
So the least outcome will be 4 screenshots, each showing one Sword slamm of each *black* Trojan from deapth 24-27. The white Trojan would be on the way, too and is easier to test than its partner on D27, although he has split shadow damage.
When I do a test run I will always do at least two screenshots in case one of them is flawed.
As conditions:
-The run has to be on Elite difficulty
-The shield has to have full health before its hit
-The shield should not have shaking effects anymore while screening (I press F12 in the moment when the purple damaged shield bar starts to decrease)
-An uploaded screen has to be unscaled
-game resolution does not matter
A final word about D27:
In D27 there are only white Trojans around except in the last room, in which there will totally 4 black trojans spawn. You probably already know this.
Mon, 10/05/2015 - 04:02
#71
great
Great explanation. Thanks. If I can fit it in, I'll try some shields. But I don't have many unusual ones.
Wed, 05/24/2017 - 21:30
#72
Mighty Shell vs Trojans
I will edit this post as I add new data into it. For time reason I will only upload one screen per value, if interest exists, I can add the other screens per trojan, too.
Note: Max shield health is 300 pixels for 4 star shields, numbers show remaining pixels after one attack.
overhead Sword Slamm vs Mighty Shell (4* ver. of GT) (Heat 1,2,3,4,5,6,7,8,9,10):
D24(Black): 156, 156, 156, 157, 157, 158, 159, 159, 160, 161 pix
D25(Black): 154, 154, 155, 155, 156, 156, 157, 157, 158, 159 pix
D26(Black): 153, 153, 153, 154, 154, 155, 156, 156, 157, 157 pix
D27(Black): 151, 152, 152, 152, 153, 153, 154, 154, 155, 156 pix
D27(White): _57, _57, _58, _58, _58, _59, _60, _60, _60, _61 pix
forward Charge vs Mighty Shell (4* ver. of GT) (Heat 1,2,3,4,5,6,7,8,9,10):
D24(Black): 102, 102, 102, 102, 102, 102, 103, 103, 103, 103 pix
D25(Black): 101, 101, 101, 101, 101, 102, 102, 102, 102, 102 pix
D26(Black): 100, 100, 100, 100, 101, 101, 101, 101, 102, 102 pix
D27(Black): _99, _99, 100, 100, 100, 100, 100, 100, 101, 101 pix
D27(White): _99, _99, 100, 100, 100, 100, 100, 100, 101, 101 pix
Heat 01-Slamms:
D24(B): http://fs5.directupload.net/images/151006/3ca2qizz.png
D25(B): http://fs5.directupload.net/images/151006/5rqrfhin.png
D26(B): http://fs5.directupload.net/images/151006/j9fd4xjl.png
D27(B): http://fs5.directupload.net/images/151006/3llrg25t.png
D27(W): http://fs5.directupload.net/images/151006/drblxyj2.png
Heat 03-Slamms:
D24(B): http://fs5.directupload.net/images/151006/kcu6x26n.png
D25(B): http://fs5.directupload.net/images/151006/8gztrytk.png
D26(B): http://fs5.directupload.net/images/151006/vwfylgb8.png
D27(B): http://fs5.directupload.net/images/151006/ugorjnkl.png
D27(W): http://fs5.directupload.net/images/151006/dvsyg2yq.png
Other screens saved locally, if proof is required, I can upload them.
Edits:
30st March 2016 - Added heat level 04 slamms and added numbers to charge attacks of heat 01, 03 and 04. I also will not upload screens for now unless they get requested. uploading just stretches the time needed to progress this data unnecessarily.
1st April 2016 - Recorded and added heat level 05 to slams and charges
2nd April 2016 - Recorded and added heat level 06 to slams and charges
24th May 2017 - Recorded and added all remaining FSC trojans with all heat levels of Mighty Shell, now only one D28 trojan misses for wiki purpose.
25th May 2017 - Just crafted another Mighty shell in order to get Trojan's Attack Buff data for wiki and stopped at heat level 2 to fill missing numbers here.
Tue, 10/06/2015 - 11:30
#73
Tue, 10/06/2015 - 15:08
#74
Ironmight Plate Shield // Omega Shell
=====================
Ironmight Plate Shield
=====================
Extracting data from Holy's new screen - IPS (Heat level 4, no relevant UV):
24 (B): 190
25 (B): 190
26 (B): 190
27 (B): (190)*
27 (W): 75
*I did messure this by cutting the left side away to the point where the shield bar is starting on the other screens.
=====================
Omega Shell
=====================
Credit to Counttabb for the data:
Heat level 05, no relevant UV:
http://fs5.directupload.net/images/151006/ddb2exap.jpg
http://fs5.directupload.net/images/151006/m4f5caco.jpg
http://fs5.directupload.net/images/151006/8pmnges9.jpg
http://fs5.directupload.net/images/151006/o45ymfom.jpg
http://fs5.directupload.net/images/151006/2luyo6ll.jpg
http://fs5.directupload.net/images/151006/etuaqyxw.jpg
http://fs5.directupload.net/images/151006/q2rbbt5n.jpg
I got some difficulty to extract data though... can someone help me please?
Tue, 10/06/2015 - 23:55
#75
So there is enough shaky proof to say that the Grand Tortoise has roughly equivalent, if not identical defense values to Volcanic Plate Shield and the Ironmight Plate Shield. This is most unfortunate, because it means that it is outclassed in both mobility and defense by the Ancient Plate Shield, its only advantage being the improved shield bash.
I haven't been following this thread vigorously as I play Spiral Knights on and off. It's nice to see so much data being submitted, and I am greatly thankful to everyone who contributed, but I was mostly interested in the answer to just this question, as well as determining the difference in effectiveness of blocking Trojan attacks, since that's what I in particular use my shield for the most. Trojans are also a good control experiment because of their mostly (if not purely) normal damage attacks and high damage values means that it's not affected by the downscaling of defense effectiveness that occurs when damage is almost completely resisted.
@Jenovasforumchar
I got some difficulty to extract data though... can someone help me please?
Yeah, those images are pretty sketchy. Got what I could from it. Most of the images don't look to be from the Slam attacks. Unbuffed ones, anyway.
Wed, 10/07/2015 - 07:55
#76
Omega Shell
Omega Shell is heat level 5, as I wrote before the screen links.
The one who contributed that data is named Counttabb (with two 'b' in the end).
Also note that Counttabb was taking different kind of Trojan's attacks, not only Sword Slamms.
Edit:
Currently I am recording all heat levels of Mighty Shell and GT (except MS, heat 2 due to a double level) for a more in-deapth record of said shields. This is a long time project and so do not expect to have instantly all data ready.
Also:
I would like as next testing subject the well known void monsters of Dreams and Nightmares. I assume that there are many people who see some data there (me included). I assume that void monsters fallow the damage values of non-void monsters without (almost) no status effects. This would offer to test different types of damage values and compare them to eachother.
Ah about status-breaking:
I got a screenshot on which Vanaduke's Mace Slamm did break Mighty Shell, heat level 3 with one single Slamm. And this exactly to zero. Just like APS broke exactly to zero. In both cases the status effects stun and fire got still applied.
But the thing is: APS got far more normal defense and should at least proof better than a 4 star shield that only got heat level 3, shouldn't it? Because how I got it, they proof *exactly* the same in this particular situation.
Wed, 10/07/2015 - 13:41
#77
"Because how I got it, they proof *exactly* the same in this particular situation."
Because of their overflow effect being similar, it's not that surprising that these two shields would have the same effects when being broken but not overflowing their health. It'd be interesting to test whether status resistances affect the health overflow or if it's solely based on physical defense and health.
Thu, 10/08/2015 - 04:09
#78
mainly @Dib
I just got another thing that sounds much more reasonable to me:
-Both shields have 124 Health (fallowing the wiki). And since we know that heating does not affect shield's health, both shields got the very same amount of it.
-Both Shields got stun defense, probably with the same resistance, but in this case I am not sure how status does work on different star grades in shields. Let us just assume this is the same, too.
-The amount of physical defense is different, as proven out of Trojan damage tests.
So my theory is fallowing:
Vanaduke deals no physical defense at all on (at least) elite difficulty!
Preparing a proving test:
-Dragon Scale Shield should be able to hold Vanaduke's mace with ease. (kinda weird, I know, but at this moment I believe it will)
Other question:
Is there any shadow damage available in his mace?
Test:
-Getting numbers of a Dread Skelly Shield vs Vanaduke's mace (elite difficulty)
=> If Dread Skelly crushes with the same aftereffects like APS and MS, then Vanaduke's mace has no (split) shadow damage in his mace attack.
TLDR:
Theory: Vanaduke has not even 1 point physical damage in his slamm on elite difficulty, let us find out what kind of damage he deals!
-----------------------
Edit:
Okay... Tested, but the outcome does differ from what I expected...
APS, heat 10 again (from a previous post)
http://fs1.directupload.net/images/150615/jupog7wo.png
Mights Shell, heat 03:
http://fs5.directupload.net/images/151008/rma3ntox.png
DSS, heat 08:
http://fs5.directupload.net/images/151008/7tjvq8zu.png
Outcome:
All 3 Shields show *exactly* the same result...
So, could it be that APS and MS do really break only by a split elemental+missing fire resistance and that DSS breaks by missing normal defense? But why do they all break to exactly zero overflow?
Does someone has a Volcanic Plate Shield? (*looks towards Holy*)
Thu, 10/08/2015 - 04:56
#79
I have a VPS, IPS, APS, as well as all Torto Shields, but they are all at level one besides my APS and one other that I can't remember right now, which are level two. If you're okay with that slight difference, I can test almost any shield you want tbh.
Fri, 10/09/2015 - 06:31
#80
@Dib
Can you do a run with your VPS by testing some Sword Slamms?
By chance:
-The black trojans in depth 24-27
-The white trojan in depth 27
-Vana's Mace Slamm. One should be enough in case your shield breaks by 1 already
Would be cool :D
Sat, 10/10/2015 - 12:57
#81
↓
-Both Shields got stun defense, probably with the same resistance, but in this case I am not sure how status does work on different star grades in shields. Let us just assume this is the same, too.
Status changes with star counts. I haven't tried to model it just yet but may be able to with your data.
Sun, 10/11/2015 - 22:08
#82
@Jenova
I would like as next testing subject the well known void monsters of Dreams and Nightmares.
Ooh, yes, thanks for bringing this up.
So my theory is fallowing:
Vanaduke deals no physical defense at all on (at least) elite difficulty!
You mean normal? Or do you mean he does "environmental" damage rather than a typed damage?
What I am thinking is perhaps there is a maximum damage that an enemy can do to a shield. Perhaps that maximum damage is a 5 star shield's health on depth 28. So therefore if an enemy attacks a full health 5 star shield, no overflow is applied. This would explain why Omega Shell can take a Vana mace without breaking
I'll have to look closely at how much damage is being taken by the White Trojan buffed Slams/Bashes to see if that confirms this theory.
Mon, 10/12/2015 - 05:54
#83
@Hex
I withdraw the theory that vana deals no normal damage (and only elemental damage). Otherwise Dragon Scale Shield would have proven much better.
Your theory about a max damage sounds plausible, but needs to be tested first - like everything.
My new theory is that status is the thing that actually causes the shield to break. Remember my post about that strange phenomena with Dragon Scale Shield? It got kinda 3 different Sword Slamms:
1. Sword Slamm breaks shield to exactly ±0 and causes stun
2. Sword Slamm breaks shield to exactly ±0 but causes no stun
3. Sword Slamm does not break the shield. Ofc no stun gets applied in this case.
I also added a screenshots in which CoA shows a Trojan's Sword Slamm, that deals significant less damage, too.
Back to Vanaduke's mace:
The shields that I saw myselfe enduring his mace are:
-Grey owlite
-Omega ward
Both of these Shields have split ele/normal defense with fire resistance, but no stun resistance.
So I am curious if Volcanic Plate Shield can resist this mace since it already has both status resistances (but missing elemental defense).
Mon, 10/12/2015 - 09:50
#84
Alright, I actually feel like playing so I'll get you those pictures. Be mindful my VPS is only level 2.
Additionally, I'd like for someone to test whether armor affects shield overflow. Maybe the reason you didn't take damage when your APS and others broke from Vana's slam was because of your armor? Perhaps try with Proto or some other low-tier armor?
"So I am curious if Volcanic Plate Shield can resist this mace since it already has both status resistances (but missing elemental defense)."
From what I see, these are consistent values so I'll only be posting one picture, but I did take multiple if you'd like me to add in more.
Depth 24 direct slam:
http://i.imgur.com/KawVAmu.png
Depth 24 Attack Up direct slam:
http://i.imgur.com/Z0TpSZI.png
Depth 25 direct slam:
http://i.imgur.com/4xqF6ib.png
Depth 25 Attack Up direct slam:
http://i.imgur.com/nk6syoc.png
Depth 26 direct slam:
http://i.imgur.com/LBIZvRr.png
Depth 26 Attack Up direct slam:
http://i.imgur.com/7MnshMV.png
Depth 27 direct slam:
http://i.imgur.com/x6IuQIa.png
Depth 27 Attack Up direct slam (sorry, I botched this picture and only took one):
http://i.imgur.com/scG1uLp.png
Depth 27 direct slam (White Trojan):
http://i.imgur.com/ao8RSr5.png
Depth 27 Attack Up direct slam (White Trojan):
http://i.imgur.com/LxyTW40.png
Depth 28 Lord Vanaduke's direct slam:
http://i.imgur.com/M5oOwb2.png
PSA: Trojans are bullies
Also tfw I can't even solo Vana anymore.
EDIT: noticed a few funny things though, including that Vana does the same damage to VPS as a White Trojan on Depth 27 does.
Tue, 10/13/2015 - 23:42
#85
@Jenova
My new theory is that status is the thing that actually causes the shield to break. Remember my post about that strange phenomena with Dragon Scale Shield? It got kinda 3 different Sword Slamms:
1. Sword Slamm breaks shield to exactly ±0 and causes stun
2. Sword Slamm breaks shield to exactly ±0 but causes no stun
3. Sword Slamm does not break the shield. Ofc no stun gets applied in this case.
I had a theory a while back that was something along the lines of the status dice being rolled, then damage and status is applied, and then the status roll result is modified by your armor to determine its application or not. The primary theory of the thread was debunked, but this secondary theory still seems plausible.
If this theory is extended to shields, then it could mean that:
1. A high status roll would successfully deal enough damage to your shield to break it, and the high roll value would cause you to be inflicted with the status despite your inherit status resistance.
2. A low status roll would successfully deal enough damage to your shield to break it, but the roll fails once it is modified by your inherit status resistance
3. A failed status roll would mean that your shield survives because the damage alone isn't enough to break it.
A summary of my theory is that damage is applied like this:
- Roll a status number from 1 to 6
- Subtract from the shield the amount of damage dealt (after reduction by shield defenses)
- If the rolled status number is greater than A, subtract from the shield X status damage (after reduction by shield resistances)
- If the shield is broken by this attack, obtain damage value overflow. If the shield had greater than M health before this attack, overflow is 0.
- Subtract overflow from knight health
- If the shield is broken by this attack, and the rolled status number is greater than A+K, where K is the knight's resistance to that status apply status to knight.
A is an value affecting the frequency of a status's infliction, it is different depending on the attack and the true chances aren't known.
X is an value not currently known, but likely dependent on the strength of the status.
K is the knight's status resistance. It will always be greater than 0 unless the knight has -4 points of weakness to the rolled status.
M is the maximum damage a shield can take in a single attack. I hypothesize that it is equal to the health possessed by a level ten 5 star shield (scaled to the current level)
Tue, 10/13/2015 - 14:40
#86
↓
@Hexzyle
In my dealings with whisps, status has only ever dealt one set damage to shields (varying by depth).
I find it more likely that the situation is explained by a trojan buff.
I ran into a lot of buffing issues with using wolvers (since alphas have a buff), despite their attacks being pure pierce damage (on elite/T2+ if I recall correctly).
Tue, 10/13/2015 - 15:19
#87
As I recall, Tier 2 monsters deal split 50/50 Normal/Special, even in Elite, it's only in Tier 3 where certain attacks go from split damage to pure damage (Kat Bites are still pure physical, but Kat Bullets go from 50/50 Normal/Shadow to 100% Shadow).
Also, don't Wisps always deal ~(8.0, 8.5) damage of status? Eight points is not enough to resist them, but I recall someone saying that 8.5 was enough to resist them (I only know that 9 is enough to resist them, not sure about the 8.5)
Tue, 10/13/2015 - 17:45
#88
↓
@Dibsville
I urge you to test the beast damage yourself. I seem to recall them being more well behaved than other family attacks.
As for the resists... I'm stating that it has always been one value - I didn't specify the value (Zeddy has in some thread).
Hexzyle seems to be suggesting that it can deal 2 different values to shields to explain the 3 observed outcomes of trojan attacks.
Also...
@Hexzyle
I took a glance at my disorganized data and on 249 I have a sage shankle dealing 9 damage past the proto shield's hp in one attack from full health. This refutes your "step 4". I thought it might be unique to the proto shield or even 0* equip...
Too bad I didn't have any data... so I went to FSC with a force buckler (1*, 175hp at that depth) and got hit by some wheel launchers. Shield broke and 87% hp left.
2* defender (284hp) 97% hp left.
I thought maybe I could do a bit of testing while GitM is the current prestige mission with those lost souls...
3* SSB (425hp+norm low UV) 3 attacks, 3 breaks + 3 times always shocked (wearing chaos).
3* SSB (same as above) 3 attacks, 3 breaks + 3 shocks (wearing divine)... maybe not enough shock resist for them...
3* SSB (same as above) 3 attacks, 3 breaks + 0 shocks (divine + shock high perk)...
I'm going to draw the conclusion that it deals damage, then status (yes/no), then armor does a standard resist against the status (roll chance depending upon status vs armor, not shield).
Wed, 10/14/2015 - 00:14
#89
Hexzyle seems to be suggesting that it can deal 2 different values to shields to explain the 3 observed outcomes of trojan attacks.
In my dealings with whisps, status has only ever dealt one set damage to shields (varying by depth).
I am suggesting that because Trojans have a chance of not inflicting stun after breaking a shield, inflicting stun after breaking a shield, or not breaking a shield at all.
Whisps don't have a chance of not inflicting status, they likely always roll the maximum value possible: that is why you only observe one set of values. This seems to further my hypothesis.
I took a glance at my disorganized data and on 249 I have a sage shankle dealing 9 damage past the proto shield's hp in one attack from full health. This refutes your "step 4".
"Full health" is referring to my previous comment
- What I am thinking is perhaps there is a maximum damage that an enemy can do to a shield. Perhaps that maximum damage is a 5 star shield's health on depth 28. So therefore if an enemy attacks a full health 5 star shield, no overflow is applied.
Proto Shield isn't a 5 star shield. I've updated my comment to make it less ambiguous what I meant.
Wed, 10/14/2015 - 05:54
#90
Here are the data what I got out of Dib's screenshots:
Overhead Sword Slamm vs Volcanic Plate Shield (heat level 02):
Source: _ _ _ Sword Slamm: Buffed Sword Slamm:
D24 (Black):_ 242_ _ _ _ _ 109
D25 (Black):_ 245_ _ _ _ _ 110
D26 (Black):_ 248_ _ _ _ _ 111
D27 (Black):_ 251* _ _ _ _ 108*
D27 (White):_ _88_ _ _ _ _ -9% Knight hp
D28 (Vana.):_ _87_ _ _ _ _ ---
___
*Screen with shaking effect
@Dib: well vana and that white trojan differ by 1 pixel. So is it a rounding thing or do they deal different damage? But it is interesting that both hits deal verey similar damage to VPS. My theory is that VPS will protect better vs white trojan as further it is heated.
@Hex's theory about a 6-side-dice:
Nope. Not 6 sides. This would mean that status could be messured in 6 different health bars on a not-breaking shield. To be honest, this is what i actually expected when I initially started testing, but the tests show obviously different: the shield health is usually the very same amount of pixels. It is just a recent occurency that shields like CoA and DSS show TWO different values (including DSS, there MIGHT be a 3rd one, but it could also be a kind of second roll vs the amor after the shield is broken).
If you search for a multi-sided-dice then I guess it is more plausible to assume a 2 sided dice (status infclicted yes/no) - though one side would be heavier.
tldr: I guess that skeptic's assumption of another roll vs amor after shield broke is plausible.
Wed, 10/14/2015 - 07:21
#91
@Hex's theory about a 6-side-dice:
Nope. Not 6 sides. This would mean that status could be messured in 6 different health bars on a not-breaking shield. To be honest, this is what i actually expected when I initially started testing, but the tests show obviously different: the shield health is usually the very same amount of pixels. It is just a recent occurency that shields like CoA and DSS show TWO different values (including DSS, there MIGHT be a 3rd one, but it could also be a kind of second roll vs the amor after the shield is broken).
The 6 sided dice was just an analogy. (also because the way the statuses and resistances work in multiples of two, grouped into three levels, I thought the number 6 would be a nice number to use)
If you know much about status infliction, you'll know that statuses are always inflicted at the same strength every time, and that resistances reduce the strength of statuses by a flat rate. Zeddy has a bunch of data regarding this.
In my comment, I use the variables A and X to represent infliction chance, and status strength, respectively. They are not correlated. Resistances reduce both infliction chance (by modifying A) and status strength (by reducing X), but randomness only affects the former, not the latter. If a status is applied, it will always be applied at the same strength, and thus always deals the same damage in the case of shields.
Anyway, I don't consider my hypothesis to be absolute, because I know that Trojans and Lumbers apply stun with a very high frequency: if this is a 100% chance of stun at inherit knight resistance levels, then why can you sometimes escape a shield-break from these attacks without being stunned? If it indeed is a 100% chance of stun, then this might suggest that status overflow is also reduced by remaining shield health or something.
Wed, 10/14/2015 - 14:44
#92
↓
@Hexzyle
You sure come up with a lot of testable ideas.
The difference in our hypothesis appears to be the reduction of status duration/strength.
This is testable:
Measure duration of stun without using shield and compare it to duration when shield is broken + stun applied.
From the images, it appears that full DS along with shield yields 5s of stun on D24.
I got close to testing this with the shocking lost souls, but I didn't check if the damage from the explosion took out the entire shield's hp or not (to have any reduction).
The other difference is a bit pedantic - I say the shield then armor rolls while you say the status rolls as many times as needed.
Mon, 10/19/2015 - 08:20
#93
Very important factors to consider
Current analysis
All 3 plate shields have 4 "points"
The shared points are shield HP, Normal resist equal to VPS, and Stun resist
Variation among shields
IPS spends it's last point on Pierce resist
VPS spends it's last point on Fire resist
APS spends it's last point on extra Normal resist
GT has a tradeoff of MSD and shield bash damage. Technically this evens it's points out to 1. So what did the GT spend it's last point on (if at all)?
I thought about this for a while.....
So I have come to ask something different. The current studies on the GT have compared it's defensive power; so what about it's offensive power?
Is the shield bash damage greater than other tortoshields?
Is the stun radius larger?
Is the Stun from the GT more powerful than other tortoshields?
Is there any factor that sets it above other shields? (faster regen, faster bash shield, etc)
If there is no distinguishing factor then I would say the GT is about as useful as the 4* plate shield..... which is insulting to a 5* shield....
Tue, 10/20/2015 - 05:11
#94
"Is the shield bash damage greater than other tortoshields?
Is the stun radius larger?
Is the Stun from the GT more powerful than other tortoshields?
Is there any factor that sets it above other shields? (faster regen, faster bash shield, etc)"
I can tell you from both personal experience and extreme testing of the shields that Grand Tortoise is exactly the same as the other shields, and also the same as the other Torto shields. Unless that was recently updated (which I doubt considering all the things that should be updated first before Grand Tortoise has some special effect), GT is the same as your regular shield, just has the bash explosion and hitbox, and the faster shield bash (as I recall, but I haven't tested it since release).
So as it currently stands the only notable reasons to use GT are its faster shield bash, which also has a Stun explosion at the end of it. Otherwise, even Volcanic Plate Shield is better, even in areas that are not Fire stratums, because VPS does not come with the MSD: Low.
Mon, 04/04/2016 - 03:30
#95
Continuum
Edited post #72.
I plan to finally finish the project of collecting data of Grand Tortoise from each heat tier of all 5 different Trojans in FSC within the next days (or until hitting a radiant wall, but I definitelly finish the 4* version and a few levels of 5* before mid april).
And yet again: you can help, too. Yes, you who read this post!
Grab your shield screen some data, upload it unscaled and post it here ;)
Wed, 05/24/2017 - 02:19
#96
I hereby call upon the dark powers and resurrect this thread to roam through the front pages of the forum once again!!
aka: Updated post #72!
Finally finished the Mighty Shell test vs all FSC Trojans. Due to the fact that no one added here, I assume that there is still no detailled Shield number gathering in Spiral Knights.
I would like to actually make a damage collection for shields similar to damage numbers for weapons as we already got them inside wiki which is would give data for each monster how much a certain attack does at the first and last level of the depth of a stratum to each shield (oh god, that will be some data collecting, hahaha).
After looking how little Mighty shell does change between the first and the last trojan in FSC, there would not be much need for actually gathering each heat level seperatelly as I did in post #72. But condition should be though to have a heat 10 shield, to keep things compareable.
TL;DR look at post #72 and use it as base for new discussion.
Wed, 05/24/2017 - 06:17
#97
@Jenova
I would be interested in gathering numerical data for shields. It might be an idea to make a new thread for it, however.
Do you know if there is any data on monster attack damage? That would prove very useful.
It might be the case that using the numerical data of a pure damage weapon (e.g. Sudaruska, Final Flourish/BTB, WRH etc) against the guardian shield to attempt to determine how much defence the guardian shield has. This can then be translated to monster attacks against shield with similar defence.
This is just a rough idea and I'll get back to it later. Thanks for continuing updates!
Thu, 05/25/2017 - 08:55
#98
Status damage to shields
Skepticraven post #88
I'm going to draw the conclusion that it deals damage, then status (yes/no), then armor does a standard resist against the status (roll chance depending upon status vs armor, not shield).
I want to determine if your thesis is correct which says that there is a status roll (yes/no) while dealing damage to shield's health and a second status roll (yes/no) in case the shield gets broken. By calculating that, let's assume that status is beeing added to the formula after the raw damage is finished.
To structure this thesis:
1. Shield health gets reduced by raw damage calculation, excluding status
-If shield breaks during this process, then overflowing damage will be done to Knight, continue with step 3
2. Only if a status roll results in a status inflict, then shield health gets reduced by Status strength
-If shield breaks during status calculation, continue with step 3, otherwise the attack got successfully blocked
3. After shield is broken, there will be a new status roll that can either apply or not apply to Knight.
To test this thesis, we need to gather data of shields with no Stun resist against Trojans' overhead slamm while the shield recives less and the shield recives more damage. If this thesis is correct, then the difference will be always the same.
In the test with CoA, the difference between that weak and strong hit was 167-30 = 137 pixels in depth 24 of the FSC mission.
Edit:
Thesis proved: in depth 1 that stun-status damage to shields of an overhead trojan smash is always exactly 72 pixels to 5 star shields.
The shocking-status part of the Attack Buff deals also exactly 72 pixels to 5 star shields at D1.
And testing overhead slams' stun-less hits is a real pain since almost each slamm includes stun and you can sometimes wait 10 minutes before you get one without stun - and I usually take 3 or more screens per value to be safe.
Thu, 05/25/2017 - 15:44
#99
↑↑↓↓←→←→ba
@Jenovasforumchar
It would also be interesting to test the converse:
An instance where the shield would have nullified the status, but broke from the normal damage... and then the status was applied to the armor.
This would tease out the point at which status was applied (during damage reduction on the shield, or after damage altogether).
Fri, 05/26/2017 - 10:25
#100
Nyan :3
@Drischa
You could fill in data to that wiki page I recently added. I believe it is the best way for people to compare directly how well a shield will do against monster x attack y.
That part about lock down has been done by Skeptic and Zeddy already, as he stated in post #9 of this topic.
What shields do you have heated at 10? (No UVs if possible)
@Skepticraven
Status has to be rolled two times, otherwise a DSS would show differently.
Btw, I started a new wiki post here with recent data collection, but it is far from finished:
http://wiki.spiralknights.com/User:JenovaSPIRIT/Shield_Damage_Chart
If you have a look at the data from an overhead slamm at depth 1, then you can see that a DSS will either break or it will have 43 pixels left.
The strength of stun status is exactly 72 pixels to 5 star shields in depth 1, so it means that DSS' health is theoretically at minus 29 pixels.
But these 29 pixels above the shield do not always result in reciving the stun status. Often the shield breaks (aka, that slamm had stun inside) but the knight has no stun at all. So after the shield broke there has to be a second, independent status roll after the first status roll against the shield's health.
Also it is obvious that damage gets calculated before status gets calculated. If it was the other way round, then you would get damage worth of 29 shield health's pixels, but you don't get any damage at all.
Edit:
Also something akward that I came along while looking at the numbers:
T1 Trojans' Overhead Slamm deal split shadow/normal damage while T3 Trojans' overhead slam deal pure normal damage. Therefore it seems that this specific white D27 Trojan are in fact a T1 trojan with stats according to D27 instead of a completelly new version. It just got lost and accidentally took an elevator down to FSC from his home in T1 and Vana trapped it forcing it to lurk in D27.
It is strange though that T1 got a split damage while T3 Trojans do not.
Another interesting thing:
DSS got in both, D1 and D24, exactly 29 pixels too less to endure an overhead slamm with stun attached. I think it is coincidence, but yet it is strange that it is 29 in both cases. Filling the chart will definitelly reveal more.
The brackets denote the inclusive/exclusive interval. (262,263] excludes {262}, but includes {263}.