Please upgrade your browser for the best possible experience.

Chrome Firefox Internet Explorer
×

Absorb with Minimal Defense Sin Tank

STAR WARS: The Old Republic > English > Classes
Absorb with Minimal Defense Sin Tank

fire-breath's Avatar


fire-breath
07.09.2013 , 07:44 AM | #31
Okey ....... I don't want to spark this 'discussion' further. I do have a question for KBN however.

How can you find how much damage is done by bosses before mitigation and such? Are you reverse engenering the damage in combatlogs to the unmitigated damage while taking your current stats into account? Or is there an other, maybe easier way to see the real damage?

Secondly, presuming that the damageprofile of TWT and DG is different (set-aside which one is the tankcheck or not). Wouldnt it be a good idea to have a second set of gear ready to switch around? I know that one maintank in my guild is using different sets of gear for Scum and TfB. In theory you could use this per boss instead, if you are willing to constantly switch gear ofc.
Afterall, if someone is tanking TfB NiM I would assume getting 69/72 tokens for specoptimalisation isnt to big a deal.

THoK-Zeus's Avatar


THoK-Zeus
07.09.2013 , 08:26 AM | #32
Quote: Originally Posted by fire-breath View Post
Okey ....... I don't want to spark this 'discussion' further. I do have a question for KBN however.

How can you find how much damage is done by bosses before mitigation and such? Are you reverse engenering the damage in combatlogs to the unmitigated damage while taking your current stats into account? Or is there an other, maybe easier way to see the real damage?

Secondly, presuming that the damageprofile of TWT and DG is different (set-aside which one is the tankcheck or not). Wouldnt it be a good idea to have a second set of gear ready to switch around? I know that one maintank in my guild is using different sets of gear for Scum and TfB. In theory you could use this per boss instead, if you are willing to constantly switch gear ofc.
Afterall, if someone is tanking TfB NiM I would assume getting 69/72 tokens for specoptimalisation isnt to big a deal.
@Fire-Breath: I can just speak for myself, but i did "reverse engineer" the migated dmg from the combatlogs with the current tanking migation i have for a specific attacktype (m/r , i/e ...) to get the unmigated dmg the boss does, like you suggested.
For the second topic: That's the same thing i alrdy suggested a page before for tfb NiM
(Beware that my results may have a huge variance as i dont have time to analyze hundreds of combatlogs, nor to analyze every possible sequence of attacks a boss can have)
TWH: 0,662 DG: 0,516 Op9: 0,790 Kephess: 0,293 TFB: 0,824
These numbers (the ratio between m/r and f/t attacks for e/k dmg for a bossfight) suggest that you more or less use kbn/dipsticks numbers for the Op9 and the TFB NiM fight (~520 defense rating for 2400 gear budget i think), to use no defense rating (unless you can't avoid it) for the DG and Kephess fight and to use a little bit of defense rating for the twh fight (about ~400 for 2400 stat Budget, can't be more excact there due to variance).
You can also use an average number out of these 5 bosses for average gear budget you should have (like kbn did for sv hm). With my numbers that would be 0,624, which means almost no defense rating at all for sin tanks in a tfb NiM operation on average.

Kitru's Avatar


Kitru
07.09.2013 , 12:37 PM | #33
Quote: Originally Posted by THoK-Zeus View Post
That's again wrong. You have the the same times on DG Fight aswell where the tank takes no damage (you forgot about force leech time)
Actually, I didn't. The amount of downtime for Kel'sara's Force Leech *pales* in comparison to the downtime on Writhing Horror. First off, the Force Leech doesn't happen until the 120 seconds of the fight, so the equal downtime only occurs for ~13% of the actual fight (last 120 seconds with even uptime/downtime). For Writhing Horror, you're only tanking the boss 50% of the time and only 50% of the time that you're not tanking the boss you're tanking an add that deals half as much damage.

Seriously, just look at the incoming damage charts for the two fights. Writhing Horror *explicitly* shows *huge* amounts of explicit tank downtime (hell, for the first 45 seconds, the second tank should only take raid damage). If you honestly think that DG shows anywhere *near* the same level of uptime/downtime, you're deluding yourself. If you can't see *that* (or even understand the fundamental logic behind "87% v. 62.5% uptime"), what the hell are you even *trying* to do discussing this?

Quote:
Force Shrouding Force Leech dmg is a very bad idea (again misinformation) as thats just about 10k dmg migated.
I'm not talking about using Resilience on Force Leech. I'm talking about making sure your healers don't slap you with heals or HoTs and taking huge damage to the face. All of the damage from Force Leech is entirely avoidable. If you're taking damage from it, it's because either you did something wrong (not removing HoTs if they were applied, standing in a heal patch) or because your healers did something wrong (healing you and/or overusing HoTs).

Quote:
It's way better to Force Shroud a Force Scream and a Discharge (thats about 25k dmg migated). That happens about 3 and 5 sec after you taunt kelsara after he force leeches the other tank.
Quote:
KBN has 1550 Damage taken on twh so thats again over a 100 Points lower then your 1690 dps you are assuming (again misinformation).
Inconsistent comparison. I was using Will's parse. Bringing in KBN's TWH parse with a Shadow tank to compare against Will's parse with a Powertech tank isn't comparing like parses. The reason why I used the 1690 v. 1750 numbers were because those were with the same person, in the same gear, with the same raid group. As many variables as could be controlled were the same.

Now, if you want to find 2 of KBN's parses that happened on the same day for TWH and DG, you can start comparing those numbers *to each other* but comparing different fights with different ACs for different people is just massively disingenuous (and outright misinformation).

Quote:
Another wrong part is where you are assuming that the spikes tanks get from twh fight are higher then the ones tanks get on third Phase kel'sara.
I'm not assuming. I'm mathing. An unmitigated Voltaic Slash or Force Scream will hit for 17k each. Discharge only hits for ~8k. Based upon Will's parse, the closest the two of those get is hitting at the extreme ends of a 3 second window with very little else on either side.

In that same 3 second window, the Writhing Horror can manage 2 melee attacks and 3 nasty bites followed up by another 2 melee attacks. The melee attacks hit for 6250 unmitigated and the nasty bites hit for 4300. If all 7 of those get through, the unlucky Shadow just took 37.9k damage to the face. The follow up for the Writhing Horror is a lot worse too since, as previously mentioned, TWH's basic attacks are *double attacks* that each hit for 6250 unmitigated a piece whereas Kel'sara's is a piddling 3k single hit.

On top of *that*, the number of potential spike incidents on TWH is *way* higher than the number of potential spike incidents on DG. TWH incidents occur roughly once every 10 seconds while you're tanking the boss. The big Kel'sara spikes happen only once every 50-60 seconds (thanks to tank swapping, it's once every 25-30 seconds split between the two tanks, or once per Force Leech swap cycle).

In a single 3 second window, TWH is going to have a *way* bigger spike potential, which is what we're discussing, those potential spike incidents are *way* more common (which more than offsets the fact that punching through said mitigation is more uncommon), and the follow up damage is *also* substantially worse and therefore more likely to kill you if you didn't get dropped before. Hell, the Kel'sara spike isn't liable to kill you off unless your healers have been slacking since everything else Kel'sara does, barring Force Leech, hits *really* light.

Quote:
Third the m/r to f/t ratio from this spikes is about 50/50 so you dont want any defense rating for these spikes aswell
If you actually look at the 3 second spike window rather than the immediate incident and the immediate follow up, the damage split is roughly 2:1.

Quote:
( well i did not answer to your Statement with the low DtPS maintenance healing on dg NiM fight )
Because, you know, it *is* low.
Walls of Text? I *love* Walls of Text!
My New Class Idea
Shadow Class Rep - Suggest/Review Questions Here
Quote: Originally Posted by Fende View Post
Listen to Kitru. Kitru knows all.

THoK-Zeus's Avatar


THoK-Zeus
07.09.2013 , 05:31 PM | #34
Quote: Originally Posted by Kitru View Post
Actually, I didn't. The amount of downtime for Kel'sara's Force Leech *pales* in comparison to the downtime on Writhing Horror. First off, the Force Leech doesn't happen until the 120 seconds of the fight, so the equal downtime only occurs for ~13% of the actual fight (last 120 seconds with even uptime/downtime). For Writhing Horror, you're only tanking the boss 50% of the time and only 50% of the time that you're not tanking the boss you're tanking an add that deals half as much damage.

Seriously, just look at the incoming damage charts for the two fights. Writhing Horror *explicitly* shows *huge* amounts of explicit tank downtime (hell, for the first 45 seconds, the second tank should only take raid damage). If you honestly think that DG shows anywhere *near* the same level of uptime/downtime, you're deluding yourself. If you can't see *that* (or even understand the fundamental logic behind "87% v. 62.5% uptime"), what the hell are you even *trying* to do discussing this?



I'm not talking about using Resilience on Force Leech. I'm talking about making sure your healers don't slap you with heals or HoTs and taking huge damage to the face. All of the damage from Force Leech is entirely avoidable. If you're taking damage from it, it's because either you did something wrong (not removing HoTs if they were applied, standing in a heal patch) or because your healers did something wrong (healing you and/or overusing HoTs).





Inconsistent comparison. I was using Will's parse. Bringing in KBN's TWH parse with a Shadow tank to compare against Will's parse with a Powertech tank isn't comparing like parses. The reason why I used the 1690 v. 1750 numbers were because those were with the same person, in the same gear, with the same raid group. As many variables as could be controlled were the same.

Now, if you want to find 2 of KBN's parses that happened on the same day for TWH and DG, you can start comparing those numbers *to each other* but comparing different fights with different ACs for different people is just massively disingenuous (and outright misinformation).

We can also compared these logs and see that tank takes more then 100 dmg on the same fight (with same gear i suppose)


I'm not assuming. I'm mathing. An unmitigated Voltaic Slash or Force Scream will hit for 17k each. Discharge only hits for ~8k. Based upon Will's parse, the closest the two of those get is hitting at the extreme ends of a 3 second window with very little else on either side.

In that same 3 second window, the Writhing Horror can manage 2 melee attacks and 3 nasty bites followed up by another 2 melee attacks. The melee attacks hit for 6250 unmitigated and the nasty bites hit for 4300. If all 7 of those get through, the unlucky Shadow just took 37.9k damage to the face. The follow up for the Writhing Horror is a lot worse too since, as previously mentioned, TWH's basic attacks are *double attacks* that each hit for 6250 unmitigated a piece whereas Kel'sara's is a piddling 3k single hit.

On top of *that*, the number of potential spike incidents on TWH is *way* higher than the number of potential spike incidents on DG. TWH incidents occur roughly once every 10 seconds while you're tanking the boss. The big Kel'sara spikes happen only once every 50-60 seconds (thanks to tank swapping, it's once every 25-30 seconds split between the two tanks, or once per Force Leech swap cycle).

In a single 3 second window, TWH is going to have a *way* bigger spike potential, which is what we're discussing, those potential spike incidents are *way* more common (which more than offsets the fact that punching through said mitigation is more uncommon), and the follow up damage is *also* substantially worse and therefore more likely to kill you if you didn't get dropped before. Hell, the Kel'sara spike isn't liable to kill you off unless your healers have been slacking since everything else Kel'sara does, barring Force Leech, hits *really* light.



If you actually look at the 3 second spike window rather than the immediate incident and the immediate follow up, the damage split is roughly 2:1.



Because, you know, it *is* low.
http://www.torparse.com/l/308310 Log for 1,6 s between Force Scream and Voltaic Slash = 34k in 1,6 s
Your biggest dmg spike is 3,4s 37,9 k (i reevulate the average time for this spike and did not pick a single spike they tend to get longer during fight) This spike happens. Then you tell me that this happens every 10 s which would indicate 30 spikes for each tank with 90 nasty bite hits (on a 6 min fight http://www.torparse.com/l/308310 ). 2 Things: I just see 63 nasty bites and i see this spike just half the time (the other half of the time we see another attacksequence) ergo this spike just happens about every 30 s.
2/1 Damage Split still means almost no defense rating at all for sin tanks (see my posted numbers...

I did never compare the uptimes.

Aronite's Avatar


Aronite
07.10.2013 , 10:07 PM | #35
I would like to preface this post by noting that I am KBN's cotank (Emital - Shadow Tank - Ebon Hawk). He's a better mind for analytics so I generally leave the forum to him.

Quote: Originally Posted by THoK-Zeus View Post
I am rly shocked at the amount of misinformation, lies, twisted truth you and kitru are giving to the community in this thread.
Thats a fairly pointed statement when all you have are your own numbers that in the context of this thread, or anything you can link to, have no analytic backing.

Quote: Originally Posted by THoK-Zeus View Post
Operator 9 is nowwhere near a tank check like dread guards. You get (out of combat logs) not more then 1,1k damage overall and not more then 1,5 k damage during some points of the fight. Operator is a way bigger raid healing issue and not sth. tanks should itemize themselves for. On the Dread Guard fight tanks get like 700 dmg more on average during the fight
The only thing you could do to optimize your gear for Op9 is, as Op9 is just about spike damage (according to Kitru), stacking hp (Underworld instead of veracity, B mods....) and that's contradicting to the things you tell ppl in your spreadsheets ("don't stack endurance").
Luckily Kitru and KBN are different people, so discontinuity between statements they make is invalid and non-germane.

Quote: Originally Posted by THoK-Zeus View Post
The real check for tfb NiM is the Dread Guard fight (every guild that cleared dg did also clear op9 so this fight is way easier) and either you optimize for this fight or for the whole operation-damage on average (i posted the numbers above they're completely different to your "data").
Your combatlogs prove you wrong,too. I analyzed the last 41 combatlogs you posted for dg NiM fight (all wipes). You did met the enrage timer in just 4 of these fights. When this fight is no healing/tanking check like Kitru and you want to believe us, how the hell can you then wipe 37 times just to mechanics?
This above all demonstrates your own total lack of meaningful experience with this fight (NiM DG), or your abject failure to analyze your own group's performance.

My assertions are numbered for your convenience:
1) The healing check for the first phase of the fight is recovering from lightening field.
2) Tanks in general should take considerably less damage from the preliminary ticks and the final burst than any other member of the raid.
3) As a function of their health and the damage any given member of the Operation is likely to take in the first phase, all lightning field damage on non tank members presents a critical imperative for the healers. Heirad is entirely capable of vaporizing a dps/healer after they've taken a full hit from the final tick of the lightening field.
4) The enrage timer for NiM DG is simply impossible for all but the very top shelf DPS. For those DPS the enrage requires brilliant execution and enviable output.

With those assertions in place it becomes clear that the loss of even 1 dps (under the best case scenario when a battle res is up) causes 5-10 seconds of lost damage output, healing output, rebuff from someone (A GCD best used elsewhere) and a repeated ramp up period.

Meaning, if one of our gunslingers misses their roll they subsequently die to Heirad or require the full attention of a healer to prevent dying to Heirad making tank death much more likely.

Show me a group for which this is not a catastrophic, cascade failure and equivalent to a long, slow, pointless wipe to the enrage timer and I'll show you one of the dozen or so groups that has downed this fight.

In short, your assumptions about what you see in any of those logs and whatever you believe they prove, that cannot be numerically derived, is patently false. If a DPS died to a lightning field in the first phase, we saved ourselves the trouble and wiped it up.

Quote: Originally Posted by THoK-Zeus View Post
I also find it not funny when you are again want people to believe that the m/r attacks (your beloved voltaic slash) are doing most of the steady damage tanks receive in dg NiM. We both know that is plain wrong (its abit over 50/50 but nowhere near the 79/20 your spreadsheet indicates) but you seem to find it good to spread misinformation and twist the truth.
Again, you acting like KBN or Kitru have some ulterior motive in providing this information doesn't help your case.

Frankly, any geared tank (or the healer keeping him up) is going to find ambient damage from sources that preclude defense to be irritating but largely irrelevant in light of the 18k Voltaic slashes that defense will give you a better chance of avoiding completely.

Your insistence that a single variable in the model is so magnificently incorrect as to negate not only the model but the credibility of the people that built it completely overshadows the larger pragmatic implications of what you're suggesting.

If you are going to argue for the complete dismissal of defense from your gear you would have to provide that the resulting gains in shield/absorb/luck/blessings of RNG string will then make it likely that you will never be below the threshold of a voltaic slash (or similar defense-able attack) killing you.

Quote: Originally Posted by THoK-Zeus View Post
I think you just want to distract from the fact that your spreadsheet is completely wrong about everything for tfb NiM. You just want to take the only Boss in this Operation, which has miraculously the same percentage of m/r attacks like your spreadsheet, for the logs and then call him "The fight you should optimize for" in this Operation.
At this point you are extrapolating embarrassment over a perceived inaccuracy into a systemic defense of a system for which you yourself can provide no alternative.

You can spit numbers all day long, as can I. However when it comes to taking issue with the analytic method used to verify this data all you seem to demonstrate is being wide of the mark in interpreting the model.

TL;DR I believe you mean "I disagree with KBN/Kitru/Dipstik on this issue" rather than "KBN/Kitru/Dipstik are lying to everyone because they're embarrassed that I don't agree!"

Grumpftard's Avatar


Grumpftard
07.10.2013 , 10:33 PM | #36
Quote: Originally Posted by Aronite View Post
I would like to preface this post by noting that I am KBN's cotank (Emital - Shadow Tank - Ebon Hawk). He's a better mind for analytics so I generally leave the forum to him.
For someone that silently reads and rarely ever comments on the forums, this was a very lengthy and detailed post.
I find this very interesting.
If the OP's comments about the "Evil Theory-crafting Community" were enough to get that kind of response from you....that right there should be a pretty good indicator that the OP is in left field with a hockey mask and swinging a football bat.
TORVA NEX! (The Site) (The Video)
"You don't have to say more than that!!"
Grumpf =Tankasin / G'rumpf =Mara / Gru'mpf =Jugg / Grum'pf =P-Tech
Grump'f =Shadow / G'rump'f =Scoundrel / Gr'umpf =Operative / Jei'Dahl =Merc

THoK-Zeus's Avatar


THoK-Zeus
07.11.2013 , 12:35 AM | #37
@aronite: KBN is not providing ANY Information. How did he get the data for his spreadsheet?
Instead he's posting 2 numbers in his thread (0.790044607 and 0.209955393) with wrong variance ( 9 Digits , means you analyzed combat dmg so well that these numbers are correct for a 0,01 Damage Difference (impossible when the mimimum damage a boss does is 1)(valid digits....) ) and with these 2 numbers he's achieving all his "results".
These 2 numbers are the only important thing of any tanking spreadsheet (as they provide the difference between d and s/a) and thers no "analytic method" kbn posted how he did get the results for his numbers at all, so dont lie... (How did KBN achieve his "Magic scalars" and how did he make them so accurate?)

I don't know which analytic method he used to get his data of the m/r f/t Ratio, so when you blame me for that its probably more a diss to kbn. (I provided how i did get my data, i analyzed combat logs for m/r f/t Ratio. You have your 41 combat logs of dg fight perhaps start to analyze them for the ratio, they provide the same results i got)


You can just recover from the lightning field if the tanks are not dieing during raid heal. When you find it not important that tanks are migating the most dmg that is mathematically possible OK. I have another opinion on that topic.

Quote: Originally Posted by Aronite View Post
Again, you acting like KBN or Kitru have some ulterior motive in providing this information doesn't help your case.

Frankly, any geared tank (or the healer keeping him up) is going to find ambient damage from sources that preclude defense to be irritating but largely irrelevant in light of the 18k Voltaic slashes that defense will give you a better chance of avoiding completely.

Your insistence that a single variable in the model is so magnificently incorrect as to negate not only the model but the credibility of the people that built it completely overshadows the larger pragmatic implications of what you're suggesting.
This single variable is the most important one and it makes the whole model completly incorrect (and because of valid digits and a missing variance these 2 numbers are definitly incorrect), if the difference is just 10% (just try 70/30, 60/40 or 50/50with kbns model if you dont believe me...)

dipstik's Avatar


dipstik
07.11.2013 , 11:53 AM | #38
based on thor's analysis a 0.624/0.376 split between w/t gives:
total d s a
2400 0 1069 1331
2500 0 1122 1378
2600 0 1176 1424

for a shadow.

jugg:

2400 670 860 870
2500 723 860 917
2600 776 860 964
2700 815 873 1011

PT:

2400 114 1337 949
2500 118 1388 994
2600 124 1438 1039
2700 130 1487 1082

dipstik's Avatar


dipstik
07.12.2013 , 12:23 PM | #39
updated my 2.2 thead after doing analysis on old tfb data. i think i might have messed up by weighting the fractions by each bosses dps instead of the overall fraction... all my fractions added to one at the end but i am still going to change my numebrs once we get NiM data from both ops.