Fixing Shadow Tank Spikiness

[Jimvinny]

Maybe I'm way off base here, but what you appear to be missing, Thok, is that while those 32k spikes from Thrasher should not kill the perfect tank with two perfect healers, this is pretty much NEVER the case. So, since we can assume that most tanks and healers are not perfect, then we need to look at what that 32k spike will do to an optimized shadow with 40k hp at 80%. Assuming no cooldowns, well, It's almost certainly going to kill him. What's it going to do to a guardian at 80%? Take him to 8k. There's your problem.

 

If you want NiM content to remain gated to only the very highest percentile of shadow tanks and their healers, while a lower tier of guardians and VG's get to waltz through, fine. But don't pretend this is balanced.

[Kitru]

(thers no instadead spike in 8 man content)

 

There aren't any "insta-dead from max hp" spikes anywhere in the *entire game*: even 16m NiM Thrasher doesn't have a 60k pre-mitigation attack. The spikes that kill you are those spikes that gib you from a reasonable non-max level of hp. This is why those Shadows that run with *absolutely amazing* healers that keep them constantly topped off have no problems with spikiness: they keep them constantly topped off such that they never dip below this threshold of being instantly killed (you would know this if you actually *read the discussion* rather than joining it late and acting as if we haven't already gone over this multiple times).

 

Shadows are *already* prone to be at lower hp levels than the other tanks (high spikiness means higher variability in incoming damage which causes the bell curve of expected hp to be smoother rather than grouped primarily at the higher max hp percents). This is basic common sense. It's exacerbated by the fact that we also have a *higher* instant-death threshold. Since we can't explicitly *expect* when we're going to get unlucky (which is the only time the spikes approach "insta-gib" levels), there's no way to reliably increase our hp *above* the instant-death threshold we're *already* more liable to be under. It doesn't help that, for Shadows, the instant-death threshold is really, really high. In fact, the instant-death threshold is very often *above* what is nominally considered to be "comfortable" hp: an unlucky Thrasher chain is going to deal ~32.5k damage in a single GCD; at 38k max hp, the instant death threshold would be *85%* of max hp. For a Guardian or VG, the instant death threshold is 65%.

 

So, if you're a Shadow running with healers that *never* allow you to drop below 85%, you'll be completely and totally fine, but, if you're running with healers that don't *constantly* keep you *absolutely* topped off, wasting *massive* amounts of healing on overheal and devoting an inordinate amount of attention to watching your hp (so they can catch and react to those situations *immediately*) you're *going* to get dropped below the instant-death threshold which *will* cause you to instantly die.

 

It's not, nor has it *ever* been a question of "one shot kills you from full health". It has always been "one shot kills you from what *should* be a *reasonable* amount of health to not get one-shot from". If you actually read the thread, you'd know this.

 

Honestly, it's getting kind of old to tell you to actually read the discussion that you've joined several months late, especially since you're just *continually* rehashing arguments that have been debunked time and time again. I realize that it's annoying to read through all of this, but you have *no* clue how annoying it is to have to *constantly* deal with idiots spouting the same tired arguments concerning a discussion that they are almost *completely* ignorant about, not to mention applying terminology in the wrong way while misinterpreting things that have been spelled out several times before.

 

Seriously, if you're going to join the discussion *late*, have the decency to actually get acquainted with what was discussed before you finally deigned to get involved. Welcome to my ignore list. Other people can deal with you until such time as you actually know what the hell you're talking about.

Edited July 23, 2013 by Kitru

[THoK-Zeus]

.....

 

Shadows are *already* prone to be at lower hp levels than the other tanks (high spikiness means higher variability in incoming damage which causes the bell curve of expected hp to be smoother rather than grouped primarily at the higher max hp percents). This is basic common sense. It's exacerbated by the fact that we also have a *higher* instant-death threshold. Since we can't explicitly *expect* when we're going to get unlucky (which is the only time the spikes approach "insta-gib" levels), there's no way to reliably increase our hp *above* the instant-death threshold we're *already* more liable to be under. It doesn't help that, for Shadows, the instant-death threshold is really, really high. In fact, the instant-death threshold is very often *above* what is nominally considered to be "comfortable" hp: an unlucky Thrasher chain is going to deal ~32.5k damage in a single GCD; at 38k max hp, the instant death threshold would be *85%* of max hp. For a Guardian or VG, the instant death threshold is 65%.

 

So, if you're a Shadow running with healers that *never* allow you to drop below 85%, you'll be completely and totally fine, but, if you're running with healers that don't *constantly* keep you *absolutely* topped off, wasting *massive* amounts of healing on overheal and devoting an inordinate amount of attention to watching your hp (so they can catch and react to those situations *immediately*) you're *going* to get dropped below the instant-death threshold which *will* cause you to instantly die.

 

It's not, nor has it *ever* been a question of "one shot kills you from full health". It has always been "one shot kills you from what *should* be a *reasonable* amount of health to not get one-shot from". If you actually read the thread, you'd know this.

.....

 

Prove your 32,5 k 1,5s spike on thrasher please with a combat log. So far out of 60 thrasher nim combat logs i did not saw a single spike like that. Biggest single hit is a 16k Swipe. I did (contrary to you) analyze these things.

But hey put me on your ignore list, if you don't wanna see the truth (e.g. everything you said is wrong).

Thrasher nim is old news,yes .

 

@Jimvinny: This spike appears over the Duration of 4 seconds. If you are not getting a single point of Healing during these 4 seconds, yes the spike (this spike is just a theoretical spike, his possibility that it happens once during a full fight mt thrasher is less then 50% as stated above) will kill you.

A guardian won't have an additional 5% healing from healers nor would he have self healing (see my calculations above). The threshold for one single global cooldown for thrasher nim is less then 50% even for assassins.

Edited July 23, 2013 by THoK-Zeus

[DarthSpekulatius]

Shield/defense Rating reduces the probabilty of getting unmigated attacks therefore they are working against spike damage.

 

no they help to keep you at high health by reducing the likelihood of spikes themselves and lowering the chance to be at low health that could kill you instantly. (so they do work but only indirectly)

 

if for example you are steady between 65% and 95% HP then increasing the constant selfheal or the non static mitigation (any mitigation that has an binary component that randomly says yes/no)

would up that lower health value without affecting the higher one (at high health your Healers regenerate energy or DPS or do nothing)

and without affecting the total health % where you may be killed without realistic possibility to react

you may still die at the same health percentage as without any defense or shield or heals (internal/external doesn't matter).

the problem is they do nothing to reduce the size of the actual spikes.

 

static mitigation on the other hand affects both the minimum health percentage you may be at during a normal maintenance cycle and lower the health percentage that risks instant or almost instant death.

 

what this thread is all about, is that many think the health % at witch shadows can die without reaction time is to high.

And to find a way to fix that without braking something else. (creating an OP monster falls into breaking something)

Edited July 23, 2013 by DarthSpekulatius

[THoK-Zeus]

no they help to keep you at high health by reducing the likelihood of spikes themselves and lowering the chance to be at low health that could kill you instantly. (so they do work but only indirectly)

 

if for example you are steady between 65% and 95% HP then increasing the constant selfheal or the non static mitigation (any mitigation that has an binary component that randomly says yes/no)

would up that lower health value without affecting the higher one (at high health your Healers regenerate energy or DPS or do nothing)

and without affecting the total health % where you may be killed without realistic possibility to react

you may still die at the same health percentage as without any defense or shield or heals (internal/external doesn't matter).

the problem is they do nothing to reduce the size of the actual spikes.

 

static mitigation on the other hand affects both the minimum health percentage you may be at during a normal maintenance cycle and lower the health percentage that risks instant or almost instant death.

 

what this thread is all about, is that many think the health % at witch shadows can die without reaction time is to high.

And to find a way to fix that without braking something else. (creating an OP monster falls into breaking something)

 

First part is excactly what i already said.

 

For the problem that self healing is not reducing the size of the actual spikes:

This problem would be really big if you are killed with 1 single ability (apart from terminate in nim on 8 man content that is not happening), so the additional healing you have cannot react to the spike damage a shadow tank gets.

 

This problems gets less and less important, the more seconds of spike damage the boss needs to actually kill the tank (we are assuming a high health pool when sin tanks are not getting spike damage). Therefore i did post the strongest possible spike in thrasher 8 man nim fight (which people think is one of the biggest challenges) a tank can possible get in this fight. (The average spikes are not that big normally)

 

So, if we can safely assume that if there's no instant killing ability from 80% to 0%, self healing has its use (as does healing) in recovering from damage that is way above the average.

I did post another calculation for a 20 k hit, where the difference in additional healing makes up for the difference in armor, so that a shadow tank is as fast recovering from a spike as a Guardian.

 

The mean migation of a sin tank is actually higher then the ones from a Guardian/Vanguard, so they should always be at a higher health level as long as there are no spikes.

Edited July 23, 2013 by THoK-Zeus

[Berjiz]

Give me a larger spike from thrasher, show me a 40k spike within 2 seconds, but no you prefer your blah blah with all the stuff everybody knows already.

Can't find it now but there was a log about a week ago from nim writhing horror that did 40k dmg in about 1.2-1.4 sec.

[THoK-Zeus]

Can't find it now but there was a log about a week ago from nim writhing horror that did 40k dmg in about 1.2-1.4 sec.

 

You probably mean the 3 second 37k spike in the minimal defense rating thread in the tanking Forum (highest theoretically possible spike for twh afaik). This spike has 7 unmigated attacks in a row (thx to the way better damage profile twh has) (all attacks from twh are split into 2-3 pieces), which gives this spike a 0,11% probabilty to happen (that's really unlikely in my opinion out of 1000 attacks sequences this will happen ~1 single time). Also most of the times both of the healers can focus the main tank in this encounter and its a way easier fight overall.

Edited July 23, 2013 by THoK-Zeus

[MillionsKNives]

I just want to throw out there that tonight I took 37k damage in the first 4 seconds of the frog mini-boss in HM TFB. I have 36k health in a mix of 69 and 72 mods. I died almost immediately after turning him around and barely after my healers even got their revivication pools on the ground.

 

Had I survived that I would have taken about 3k extra, because the ability is split into 3 parts and I took only 2/3 for the third one.

Edited July 24, 2013 by MillionsKNives

[THoK-Zeus]

I just want to throw out there that tonight I took 37k damage in the first 4 seconds of the frog mini-boss in HM TFB. I have 36k health in a mix of 69 and 72 mods. I died almost immediately after turning him around and barely after my healers even got their revivication pools on the ground.

 

Had I survived that I would have taken about 3k extra, because the ability is split into 3 parts and I took only 2/3 for the third one.

 

I looked at my last nim parse of the frog mini-Boss (don't know whether he got stronger in nim i doubt it). Here the boss can hit me for a maximum of 12,3 k per 3 hits every 1,5 - 2 seconds, so yeah your parse must have been really unlucky (9 unmigated m/r attacks in a row have a 0,004% of happening).

Such a spike will just happen 1 time out of 25000 times, so it's nothing one should really care for .

[swtordrz]

This is an awesome thread, thanks to all the contributors, just two quick questions:

 

1. I was of the view that different classes intentionally had unique strengths and weaknesses to optimally cater to different game content so would it not make more sense to bring the "right" type of tank for Nightmare PvE content rather than remove a weakness of Shadow/Assassin damage spikiness?

 

Follow up question:

 

2. If Shadow/Assassin spikiness as a weakness is appropriately addressed, given its already superior class/mirror versatility for PvE and PvP content, why would anyone be interested in playing or welcoming any other type of tank?

 

Look forward to your views, seeking some concise clarity, thank you in advance.

[THoK-Zeus]

This is an awesome thread, thanks to all the contributors, just two quick questions:

 

1. I was of the view that different classes intentionally had unique strengths and weaknesses to optimally cater to different game content so would it not make more sense to bring the "right" type of tank for Nightmare PvE content rather than remove a weakness of Shadow/Assassin damage spikiness?

 

Follow up question:

 

2. If Shadow/Assassin spikiness as a weakness is appropriately addressed, given its already superior class/mirror versatility for PvE and PvP content, why would anyone be interested in playing or welcoming any other type of tank?

 

Look forward to your views, seeking some concise clarity, thank you in advance.

 

1. Sin Tanks have some unique strenghts for nim pve content (atleast for 8 man) (Phase Walk for example...) Look here for a more detailed view upon the different Advantages/disadvantages all the tank classes currently have:

http://www.swtor.com/community/showthread.php?t=660876&page=2

 

2. As for 8 man spike damage is not that big of a problem probably not, if they fix spike damage completely sin tanks might be definitly ahead in 16 man Content (600 hps just from phase walk sounds amazing)

[Ancaglon]

Such a spike will just happen 1 time out of 25000 times, so it's nothing one should really care for .

Even if your numbers are correct, such a damage spike still would not have killed an equivalent-geared Vanguard or Guardian. Would you care to calculate what their odds of getting hit by such a killing spike, and compare?

Edited July 24, 2013 by Ancaglon

[THoK-Zeus]

Even if your numbers are correct, such a damage spike still would not have killed an equivalent-geared Vanguard or Guardian. Would you care to calculate what their odds of getting hit by such a killing spike, and compare?

 

The odds that pt/juggs get hit by the excact same amout is zero (as the spike we are talking about is again the highest possible one). They would instead get hit by about 29 k from the same spike.

But honestly when one probabilty is 0,00004 and the other probability is 0,00000, both is more or less impossible to happen (Exceptions prove the rule ).

We are talking about a theoretical thing that will happen about every 25000 time you attack the frog miniboss. Is that really important?

Edited July 24, 2013 by THoK-Zeus

[Stormskimmer]

The odds that pt/juggs get hit by the excact same amout is zero (as the spike we are talking about is again the highest possible one). They would instead get hit by about 29 k from the same spike.

But honestly when one probabilty is 0,00004 and the other probability is 0,00000, both is more or less impossible to happen (Exceptions prove the rule ).

We are talking about a theoretical thing that will happen about every 25000 time you attack the frog miniboss. Is that really important?

 

 

Yeah.

First, he DIED from it, so it is pretty bloody important to MillionsKNives.

Second, it proved your "rule" was wrong. So, it should be important to you.

[Kitru]

1. I was of the view that different classes intentionally had unique strengths and weaknesses to optimally cater to different game content so would it not make more sense to bring the "right" type of tank for Nightmare PvE content rather than remove a weakness of Shadow/Assassin damage spikiness?

 

The suggestions being made aren't to *remove* the weakness Shadows have to spikiness. The suggestion is to bring said weakness down to a *reasonable* level. By shifting some of the reactive/RNG mitigation mitigation over to static mitigation, Shadows still remain substantially spikier than Guardians or VGs while not suffering from the oft mentioned "insta-dead".

 

2. If Shadow/Assassin spikiness as a weakness is appropriately addressed, given its already superior class/mirror versatility for PvE and PvP content, why would anyone be interested in playing or welcoming any other type of tank?

 

I have to wonder why you think that Shadows have "superior class/mirror versatility for PvE and PvP content"? At the moment, the tanks to bring are Guardians because of their ridonculously good CD suite (that only gets better the further you get into content), excellent utility, and smooth incoming damage profile. Even in PvP, Shadow tanks are only brought as node guards; Guardians are the actual *good* PvP tanks for actual tanking purposes.

 

Even if Shadow spikiness were addressed, unless something happened to Guardians (which I doubt will happen since that's all the devs seem to play), Shadows wouldn't be uber-tanks. It would simply make them *not* be a massive liability to ops groups. It's not a question of making Shadows better but making Shadows *viable*.

[DuEldrvarya]

I think alot of people forget that the reason most shadow tanks are knowledgeable about their effect sets and optimal rotations is because literally, (even in story mode unless you are way way way over-leveled) it is know or die. Assassin's are so flipping intense as tanks compared to any other class because if you let your concentration on your defensive buffs (and even on projects/throw since its self-heal is so important) you die. I mean there is no break, you just die...

 

In order to make it to lvl 55 66+ gear as a shadow tank, you will have had so many trying experiences tanking for teams because of your spikiness that you almost are required to become high functioning tanks.

 

I can **** around with guardians and still do ok. With an assassin, if you are not doing a good job, you are dead. I mean just dead.

 

Its pathetic that will almost full 69 gear and full 66 augments, I actually have to worry about spikiness on 50 hm's...

[Farstrider]

I finally specced back into Kinetic Combat this evening, to run a level 55 hardmode flashpoint, my group and I ended up in Cademimu, which I recently ran as my Defense tree Guardian alt. I hate to say that its almost unplayable comparatively. Edited July 25, 2013 by Farstrider

[nagatamen]

I'm sure this has been mentioned, but has it ever been fully fleshed out? Would a "Supercharge Shields" type cooldown help out with the spikiness? What I mean by Supercharge Shields is an ability with a cooldown (I don't know what length needed) that gives a 100% shield chance for a set number of hits and/or length of time. And when I say 100% shield chance I don't mean like Resilience's 100% chance, I mean bypassing that stupid 5% unblockable nonesense.

 

Assuming that it would help, what kind of specs would we need for it? Specifically I'm looking for how many hits, how long it should last and what's the length of the cooldown? If a force cost would help keep it balanced, what should that be? If it's too powerful, any way to tone it down?

 

And if all of it has been answered somewhere else can I get a link?

[EFSoupy]

I'm sure this has been mentioned, but has it ever been fully fleshed out? Would a "Supercharge Shields" type cooldown help out with the spikiness? What I mean by Supercharge Shields is an ability with a cooldown (I don't know what length needed) that gives a 100% shield chance for a set number of hits and/or length of time. And when I say 100% shield chance I don't mean like Resilience's 100% chance, I mean bypassing that stupid 5% unblockable nonesense.

 

Assuming that it would help, what kind of specs would we need for it? Specifically I'm looking for how many hits, how long it should last and what's the length of the cooldown? If a force cost would help keep it balanced, what should that be? If it's too powerful, any way to tone it down?

 

And if all of it has been answered somewhere else can I get a link?

 

I think part of the problem with making it a cd is we're still so reliant on a cd so one of two things happens. 1) cd timer is too long and we're still susceptible to spikes, just fewer of them or 2) cd it short enough to use it on every burst and it becomes too powerful. There have been a lot of talks of adding new mechanics to try to fix the damage spikes, but it seems that the best and simplest solution really is a small armor buff and a small healing nerf. If memory serves, the number crunchers like a 4-5% gain in DR and something like a 25% nerf to self heals.

[Kitru]

And if all of it has been answered somewhere else can I get a link?

 

It hasn't been answered because it's a fuzzy question to ask. To really take care of the spikiness problem, it would need to be on a very short CD, so that it's up for every major spike interval (since, without it, you'd end up with occasions where you still have no spike protection), which suggests something like a 20-30 sec CD. For it to be that short, the duration would need to be, at most, 3 seconds (long enough for a precision shielding mechanism), though it would create potential balance issues since it provides a *lot* of burst mitigation on a short CD. Combined with KW, it might get a bit bolloxed. Of course, if it *were* implemented as such, I would expect it to either replace KW (such that KW *becomes* said guaranteed shield CD) or be a rider attached to KW's activation (i.e. for ~2-3 seconds after KW is activated, 100% shield chance).

 

To really get into it, one would have to actually come up with various potential implementations (durations, CDs, whether or not there are charges,etc.) and derive from there whether it would be effective and balanced. From a sheer abstract, it's a bit too vague to get into specifics of balance.

[Kitru]

If memory serves, the number crunchers like a 4-5% gain in DR and something like a 25% nerf to self heals.

 

It's closer to a 40-50% reduction in self heals. The changes that I did the math for asked for the self heal on Combat Technique to be removed completely and the self-heal on Harnessed Shadows to be reduced from 2% to 1-1.5%. I recall, but am not entirely *sure*, that the 1.5% nerf brought mean mitigation to the same number as we currently get whereas 1% brought mean mitigation to slightly *worse* than we currently get, under the auspices of balancing out for a more stable incoming damage profile.

[THoK-Zeus]

Yeah.

First, he DIED from it, so it is pretty bloody important to MillionsKNives.

Second, it proved your "rule" was wrong. So, it should be important to you.

 

First awall it did not prove my rule wrong (40k in 1 gcd, 37k is not rly killing most of the sin tanks). If he dies 1 time out of 25000 times he engages the Boss, plz explain me why that is important. On average this would happen to him (if he goes every week to tfb) every 500 years....

Edited July 25, 2013 by THoK-Zeus

[Snarkasms]

First awall it did not prove my rule wrong (40k in 1 gcd, 37k is not rly killing most of the sin tanks). If he dies 1 time out of 25000 times he engages the Boss, plz explain me why that is important. On average this would happen to him (if he goes every week to tfb) every 500 years....

 

I went ahead and mocked up a full-BiS 72 shadow (since I left mine mid-69->72 transition) and you know what? If you're stacking mitigation to make yourself generally easier to heal (which it seems like we've mostly established you should do), you're topping out around 39/40k HP. That's with buffs, stims, stance, everything. Now let's examine what it takes to get said BiS tank to 37k HP:

 

One 3k hit.

 

A full BiS shadow/sin is probably normally operating within the 33k-38k range (82-95% health), so YES, it is entirely likely that a single unmitigated 37k hit will decimate a shadow/sin who's gearing for helping their healers out.

Edited July 25, 2013 by Snarkasms

[Leafy_Bug]

I went ahead and mocked up a full-BiS shadow (since I left mine mid-69->72 transition) and you know what? If you're stacking mitigation to make yourself generally easier to heal (which it seems like we've mostly established you should do), you're topping out around 39/40k HP. That's with buffs, stims, stance, everything. Now let's examine what it takes to get said BiS tank to 37k HP:

 

One 3k hit.

 

A full BiS shadow/sin is probably normally operating within the 33k-38k range (82-95% health), so YES, it is entirely likely that a single unmitigated 37k hit will decimate a shadow/sin who's gearing for helping their healers out.

 

 

I am BIS 2500 mitigation budget and my HP is 36800

[Snarkasms]

I am BIS 2500 mitigation budget and my HP is 36800

 

Exactly. My theoretical 72 BiS build has a 3342 mitigation budget. The thought that shadows need to be kept at 95% health or better to survive boss encounters is ludicrous, especially with all this mitigation.