test content
Logo
What is the Arc Client?
Install Arc
NeverBleed v0.1
kurahavi
Member Posts: 87
Because I have been cursing rogues in t2 for a while, I decided to release my not so awesome tool. What this thing does is, it parses combat log and finds dueliest's flurry bleed, and reports that damage in a small window. I tried to make it to work as on top of all windows, which should work if the game is run as a borderless window.
to get combat logs showing, just type in game in chat box: /CombatLog 1
After that in game folder/Neverwinter/Live/logs/GameClient/Combatlog.log file should exist. When you open the program it asks for that specific file, so you must pinpoint that file for the program.
The program parses the last lines every second, and tries to find duelist flurry bleed in it and reports this bleed damage per target both as true damage (doesn't include armor pens of any kind) and as the damage that you can see. Meaning if there was bleed before the program was started, it doesn't show it, only bleed ticks after the combatlog was given.
The program should save its position and sizes after closing it down, so you might want to first adjust those and close it.
Edit: Had a small bug showing the ticks from 6 to -3 instead of 10 to 0. Also the resistance was counted wrong way it was:
100 * ( 1 - ( true damage / damage ) ) opposed to 100 * ( 1 - ( damage / true damage ) )
The version is now 0.17
http://filebin.net/nx964xfbck
Direct:
http://filebin.net/nx964xfbck/Neverbleed_v017.zip
Also a virus scan for the file:
https://www.virustotal.com/en/file/76eb361f981e3e1ba462897dddfe0ab0efca472d6cb57901568bc1f6b516dacc/analysis/1370577333/
Git for source, quite not much comments + terrible code.
https://bitbucket.org/Kurahavi/neverbleed
If this doesn't work then either I made some mistake with .net 3.5 or something else that I have no clue for.
1. column is estimated ticks left for bleed,
2. column is for estimated stacks ( this number adds up every time the bleed damage changes ). The true stacks can be higher than what this column shows.
3. column is how many times this bleed damage has ticked, can be used for checking if all of the bleed stacks are in.
4. column is the name of the target
5. column is for the true damage the game reports, damage with no armor / debuff / armor pen counted, this is the damage you should be watching to determine if the bleed was good or not.
6. is the resistance % calculated as 100 * ( 1 - ( damage / true damage ) )
7. column stands for damage shown for the player.
8. column shows 'X' if the bleed is critical and empty if not.
- The application checks the file only once per second, so there might be slight lag on how many stacks there are or how much the damage is hitting.
- You should be able to move around the window by left clicking on the black area and moving mouse.
v0.16
- The program checks the combatlog-file now within 400ms intervals. I do hope it is not too often for some computers, but from my observations, 1s intervals is too rare to update the damages.
- Right clicking the black area of the screen, you can open filters menu. From filters menu, you can add a filter, and give it a color. The filter checks the line if it contains the string. In other words, if you put word bleed in it, it will take ALL of the bleed lines. So make sure you use something to recognize it such as player name or account name. Also IMPORTANT, DO NOT put something in it that is not in bleed lines of the combat, such as other skill names. It does not include things other than duelist flurry bleeds.
v0.17
- Fixed a bug that came with the flags thing. The flags really showed whatever they wanted, for example: "CFFFFFFFFFFFFFFFFF" etc. Now it should work correctly
- Added a font change option to increase size or style.
- Added an option to try to adjust the interval for checking the log file, altho the minimum is still 400ms, but you can make it happen less often.
to get combat logs showing, just type in game in chat box: /CombatLog 1
After that in game folder/Neverwinter/Live/logs/GameClient/Combatlog.log file should exist. When you open the program it asks for that specific file, so you must pinpoint that file for the program.
The program parses the last lines every second, and tries to find duelist flurry bleed in it and reports this bleed damage per target both as true damage (doesn't include armor pens of any kind) and as the damage that you can see. Meaning if there was bleed before the program was started, it doesn't show it, only bleed ticks after the combatlog was given.
The program should save its position and sizes after closing it down, so you might want to first adjust those and close it.
Edit: Had a small bug showing the ticks from 6 to -3 instead of 10 to 0. Also the resistance was counted wrong way it was:
100 * ( 1 - ( true damage / damage ) ) opposed to 100 * ( 1 - ( damage / true damage ) )
The version is now 0.17
http://filebin.net/nx964xfbck
Direct:
http://filebin.net/nx964xfbck/Neverbleed_v017.zip
Also a virus scan for the file:
https://www.virustotal.com/en/file/76eb361f981e3e1ba462897dddfe0ab0efca472d6cb57901568bc1f6b516dacc/analysis/1370577333/
Git for source, quite not much comments + terrible code.
https://bitbucket.org/Kurahavi/neverbleed
If this doesn't work then either I made some mistake with .net 3.5 or something else that I have no clue for.
1. column is estimated ticks left for bleed,
2. column is for estimated stacks ( this number adds up every time the bleed damage changes ). The true stacks can be higher than what this column shows.
3. column is how many times this bleed damage has ticked, can be used for checking if all of the bleed stacks are in.
4. column is the name of the target
5. column is for the true damage the game reports, damage with no armor / debuff / armor pen counted, this is the damage you should be watching to determine if the bleed was good or not.
6. is the resistance % calculated as 100 * ( 1 - ( damage / true damage ) )
7. column stands for damage shown for the player.
8. column shows 'X' if the bleed is critical and empty if not.
- The application checks the file only once per second, so there might be slight lag on how many stacks there are or how much the damage is hitting.
- You should be able to move around the window by left clicking on the black area and moving mouse.
v0.16
- The program checks the combatlog-file now within 400ms intervals. I do hope it is not too often for some computers, but from my observations, 1s intervals is too rare to update the damages.
- Right clicking the black area of the screen, you can open filters menu. From filters menu, you can add a filter, and give it a color. The filter checks the line if it contains the string. In other words, if you put word bleed in it, it will take ALL of the bleed lines. So make sure you use something to recognize it such as player name or account name. Also IMPORTANT, DO NOT put something in it that is not in bleed lines of the combat, such as other skill names. It does not include things other than duelist flurry bleeds.
v0.17
- Fixed a bug that came with the flags thing. The flags really showed whatever they wanted, for example: "CFFFFFFFFFFFFFFFFF" etc. Now it should work correctly
- Added a font change option to increase size or style.
- Added an option to try to adjust the interval for checking the log file, altho the minimum is still 400ms, but you can make it happen less often.
Post edited by kurahavi on
0
Comments
Do you have any sort of "guide" to link me in order to understand a bit more how does bleeds work?
For the program, I noticed one small bug, where the first column started from 6 seconds and went on to -3 when it was supposed to start from 10 seconds and go to 0. The time is taken from last time someone hits duelist flurry to that mob, so it might not be completely accurate when the stacks wear off.
So the way bleed works for what I've seen is: First off you will attack 2 times the normal attack with duelist flurry and third one starts hitting attacks called Duelist's Flurry Hit, and in every flurry hit, there is a chance to apply 1 stack of bleed. From my observations and parsing combat log, the bleed damage is randoms again every time the bleed stack is applied. Once there is 10 stacks of Bleed, the Bleed damage remains the same for 10 seconds, or if you hit Duelist Flurry Hit to the target it applies stacks again, but this time it does not increase the stacks, but keeps it at 10, and the same bleed damage has again 10 seconds timer on it.
So the Bleed that gets hit on 10th stack is the most crucial one. The damage on this Bleed will remain as long as you do Duelist Flurry Hit during those 10 seconds, and so you can keep this bleed damage as long as you do the Duelist Flurry Hit within 10 seconds of doing a last one. In other words, you might actually keep having 22k Bleed ticking from the start of the boss battle to the end of it, or in a worst case your Bleed might be ticking 500 damage from beginning of the fight to the end, if you do not let the Bleed damage reset.
Now people keep saying that you can apply 20 stacks of bleed to a target with 2 rogues, but sadly from 10 onwards, the stacks mean nothing. It just keeps applying the same Duelist's Flurry Bleed damage as it did on the 10th stack. The only reason why you would take second rogue in to your party is to maintain a lucky good crit bleed. Since it is easier for 2 rogues to keep hitting Duelist's Flurry Hits within 10 second interval than one rogue, the bleed damage remains more likely on the mob with 2 rogues than one. This also means if 2 rogues are maintaining a bad bleed on a rogue, it really is harder to let the bleed damage reset.
I hope this clarifies things.
Thank you for taking some time to write the post. It surely explained me a bunch of things but opened a few more questions. Is there any easy way to tracks the ticks of the bleed (not using your program)? Is there any easy way to keep count of how many stacks does the target have? Also, is it hard to maintain the 10 stacks on mov. heavy fights?
Again, thank you for uploading the tool you have made. I was not able to download it, since I got the following error:
"Uh oh,
The tag you are trying to access does not contain any files. The tag is read only."
For keeping a number how many times the damage has ticked is possible, but its quite hard. I guess easiest way to do it would be setting a timer every time someone hits Duelist's flurry Hit ( the third attack ) for 10 seconds, as a timer how long you have time to attack again. As for how high the bleed is, you can see it by stopping the attack for a second and seeing the number ticking. If the bleed is caused by other rogue on the other hand, you would have to enable in game so that it shows every damage in your party in your screen, and trying to find the bleed hit from all that, well... I guess it is not really humanly possible.
As for keeping stack count, well you can easily see the amount of stacks next to the mobs name. There is debuff and my application doesn't keep count on stacks. I guess it would be possible somehow maybe add such feature, but can't really for now think of a way. As for the tag being empty, I suppose it was due to the tag being private, I changed it into public. If that didn't help here is a direct link http://filebin.net/nx964xfbck/Neverbleed_0_12.zip. If that doesn't work either, then I have to figure out another file upload site thingy.
Unfortunatelly I am not able to test this since I won't be able to log into the game for about 2.5hs or so, but I'll try this as soon as I can. I'll try to take a loot into the stack count and tick damage and will try to min-max a few things, it may not be worth at my level (36 or so) but it will give me the habit of doing so for lvl 60 dungeons.
I was able to get the file, I'll give it a test tonight.
Thanks again!
Each hit of the flurry that applies a bleed adds some portion of the flurry hit's damage to the prior bleed total, then hastens the bleed tick to be immediate. It isn't actually applying a "new" bleed, it simply increases the damage of the bleed already on the target and makes the next tick happen instantly, which resets the timer.
Also the app works great for me. The only 2 other features that maybe would be useful are a crit flag, and a way to have it selectively parse only your own bleeds (IE it will parse all bleeds if there are multiple rogues nearby). Also, is it written in such a way that I could easily expand it to track deadly momentum and wicked reminder stacks? IE is it capable of tracking custom skills without tons of hassle, or just watching for "Duelist Flurry Bleed" in the log and parsing those lines? Thanks a lot!
@matutin7 If you resize it one way, then downsize it the opposite, then the window will "move".
PS: There seems to be some bug that causes it to occasionally reset the tick count for no obvious reason. Edit: This was due to combat dropping. If a bleed is still going when combat drops the app resets the tick count.
You can also see a better formatted version of the guide here.
I dont get it, this is for every new stack before the 10th stack or just everytime the flurry applies a bleed? If so, everytime you flurry you should get a bigger bleed dmg rolling.
Also, thank you for pointing me how to move it. It was indeed an stupid question.
You can also see a better formatted version of the guide here.
Ah got it. So every flurry hit (that applies a bleed) after the 10th stack just makes the bleed tick instantly, so there is no way to "clip" the bleed dmg.
You can also see a better formatted version of the guide here.
I really like the idea. I used to play another mmo and my 2 pref chars were 2 melee classes, both relied heavily on bleed dmg and both had a way to extend the bleeds, but usually it clipped the dot in a way or the other.
Or maybe you can post it on some colaborative site like sourceforge or google code to allow contribution from other coders. This could speed up feature inclusion and bug tracking.
Countdown goes from 10 -> 7, then back to 10 -> 4.
https://bitbucket.org/Kurahavi/neverbleed
The 0.14 link is on the first post.
I tried to fix the tick thing not to reset anymore unless the bleed damage doesn't appear within 2 seconds. This also means if you hit a new bleed in after the bleed resets BUT its still within 2 seconds of the last bleed tick, the application assumes that the bleed continues and starts adding stacks to it. Also this means if the bleed damage now changes the ticks do not reset unlike in the original version. I guess this is the more logical way how it should work.
About having to parse players out, honestly I cannot think any other reason than doing non-dungeons, and then again would one really need the bleed ticks in non-dungeons. Since I am a lazy person I didn't go add a feature that I would find defeating the purpose of the whole application. The point is that you actually know the bleed tick on a boss monster even if you have 2 rogues. Maybe in a future if i get less lazy I might do this tho.
Parsing hits other than bleed, well it might not be that hard, but personally I just feel that it should either a) be completely different program or b) different window (form). This is because I don't think one really needs to know the damage count from certain skills or status effects. Also the fact that Duelist's flurry and brand of the sun are both dots so they work bit differently compared to other skills. Honestly tho parsing usage of wicked reminder and giving it a small timer is probably easier than having to parse bleed, since the bleed timer is suppose to reset during duelist's flurry hit, and thus the timer is not necessary completely correct, because it assumes the timer is reset always on a last duelist's flurry hit.
Keeping tracks of how many stacks of wicked reminder a mob has is semi impossible to do accurately, since there is no indication if wicked reminder was used during stealth or not. The only possible way to bit distinguish between non-stealth and stealth is a flank hit, and you cannot really even count on that, since flanks do happen even when you're not in stealth.
As for the countdown going from 10 -> 7 and 10 -> 4... Maybe this is due to the program only reading the combat log once per second. Maybe the newer version fixed this, if not I would appreciate getting more information.
Edit: If the application works as it is supposed to be, you should be able to move the window around by moving the mouse while holding the left mouse button down on the black area (listview).
If you mean that once the stacks hits a count of 10, the damage doesn't change.... well the thing is, the damage does not change after the target has 10 or more stacks. I can't help it that the way the game is made works like that. Which is why I have said that second rogue does not really bring that much to the group.
Or if you mean that there is a slight delay of updating the counts, that is due to the combat log being parsed every second, + there might be slight delay on the game writing the logs. I figured that should be good interval, not too fast so that it doesn't hang up computer and should not be too long delay.
Altho the resistance update is a bug I guess. Have to look at it.
Damage (player shown damage) is also stuck however. I'm guessing it's directly related to Res being stuck too.
Example: Stacking to 10 stacks. Reaching 8600 dmg per tick. Use a Wicked Reminder, now the damage shows 8900 on screen.
The -resistance and player shown Damage is not updated however, it's still stuck on whatever the value was when it reached 10 stacks.
It updates just fine while it's stacking, but after reaching 10 stacks it doesn't update just like TrueDmg doesn't.
Ty for reporting, fixed the small bug and also that wrongly blinking if it crit or not thingy.
I tried to figure out how the bleed damage is calculated but the critical overrun makes things a hassle + the fact that there is not 100% certain way to know the exact number of stacks for deadly momentum when the hit happens.
I used low level daggers to get as small as a possible for damage variation. This helped me to identify the overrun damages easier from combat log, using flourish, that actually has 4 different skills. With the equipment I have, the 3rd hit hits roughly 3.5% harder than first or second, and the fourth one 18.5% harder than the first or the second one.
Also critical overrun bleed can actually hit harder than bleed stack that is applied after this. Also if you apply 6-7 stacks first, and let a low level rogue to apply the rest of the stacks, the bleed hits about as hard as if the rogue had applied all 10 stacks by himself.
The critical overrun hit with flurry hit about 4.5% harder with each 15% CS on non-crit hit. Cannot say the values are absolutely correct since the damage varied only few points per stack + they kind of overlapped on different stack counts.
75% CS = 1.224 * original damage
90% CS = 1.269 * original damage
105% CS = 1.328 * original damage
120% CS = 1.373 * original damage
135% CS = 1.418 * original damage
150% CS = 1.462 * original damage.
The values above are rough estimates, and the numbers in middle are more of interpolation between full stack and 0 stack ones.
My guess would be that the bleed damage is some random value * bleed stack count, as in 10th stack * for example 100-200 => 1000 - 2000 damage. The reason why I think this is randomed is, because even if the damage is about same between flurry hits that add bleed stacks, the damage change doesn't seem to follow the flurry hits. On the otherhand if the flurry hit that applies the stack happens to crit, the bleed also crits. Also if the critical overrun is on while the flurry hits, it seems to apply extra damage on the bleed aswell.Also it is possible that the bleed damage is additive as in every stack applies some base damage to the bleed, but that would not explain why 6th stack bleed can tick harder than 10th if the first 6 ones are applied by better geared/leveled rogue.
Edit I should probably test with the low level daggers too, what values it gives for the overrun crit. The 22% was with purple (not so good) lvl 60 daggers.
With lower level daggers the critical overrun affected roughly same amount with 75% CS. The damage was about 22.5% higher from the CO.
You can also see a better formatted version of the guide here.
Gotta admit that you seem to be right. It adds 30% of your critical severity to the next attack as extra damage. Also it does work on non-crits too.
The sun flourish consists of 4 skills, 1st and 2nd ones hit about same, 3rd one about 3.5% harder, and 4th around 18.5% harder than the first one.
With my level 5 daggers
1st damage varied between 204 - 214, avg 209.5 n = 84 only counting non criticals, non-flank ones
2nd damge varied between 204 - 213 , avg 209.3 n = 44
3rd damage varied between 212 - 223, avg 217.6 n = 60
4th damage varied between 245 - 257, avg 251.9 n = 78
Now the damage in real actually varied higher, but the thing is, for example first hit between 214 and 250, there is no damage count, only from
1st 250 - 262, avg 256.3 n = 61, which is supposed to be 1st hit, non crit critical overrun damage.
2nd 251 - 263, avg 256.7, n = 26
3rd 260 - 273 avg 266.9, n = 60
4th 301 - 317 avg 308 , n = 50
Now if we count using for example 1st hit lowest damage:
250 * 1.75 = 437.5, which is about what the crits hit.
Applying your formula:
(Damage + (Damage*0.3*0.75)) + CS
( 204 + ( 204 * 0.3 * 0.75 ) ) * 1.75 = ( 1.75 * 204 ) ( 1 + ( 0.3 * 0.75 ) ) = ( 1.75 * 204 ) ( 1,225 ) = 437.325
So our results are pretty much the same.
If we now assume I would have 5x deadly stacks = 150% CS, with some testing I hit around 300 ( 298,75 ) damage on dummies, so that makes it 300 * 2.5 = 750 if it were a crit.
( 2.5 * 204 ) ( 1 + ( 1.5 * 0.3 ) ) = ( 2.5 * 204 ) * 1.45 = 739.5 So the formula seems to actually work quite well, even tho the critical overrun can be applied on non-crit attacks aswell.
You can also see a better formatted version of the guide here.
I've had instances where for example:
The true damage is 4600. Resistance is 0. Player shown damage is 4600. It's a crit.
Obviously the true damage can't be 4600 if it shows a 4600 crit on screen.
Hmm if you hit dummies you're doing true damage. Maybe the name of true damage is a bit misleading. When you watch the log in the chat-window. It shows damages something like this (cant remember fully): Lashing blade 2000 (2200), the number inside the brackets is what I call true damage, and if there is no bracket thing shown, I assume it is doing "true damage".
In the case you mentioned, if it is showing 0 resistance then it should be doing true damage. Although I can't say 100% there won't be some kind of a bug in the application, but I feel like there is more of a misunderstanding with the term true damage. Maybe it should be called unamplified damage or something like that. Anyways if the chats log is not showing brackets with numbers, then the application is working correctly and the game is giving false data, if this is not the case, then there is definitely a bug in the application.
For now I've gone back to v0.14 since I primarily use this to keep track of truedamage.
Mobs do have armor in this game, like some mobs have around 22% armor. That is where armor pen comes for rogues. Now if your CW uses some abilities the resistance of the mob can go negative. On the otherhand apparently armor pen itself can never make
Hmm another possibility what I think of is the fact that the thing updates once per second, which now that I actually did finally have a chance to do some t2 dungeons with my rogue, I realized it is not frequent enough. Also I can now fully understand why someone wanted some way to filter the other rogue bleeds out. The 1s update time worked well when i wasn't one playing rogue, just checking out the rogues bleeds, but it becomes more obvious that 1s is too long when I myself played one. I thought about giving people a choice to adjust the timing themselves, but I figured it might actually cause more complications than fixing things, if people put it to too low value, it might just hang up the whole computer.
As for the filtering thing, I was thinking about some easy list where people can add strings to be matched, for example name@account or whatever is written in to the combat log and giving ability to change background color / text color to indicate your own bleed from others. I do not exactly know if this actually works or how long it takes to implement such, only time will show. Just to clarify, I don't mean custom skill parsing, but rather than having a filter with the duelist bleeds, who caused them etc.
So since I got to play rogue in a t2 dungeon, I again thought spending some time to figure out bleed damage. I think I actually finally got the correct formula for counting bleed. Most likely this formula will kill almost EVERY single rogue build out there.
What I did was I bought low level dagger, put it in main hand, and slashed dummies. As I thought from my previous observations, bleed damage is somewhat related on stack count. Now I didn't bother looking combat log after taking down several damage numbers, I could pretty much see that the damage has 4 different ranges. Normal, critically overrun normal, crit and critically overrun crit.
So I went to smash some lvl 35 mobs with lvl 4 dagger in my hand, and I noticed the true damage was actually same despite these mobs having 18% armor resist. I checked my arp and it is 5.5%. I noticed that during flurry you actually hit harder bleed, and this is due to arp, after duelist flurry hit is over, the damage returns to the non armor pen damage. This means, arp does affect, but rarely.
So I dropped few pieces of armor to see how armor pen affects, and I noticed the resist during the flurry dropped, but I also noticed that the damage was still the same, even tho i had 700 less power.
With this thought in my mind I went back to the testing are of hitting dummies and yup, as expected bleed damage is unaffected by power.
I took off all the armors, rings, belt whatever, and the bleed damage was exactly same. The only thing that affects bleeds damage is your weapon damage.
So basically the formula goes:
Weapon damage * stack * constant. By constant i mean its combined damage from str etc.
So how to get high bleed, hit crit + critical overrun on 10th stack while having lurkers on, and that is about it.
Very simple, this means stacks 1-9 has NOTHING to do with the end bleed damage. You can even hit 1-9th stack with crappy weapon, swap weapons back and STILL get same bleed damage on 10th stack. In other words power is kind of useless stat to dungeon rogues.
- Now you should be able to filter bleeds and add different color for the lines that match the filter. Just make sure you add something in filter that you want to be shown, such as your name or char@acc name. If you put something such as bleed, it will simply accept everything or such as duelist. You can even put charname-level to it. The filter itself is rather lazy way, so its up to you to input correct string. Also note that this IS case sensitive, for example "imakeubleed" is not the same as "IMakeUBleed".
- The application now checks the log file every 400 milliseconds to cause less delay on updating the bleed damages. I noticed it myself that 1 second update time is just too long interval. With 400ms you can pretty much see the stacks growing without it doing some huge leaps. I am afraid that this might actually cause some problems on some computers for too frequently checking a file, but I really hope it does not do that. If it really becomes a problem, I can always make some option to manually set the interval. For me it felt like 400ms is often enough and hopefully not too often enough.
- You can switch order of the columns by holding left click on the columns header.
- Now instead of using appfolder for saving the settings, the application writes settings on the folder where the application is in xml-format.
- Now saves the location of combat log, so that you do not have to give the same file every time you start the application.
Known bugs:
- Sometimes the stacks grow past 10, or it does not reset the stacks. This is because the timer is mostly an estimate how long you have left, since after you get max stacks on the target, there is no way to know when it actually tried to apply a stack as in reset the timer from reading the logs. So basically it is only assuming that the last duelist flurry hit always resets the timer. Another reason why this bug happens is due to the wait time for the tick to happen. The application waits for 2 seconds for the tick, because the times keep varying a bit like 0.9 seconds to 1.1 second. Also because the neverwinter doesn't seem to write down everything instantly, but does it for like maybe once per maybe few hundred ms, the data might not be readable instantly when 1.1 seconds have passed from last tick. I might try to improve the realibility of this a bit in some time in future.
1. Bleed damage only changes after 10 stacks via debuffs on the mob. So if he was seeing a change it was not due to ArP or using a flurry. Either he was not yet at 10 stacks, or he was applying a debuff of some kind.
Also it has long been established that the bleed damage itself ignores damage resistance. Most of us have accepted at this point that this is due to the actual application of the bleed taking it into account. He said he "tested" this by dropping a few pieces of ARP gear. A few percent of ARP (the most he could have dropped is 5.5%) is well within the random values you can get from the base damage. So this is not an accurate way to test how or if ARP affects the bleed damage.
2. You can test very quickly that this formula does not work. He is trying to use basically a linear formula to qualify the 4 damage states that we can do (normal, normalOC, Crit, CritOC). However you can do 2 minutes worth of testing and get numbers that break this simple linear "the 10th hit determines the damage by itself" idea. Case and point, these are all examples of 10 stack bleed ticks that I got with just a few minutes of testing, no buffs, no combat advantage, etc. Just sitting and stacking my bleed on a dummy:
2000, 2100, 2800, 2900, 3000, 3100, 4800, 5850, 6500, 8443. You could argue that most of these can fit into one of our four normal cases with respect to the range of damage values you can get randomly. However, it cannot account for the outliers. This is pretty simple proof that the damage is not in fact a linear formula related to the last hit's damage only. Also it is not quite so random as to assume that it's all multiplied by a random number like some have suggested.
3. Using his formula with my numbers you can quickly see how unlikely this formula must be. My weapon damage is 393-481 +200 for power. This ends up in 593-681 weapon damage. You can average that to 637 weapon damage. So we get 637*10*C = Bleed damage, or 6370C, for a noncrit. C could be a fraction, but where was the math to try to show this?
In other words, no, this is not how the bleed damage is calculated.
You can also see a better formatted version of the guide here.
For point 1 I actually posted combat log showing where I Only use duelist flurry hit to attack a monster and the armor drops on bleed hit during duelist flurry hit. This was in your thread.
For the damage, yeah I forgot a huge detail, definitely my mistake. I was supposed to say it is 1/10th of weapons total damage * 2 roughly. Meaning:
You got 2 daggers (( 393-481 ) * 2 ) = ( 786 - 962 ) note that my point was that power is not affecting the damage at all, it is not included into weapon damage IF I am right, if i am wrong on the other hand then yes. So yes I do admit I don't have definite way to determine the base damage.
So I made about ~40 1oth stack of bleed with gear on, ~40 times 10th stack during stealth, ~40 10 stack bleeds with only daggers on hand, and about 40 10th stack bleeds with no gear but 10th stack on stealth.
I may point out that I had varying amount of stacks in deadly at the point i put first stack in, but I had 5 stacks at the point i was hitting 10th stack. Also when I used stealth I used the stealth just at some point at random stack count, but still enough to have stealth on 10th stack.
The numbers I got from smashing the training dolls are below. The first number represent the 10th stack damage, and the number right to it would be:
damage / 2.55 for crit
damage / (2.55 * ( 1 + ( 0.3 * 1.55 ) ) ) for overrun critical crit
damage / ( 1 + ( 0.3 * 1.55 ) ) for overrun critical non-crit
On right side of the columns we see damage without gear on except daggers, no armor, no rings, no necklace, no shirt no pants nor belt.
damage / 2.80 for crit flank
damage / (2.80 * ( 1 + ( 0.3 * 1.80 ) ) ) for overrun critical crit
damage / ( 1 + ( 0.3 * 1.80 ) ) for overrun critical non-crit
"2000, 2100, 2800, 2900, 3000, 3100, 4800, 5850, 6500, 8443."
If all of these are 10th stack bleed damage with equal CS (always same amount of deadly stacks and no cw or anything to boosting it) and none of them are flank ones, I will admit that I was wrong with my thought about the damages.
Edit: Ordered the data into ascending damages. Also corrected the flank normalised ones, I forgot to add extra 25% cs from using stealth.
Edit2: Still had wrong multiplier on cs using 175% CS when stealthed although it should be 180% CS. Although with these changes the damage seems to go to too low when counting flank and CO or crits. Flank crit => damage / 2.8 gives range of ~2450 - ~2800 where as pure flank was in range of ~2650 - ~2900. Same thing seems to apply for crit CO and CO 10th stack hits aswell. But when comparing to non-stealth ( non-flanked ) ones the numbers seem to be in a similar range.