Jump to content

Understanding Tank Theory Crafting


KeyboardNinja

Recommended Posts

KBN and Thok tell people to learn to play because they are bad, nobody bats an eye. Someone comes and says KBN makes mistakes and needs to learn to play, everyone loses their minds :D.

 

Baseless? We've tested it, we've used it, several guilds on our server who compete with us use it. The videos are out there, if you care that much you can find them. I'm not your lackey, do the work yourself. This isn't English Comp 2, citations aren't required.:cool: I don't remember which guilds nor do I really care, we didn't come up with the idea to HP stack PT's and Sins on our own because everyone in the guild disagreed with me on my Sin gearing prior to NiM TFB.

 

KBN puts out math, and in the end it's just that, math. Reference his swtorboard.org Sin/Shadow guide, which had an improper initial spike threat rotation, as well as several other critical play mistakes. Someone else put up a hilarious comment in the responses, tactless as it was the stuff had merit and was kind of funny. I have a hard time taking gearing advice from a guy who is posting incorrect class guides.

 

 

Good to see some variety in opinions, if it works for you guys and other guilds, you should not be slapped on the wrist like many who disagree. Math has always stayed math and assumed perfect gameplay. When perfect gameplay was pointed out, from the player manning the shadow, other excuses were found of the sort : your healers are bad they need to learn to play.

 

 

For 16M NiM certain guilds, like NGE, do not gear up according to KBN's numbers and it works for them. They have Hateful Entity on farm on a weekly basis and they had quite a few achievements up there with the best 16M NiM guilds. I myself will customize my shadow for 16M NiM differently than what KBN is suggesting. Granted for 8 MAN I am using his math but 16 is another cup of tea. I wont be doing 16M NiM Raptus with 43k+ HP :cool:.

Edited by Leafy_Bug
Link to comment
Share on other sites

For 16M NiM certain guilds, like NGE, do not gear up according to KBN's numbers and it works for them. They have Hateful Entity on farm on a weekly basis and they had quite a few achievements up there with the best 16M NiM guilds. I myself will customize my shadow for 16M NiM differently than what KBN is suggesting. Granted for 8 MAN I am using his math but 16 is another cup of tea.

 

If I were doing 16 man, I would most definitely be using B mods. No question about it. I would also have a different stat distribution, since the ones that are generated use boss pre-mit damage numbers from 8 man, not from 16 man. The ratios are slightly different.

Link to comment
Share on other sites

KBN and Thok tell people to learn to play because they are bad, nobody bats an eye. Someone comes and says KBN makes mistakes and needs to learn to play, everyone loses their minds :D.

 

 

Good to see some variety in opinions, if it works for you guys and other guilds, you should not be slapped on the wrist like many who disagree. Math has always stayed math and assumed perfect gameplay. When perfect gameplay was pointed out, from the player manning the shadow, other excuses were found of the sort : your healers are bad they need to learn to play.

 

 

For 16M NiM certain guilds, like NGE, do not gear up according to KBN's numbers and it works for them. They have Hateful Entity on farm on a weekly basis and they had quite a few achievements up there with the best 16M NiM guilds. I myself will customize my shadow for 16M NiM differently than what KBN is suggesting. Granted for 8 MAN I am using his math but 16 is another cup of tea. I wont be doing 16M NiM Raptus with 43k+ HP :cool:.

 

Lol'd at the meme. Our main tanks don't change the methods they gear with for fights, the defensive focus (Def/Abs) gearing shifts but they all are at somewhere between 47-51k and they do just fine in both 8/16, we've killed every boss in the game post NiM TFB with HP stacked tanks and it's never a tank going down from want of healing that wipes us. Pretty much no one competitive on our server goes mitigation actually (not saying much, TBH :p). Math is just that, it's numbers. Without testing, they mean nothing. http://www.funny-memes.org/2014/02/its-theoretical-of-course-but-i-believe.html

 

To Keyboardninja, in a half-assed attempt to keep this civil, have you ever actually run TTK tests against incoming damage? Not in raid, but as a straight factor of "How long does this gearing keep me alive against a player/mob doing a preset static attack rotation"? I know my guys have, and heavy HP kept the tank alive longer in every iteration run for every class, especially PT. Probably also why every PvP tank on our server runs high HP. Dunno how they do it on other servers but our Ranked scene is all Juggs and PT's with boatloads of health. One other note, The Math doesn't factor in hybrid builds, which we've tested to be worth taking a look at for some fights.

Link to comment
Share on other sites

To Keyboardninja, in a half-assed attempt to keep this civil, have you ever actually run TTK tests against incoming damage? Not in raid, but as a straight factor of "How long does this gearing keep me alive against a player/mob doing a preset static attack rotation"? I know my guys have, and heavy HP kept the tank alive longer in every iteration run for every class, especially PT. Probably also why every PvP tank on our server runs high HP. Dunno how they do it on other servers but our Ranked scene is all Juggs and PT's with boatloads of health. One other note, The Math doesn't factor in hybrid builds, which we've tested to be worth taking a look at for some fights.

 

I'm all for civil…

 

I have run TTK tests. I've also done math on it. What you're talking about is generally known as "effective health", and it's simply the amount of damage required to drop you from 100% to 0% with no external healing and mean mitigation (I suspect you know this, but just clarifying for the record). It's fairly easy to compute as: HP / squish, where "squish" in this case must include the mitigation contribution from self-healing mechanics but not heals received bonuses. This hints at the fundamental problem with TTK as a tanking metric: it pretends you have no healer (ever) and optimizes for that case.

 

Brief reminder: if you test TTK in the open world, be sure to also get the pre-mit damage ratios for the boss you're testing on and remember that your results will be very specific to bosses with those damage ratios.

 

It is trivial to see from the expression for effective health and the way that the stat ratings interact that stacking endurance is by far the best way to increase your TTK. There isn't even a question about this. The problem is that you're optimizing for a fight scenario which does not exist! Dipstik's math from earlier is pretty clear on the implications. By endurance stacking, you can stretch your TTK from around 17 seconds to around 20. The problem is that a fight lasts longer than 20 seconds, which is why we need healers in the first place. The longer the fight lasts, the more that you want to optimize for external healing, rather than intrinsic durability. Dipstik calculated the crossover point as being 113 seconds. Even 8 man Nefra with four DPS pumping over 3.7k on the boss takes longer than that to die (and I've personally tested that particular situation). No other boss is even remotely close to that threshold.

 

Nevertheless, TTK is still one of the metrics that I peg as being something that is worth considering when theory crafting about tanks. Not in the OP, but in other posts I have made. Specifically: TTK, survivability and spikiness are what I generally look at. Because of the way that content, tanking mechanics and healing play off of each other in SWTOR, TTK is by far the least significant of these mechanisms. I mean, seriously, a 17 second TTK on a mitigation-stacked tank?! Unless your healers are AFK or dead, that's an unholy amount of time. Survivability and spikiness are major concerns. The former has to do with efficiency with respect to external healing, while the latter has to do with reaction time for external healing. Endurance is very, very important in reducing spikiness, and it is precisely why I would use B mods if I were tanking 16 man progression content.

 

Anyway, don't assume that I haven't tested this stuff. Just because I prefer to work in clean formal models doesn't mean that I haven't tested those models in the actual game. I have. Others have too, and there is a lot of really extensive research that weighs these various alternatives against each other.

Link to comment
Share on other sites

Hatred doesn't. Severity doesn't. We (DnT) don't. Provectus doesn't.

 

I actually can only think of a couple of tanks that HP stack, and their healers usually hate them.

I believe Thiol (of Hatred's A group/world first group) runs B mods. It would be nice to get a Hatred member to confirm this.

 

In the end, it's a role-playing game. If stacking endurance helps people to role-play as a big bad dude, then they're enjoying the game and no amount of math will correct them, because their enjoyment of the game and the development of their in-game persona derives from perception rather than truth.

Edited by MGNMTTRN
Link to comment
Share on other sites

I believe Thiol (of Hatred's A group/world first group) runs B mods. It would be nice to get a Hatred member to confirm this.

 

In the end, it's a role-playing game. If stacking endurance helps people to role-play as a big bad dude, then they're enjoying the game and no amount of math will correct them, because their enjoyment of the game and the development of their in-game persona derives from perception rather than truth.

 

He sits at just under 41k in some mix of 72/75s (just checked their NiM TfB Title run video). Anakyn sits at 39k. I'm sure someone more familiar with the HP given by 75s could tell if that's B mods or not. B Mods is not "going pure END" though. I've run B mods before as well.

Link to comment
Share on other sites

He sits at just under 41k in some mix of 72/75s (just checked their NiM TfB Title run video). Anakyn sits at 39k. I'm sure someone more familiar with the HP given by 75s could tell if that's B mods or not. B Mods is not "going pure END" though. I've run B mods before as well.

 

Depends on whether or not he is using the endurance talent. I'm fairly certain though that both of them would need to be in B mods to get the HP levels they have. Remember that in their TfB title run, they wouldn't have had very many 75s at all.

 

Also, Carl has mentioned on several instances that Hatred's tanks prefer the B mods paired with high mitigation enhancements and augments. As you said, this isn't really endurance stacking. I used to do this myself (even for 8 man content) pre-2.0.

Link to comment
Share on other sites

There's a lot of dogmatism on both sides here.

 

KBN has shown how to compute a very nice quantity: expected damage taken passively. That's great, and is a great start to tank theorycrafting. It has the benefit of being straightforward to compute, only depending on the damage profile.

 

This doesn't imply anything about the mitigation vs endurance debate. Expected TTK is another metric, which KBN describes how to compute again in the post above. This is another useful number to think about, but it doesn't tell us whether to take mitigation or endurance without further careful thinking or further tests.

 

To really say anything about mitigation vs endurance, you either need experience with both (which maybe some of you have), or you need some good simulations. Here you need to make more choices, such as a more detailed damage profile (which the results will depend upon), and you need to simulate healing somehow. If there's a next step in tank theorycrafting, this would be it. It requires judgement to set up properly, but this is where I'd go to really figure out something to say about mitigation vs endurance.

 

The simplest sort of simulation: Put a generous hot on the tank. Have some incoming damage profile (hits of X amount at Y intervals). Calculate expected time to die, or expected chance of death over 5 minutes, or something.

 

You'd do this for various sizes of hots, and various incoming damage profiles (some with bursts at greater intervals, etc).

 

Next steps: Add in some AI healing based off one or more of the classes instead of just a basic hot (so you get more heals when you get low, etc). Add in AI cooldown use.

 

You can go further, of course, and have the program learn how to heal and how to use cooldowns instead of hard coding it.

 

This would actually say something about mitigation vs endurance. Until then, we're just all bsing on the forums.

 

The highest-burst single hit in 8 man content today is a 29k hit: unmitigated (Driving Thrust). That's on a shadow/assassin, btw. It's closer to 24k on a vanguard/powertech. But let's take the 29k. 29000 HP is just shy of 68% of a fully buffed, fully stimmed mitigation stacked shadow/assassin tank. If your healers can't keep you above that line (even during progression) when the hit is coming at long and highly predictable intervals, then you have a much bigger problem than how your tanks are itemized. Mitigation stacking yields more than enough HP to sustain even the highest burst window in the game.

 

Will the damage be higher in nightmare mode? Absolutely. How much higher? Well, previous tiers of NiM content have very consistently buffed tank damage by a maximum of 15%. Some fights much lower than that. 15% buff on Driving Thrust gives an unmitigated hit of 33.4k, which is up to 78% of maximum health. Much scarier, but still well within safe margins if you have healers who are awake and a tank who can react appropriately when mitigation fails (which it will 40% of the time on that ability). Also remember that, by the time we get to Raptus (even at cutting-edge progression), mitigation stacking tanks will have almost 47k HP. That's back down to 71%.

 

This argument you've made multiple times. It doesn't sound too good. You're essentially assuming the healing is so easy that tanks are topped off at all times prior to big hits. If that's the case, then why does average dtps matter at all (low average dtps is the whole benefit of mitigation-heavy gearing)?

 

Remember, gearing only matters during progression. (For any raid group.)

Link to comment
Share on other sites

snip

 

I actually think the most useful tanking statistics, particularly during progression are "frequency of spikes", "depth of spikes", and something my healers and I call 'set and forget time' (which we think of as the amount of time I can go without a heal without risking death). The lower the frequency of spikes the better, the depth of spikes is an uncontrollable number since it's mostly just armor based, and "S&F" (which is in someways close to TTK) is some less than concrete measure of how long you go on average without a healer needing to devote multiple globals to you, which can be approximated in some way by looking at average length of time required to reach a damage total roughly equal to 3 unmitigated attacks (3 unmit, 6 shielded, or some combination thereof to total the amount of HP). KBN's model has, at least in my personal experience, made that core measure of "S&F" the longest. I've tried a lot of mixes, B mods, full END, full END with miti augs, etc, but in the end I always come back to slightly higher DEF than KBN suggests as what feels best for my healers and I. It's hard to argue using personal experience, but the best I can suggest is to try the two extremes, then try something in the middle (B Mods, half fort augs) and ask your healers what's easiest to keep you alive with, since that's really your whole goal. If they don't say max miti, I would be astonished.

Link to comment
Share on other sites

I actually think the most useful tanking statistics, particularly during progression are "frequency of spikes", "depth of spikes", and something my healers and I call 'set and forget time'

 

...

 

I've tried a lot of mixes, B mods, full END, full END with miti augs, etc, but in the end I always come back to slightly higher DEF than KBN suggests as what feels best for my healers and I.

 

These sound like some nice heuristics, and my whole point was that something like the experiece you claim, or failing that some detailed simulations, is what's needed. (Of course, as your personal experience it's hard to share a detailed account of and understand any bias, but at least it's some evidence for the correct question.)

 

An interesting point here is that a given raid group's particular choice of tanks & healers, as well as individual skill, is going to affect things. Your heuristics (which are similar to what I was suggesting inputting to a simulation) will vary depending on these factors, as well as from boss fight to boss fight. Other than class differences, the simplest of these is whether your group takes ticks from avoidable sources.

 

As a general comment (not to tenebras): Simulating (and accounts of heuristics and personal experience too) sounds inherently biased, but what you do is try a variety of inputs (based on particular [or even just typical types of] fight data hopefully), look at the results, and then try to interpret what you see. This of course isn't as straightforward as a calculation of something, but the idea is the result may be more meaningful if done with skill.

 

----------

 

Just a note: I don't mean to imply one way or another what my best guess is (and it isn't worth that much and isn't as experienced as many here anyway). I was more arguing that KBN's calculations are great at what they do, but what they do does not tell us which to choose between mitigation and endurance.

 

I certainly haven't tried full endurance, and have no experience with it outside of seeing those behemoths in random pug storymodes. I may take your advice: it would be an interesting thing to try on one of my tanks just to see (or maybe convince some of our tanks to try, as I don't main a tank).

Edited by cxten
Link to comment
Share on other sites

I actually think the most useful tanking statistics, particularly during progression are "frequency of spikes", "depth of spikes", and something my healers and I call 'set and forget time' (which we think of as the amount of time I can go without a heal without risking death). The lower the frequency of spikes the better, the depth of spikes is an uncontrollable number since it's mostly just armor based, and "S&F" (which is in someways close to TTK) is some less than concrete measure of how long you go on average without a healer needing to devote multiple globals to you, which can be approximated in some way by looking at average length of time required to reach a damage total roughly equal to 3 unmitigated attacks (3 unmit, 6 shielded, or some combination thereof to total the amount of HP). KBN's model has, at least in my personal experience, made that core measure of "S&F" the longest. I've tried a lot of mixes, B mods, full END, full END with miti augs, etc, but in the end I always come back to slightly higher DEF than KBN suggests as what feels best for my healers and I. It's hard to argue using personal experience, but the best I can suggest is to try the two extremes, then try something in the middle (B Mods, half fort augs) and ask your healers what's easiest to keep you alive with, since that's really your whole goal. If they don't say max miti, I would be astonished.

 

Serious question here: do you (or your healers) have a decent way of weighting depth of spikes vs frequency of spikes? I would actually argue that depth of spikes is something that we can control, because spikes are measured in the eyes of healers as percentage losses, not as absolute losses. Sure, it will take more healing to pop up back up after the fact if you have more health (relatively speaking), but that has nothing to do with spikiness.

 

Anyway, the question of "depth vs frequency" is very directly the fundamental conceptual issue that I, Thok and Dipstik all have different opinions on relative to quantifying spikiness. I think more quantitative input here from progression tanks would help a lot.

 

I was more arguing that KBN's calculations are great at what they do, but what they do does not tell us which to choose between mitigation and endurance.

 

Bingo. I should put that in my signature.

Edited by KeyboardNinja
Link to comment
Share on other sites

Serious question here: do you (or your healers) have a decent way of weighting depth of spikes vs frequency of spikes? I would actually argue that depth of spikes is something that we can control, because spikes are measured in the eyes of healers as percentage losses, not as absolute losses. Sure, it will take more healing to pop up back up after the fact if you have more health (relatively speaking), but that has nothing to do with spikiness.

 

Anyway, the question of "depth vs frequency" is very directly the fundamental conceptual issue that I, Thok and Dipstik all have different opinions on relative to quantifying spikiness. I think more quantitative input here from progression tanks would help a lot.

 

 

I was referring to depth of spikes as total damage taken, not a percentage, but I'm in active discussions about potential NiM damage with my healers now about whether to take slightly more frequent spikes to give them a cushion. When we come to a decision after testing I'll come back with more info.

 

The one thing I can say for sure (and this is coming from Kakuzzu, highest parsing Merc healer, and the best healer I've ever had heal me by far from personal experience), is that generally, refilling a tank's HP infrequently is better than throwing heals at him constantly. Of course, we run Merc/OP so YMMV on that theory. The less frequently our Merc has to dump 2 or 3 scans into me, the better.

 

I'm going to get a bit ballparky with numbers here so bear with me:

 

Our merc can one global me for roughly 27k HP, consistently. If I'm taking spikes bigger than that (or if 2 spikes exceed that), he has to spend more than 1 global on me and miss a raid healing global (and hurt his heat). If I need to be above my current HP + 27k for any particular incoming damage spike, he must focus me for more than one global, or both healers need to hit me with a heal. Our general healing practice when he's healing me is that he will refill me to full, shell me, and only come back when necessary. The longer I can make that time, the better, because he can use Missile on the raid, help DPS, aid the other healer, or do anything else that keeps his heat lower. What we are currently trying to figure out is whether I should increase my HP such that with just probes + shell I can survive an additional spike. I think the only way to know this for sure will be to get on PTS and test, since right now, the only scary spikes (NiM Thrasher 16 and Driving Thrust) are not exceeded by adding HP.

 

More spitballing/guessing follows:

 

A rough guess that is purely speculation is that Nef'ra will be hitting for about 39k pre-shield/defense. A poor roll twice in a row is death without big heals whether I have my current 43k, or 56k. The likelihood of those 2 rolls being bad is increased with a higher HP pool, but does not, as it seems to me, to be enough HP to survive. However, 2 mediocre rolls (say 26k + 13k) taken at 43k leaves me in extreme immediate need of a heal before the next swipe because anything but a full defend is death. A third mediocre roll of 13k with 56k actually leaves me alive. This is where my consideration for additional HP comes in. My goal is to give my healers as many rolls as possible before I MUST take a heal.

 

If my numbers are crazy, let me know, but that at least sums up my current thinking.

 

EDIT: Something that would GREATLY help my calculations would be the defend/shield/absorb percentages of a tank running HP over mitigation (B mods and max endurance would be optimal, but I'll take either if they are easily available).

 

EDIT 2: SWTORMiner pulled some #s for damage increase on Nef'ra out of the PTS patch, and the pre-shield/def damage is too high. I'll rerun this #s later on today if possible, or tomorrow if not. However if you want to redo the math switching 39k for ~32k before I get to it, that's the # we're working with based on datamining.

Edited by vVvCheese
Link to comment
Share on other sites

I'm all for civil…

It is trivial to see from the expression for effective health and the way that the stat ratings interact that stacking endurance is by far the best way to increase your TTK. There isn't even a question about this. The problem is that you're optimizing for a fight scenario which does not exist! Dipstik's math from earlier is pretty clear on the implications. By endurance stacking, you can stretch your TTK from around 17 seconds to around 20. The problem is that a fight lasts longer than 20 seconds, which is why we need healers in the first place. The longer the fight lasts, the more that you want to optimize for external healing, rather than intrinsic durability.

 

This is if, and only if, the healers are bottoming out on energy. If energy management is not an issue, then it becomes primarily about TTK, because it buys GCDs for the healers to put out raid heals. That extra 3 seconds translates to 2 globals that the heals can use to pump out green. Globals are a secondary resource for healers, not enough people recognize this. I've never seen our heals fail to keep up because they're straight OOE. They run out of the GCD's required to put the heals where they need to be when mistakes happen and the raid takes too much damage. Energy management hasn't been an issue for us since the first 2 times we 8man'd 16man HM Nefra, and most of that had to do with the cleanse globals required ended up costing the time needed to maintain tank HP. It was never energy, it was again, GCD's.

Link to comment
Share on other sites

This is if, and only if, the healers are bottoming out on energy. If energy management is not an issue, then it becomes primarily about TTK, because it buys GCDs for the healers to put out raid heals. That extra 3 seconds translates to 2 globals that the heals can use to pump out green. Globals are a secondary resource for healers, not enough people recognize this. I've never seen our heals fail to keep up because they're straight OOE. They run out of the GCD's required to put the heals where they need to be when mistakes happen and the raid takes too much damage. Energy management hasn't been an issue for us since the first 2 times we 8man'd 16man HM Nefra, and most of that had to do with the cleanse globals required ended up costing the time needed to maintain tank HP. It was never energy, it was again, GCD's.

 

Rewatch Ryz's NiM Cartel Warlords farm video (it wasn't your first kill). Energy was the gating factor on your healers for the entire fight, especially for your operative. Ultimately, there were two separate deaths from Stabbing Spree which could have been avoided if the healers had any energy at all when they hit that phase. I don't doubt that energy is a non-issue now, given that everyone is over-geared for the old NiMs and the new HMs are very easy.

 

The argument about having extra time is valid, but remember that you lose time by sacrificing mitigation to pick up TTK, since you're taking more damage. See dipstik's calculations on exactly this point. You're gaining time only in the short term; you lose that race very quickly (not quite at the 2 minute mark). Thus, TTK is only interesting for spike phases, which is precisely why I made the point about Driving Thrust.

Link to comment
Share on other sites

The only reason for me to mix some b-mods into my equip is to be able to survive 2 spikehits. Imagine a bossburst. First hits puts me with defstacked gear to 48%, healers would start casting a big heal or smth. lets assume a 2.5 second cast. Bosses although have a swing timer of 2 seconds ( some attack quicker ( Bestia) some slower (Nefra). Therefore during the healers cast i might get hit by another bos****. Imagina i would now again take such a spikehit that kills me ( both hits hit for 52%). If i atleast absorbed 1 of those hits i wouldnt die. 5% more hp would also made me survive. Stacking def might look a little like gambling, but in the end dtps is the lowest with that choice. While more hp might give you a bolster to survive spikephases ( or as i call them burstphases). But dtps will be higher and which is also a factor total healing needed to get back to 100% will also be higher. Therefore healer might invest too much ressources to get their tank back to 100% or instead they might let another raidmember die ( If healers see a dps at 70& and a tank at 80% they often tend to heal the tank not the dps).

 

There are however challenges for the individual raidroles. A Healingchallenge would be warlords. Tu'chuk tank gets nearly 3k dtps while that very tank needs also to be cleansed and the rest of the raid also gets dmg + there are phase of burstdmg on 1 dps. Overall you need about 7 hps. To make it easier for the healers you could swap to full def instead of tanking tu'chuk with 50k hp leading to 4k dtps. Tried both and the tradeoff is just to high 1k dtps against 9k more hp. But there are other bosses where some extra hp give the healers more time in order to heal the dps or themselves ( Tyrans for example).

 

For 16man some extra hp might reduce "spikiness" when looked at % but total (in numbers) the spikiness is the same or even higher. If you have 50k hp and gett hit for 30k you dont drop below 30%, if your at 40k hp defstacked 30k would put you below 30%. Hp might decrease the heartrate of the healers a bit more than defstacking however the total dmg (numbers) are the same and 10k more hp dont make you survive 1 bos**** more (at least not at bosses that count).

 

The last reason to stack hp is to avoid onehits. This however occurse very rarely (even in 16man). I personally never got 1 hitted although i am walking around fully defstacked. This might change with the new nm op but the first thing i want to do is to mitgate the dmg as most as i can to help the healers. Otherwise you could just use a dps with b-mods in heavy armor in front of the boss.

Edited by Methoxa
Link to comment
Share on other sites

Dear Abby,

I understand I may be poking a hornets nest here, but I was hoping for some help. I am in a more laid back guild that works on progression a few nights a week. Currently we are working on the last few bosses of DF HM and DP HM. One of the issues we have been having is our healers telling me that one of our main tanks (long time in the guild and one of our better players) who stacks endurance (51K hps) takes too much damage and requires too much healing.

 

I understand that many of you can clear the current HM ops with either an endurance tank or a mitigation tank. It would help if we were optimized since many of us are casual players. Reading some of those formulas hurts my brain but my personal opinion from looking at numbers from our parses:

 

The extra 4480 hitpoints gained from using endurance augments is only good for an extreme burst phase. In that case time to kill (TTK) may apply since those extra hitpoints may outweigh any damage mitigated. Is it safe to assume that a burst phase (basically in the 40-50k range) that extreme is rare?

 

To me, any damage taken in the raid represents healing resources (GCD's) needed to fix that damage. Therefore, it is best to avoid as much damage as possible as opposed to "absorbing" it with a hitpoint cushion that consumes raid resources to fix. Our healers are usually fighting against the GCD clock to heal up the raid from adds / DoT's. I may be able to leave the endurance tank for 1 extra GCD (he absorbed with extra hps) while raid healing but when I come back, it's going to cost me an extra GCD (and energy/force) to heal him back up anyway right? Wouldn't it be better if he just took less damage in the first place?

 

Thanks in advance for any helpful and positive responses

- Hopeless in Noobville

Link to comment
Share on other sites

A question regarding shelter for assassin/shadow tanks. Is the shelter taken in considerations into assassin/shadow survivability, individually, or not considering it is a buff for healers, but not tanks?

 

While this makes no difference in ops, if you have one assassin/shadow tank. It makes a huge difference in arenas where you only have one tank, thus becomes +5% survivability to all group members.

Link to comment
Share on other sites

its difficult to weight mitigation against endurance. if i was getting three shot by a boss, i might consider using B mods and ask my healers if they prefer my b mod build. using augments or enhancements is a bad trade compared to b mods (1:1 versus 2:1).

 

the 5% healing buff is troublesome to account for, much like oil slick, in that both tanks would benifit from it. i understand for arenas this is not a problem, but arguments could be made that the other teams would kite the other group out of the salvation/shelter etc.

Link to comment
Share on other sites

Rewatch Ryz's NiM Cartel Warlords farm video (it wasn't your first kill). Energy was the gating factor on your healers for the entire fight, especially for your operative. Ultimately, there were two separate deaths from Stabbing Spree which could have been avoided if the healers had any energy at all when they hit that phase. I don't doubt that energy is a non-issue now, given that everyone is over-geared for the old NiMs and the new HMs are very easy.

 

The argument about having extra time is valid, but remember that you lose time by sacrificing mitigation to pick up TTK, since you're taking more damage. See dipstik's calculations on exactly this point. You're gaining time only in the short term; you lose that race very quickly (not quite at the 2 minute mark). Thus, TTK is only interesting for spike phases, which is precisely why I made the point about Driving Thrust.

 

That kill was our first after they reworked it so you couldn't cleave-double-kill cheese the fight, so you can't really use that phase as an example, I think it was the second or third pull after change. The deaths were also not energy related, and since in that phase whoever he happens to start stabbing essentially becomes the tank it provides a poor sample as far as mitigation vs HP/triage. The first phase is a better gauge of how tank HP/mit correlates to raid triage. FYI Ryz plays an incredibly non-standard form of Sorc healing, he is probably the best Sage/Sorc heals on the server but none of us know how he makes it work. He will sometimes call OOE with 75% force remaining, and say he is fine at sub-10%.

 

Overall in NiM fights, the raid ends up taking more critical/spikey pull-ending damage than the tanks. Sometimes this is l2p, madcuzbad, standing in red (Thrasher), others it is unpredictable RNG based LOLWIPE raid spike (Thrasher, lolz). The NiM raid damage is rarely a static X value, compare Thrash's release sniper burst on 72/75 minmax to HM Draxus raid damage for current context as that's the fight where damage was going bonkers without any rhyme or reason. The tanks almost never went down and it was multiple DPS dying to spike in the same pull causing enrage. It had nothing to do with whether or not the heals had energy to spare, it was a matter of if they had the GCD's to triage. Our tanks haven't had a death during progression related to straight lack of heals since pre-nerf NiM DG. Part of that is probably luck, part of it may be that they swapped to HP stacking at the same time. Correlation data can't 100% prove or disprove this unfortunately. Keiki has been healing with me for almost 2 years, Ryz for well over a year. They say they don't notice a difference in healing the tanks, but there are significantly less deaths to tanks. Again not because of resource issues one way or the other, but because they have more GCD's to toss green.

Edited by countpopeula
Link to comment
Share on other sites

  • 2 months later...

This is by far the best topic for a newb tank specialist for a subscribe and rated 5 immediately. Thanks, KeyboardNinja and everyone who contribute.

 

For the mitigation vs. endurance debate. I was the tank that stack HP in every other game before I knew better but everytime I did that, healers had a resource problem which lead us into wipes many times. Only then I tried for the reverse and both in PVE and PVP we got rid off that problem and focused on other things. So, Im telling you this in experience not by numbers and im fairly new to this but same thing applies to SWTOR as well. So, my vote is for KeyboardNinja's by experience.

 

* * *

 

Could you help a newb tank specialist to understand some specific things. I still have too much to test and learn.

 

Internal/Elemental damage bypasses shields and armor, but is subject to internal resist

 

For the record, neither "resist" nor "internal resist" come from gear in any way (aside from set bonuses).

 

1) I guess internal resist isn't in the character sheet, so how can we check this?

2) Which sets gives bonus to internal resist or any resist at all? You mean the ones for Blade Barrier?

3) I want to test that sets for the reduction of internal damage (also after that, with elemental dmg) and how do you suggest for me to do these tests? As I said I'm fairly new to advanced tanking.

 

Sidebar: It is possible for Guardians/Juggernauts to increase their self-healing by stacking power or strength.

 

So, would you suggest taking 6% strength over 2 points of Accuracy for a Guardian who wants to increase his survivability without losing dps/threat in Defense Tree? Considering I'm mostly PvP Tank.

 

More questions, yet to come as I read again and again. And thanks in advance.

Edited by Xtremophile
Link to comment
Share on other sites

1) I guess internal resist isn't in the character sheet, so how can we check this?

 

Hover your mouse over the "Damage Reduction" category in the character sheet. You'll see it broken down into different categories, one of which is internal/elemental.

 

2) Which sets gives bonus to internal resist or any resist at all? You mean the ones for Blade Barrier?

 

Blade Barrier/Sonic Barrier is a heal, not really a form of damage reduction. It is most closely analogous to Force Armor/Static Barrier for sorc/sage healers.

 

The only set bonus which increases internal resist is the shadow/assassin tank 4pc set bonus, which increases it by 2%. Several talents also exist which increase internal resist, such as Jedi Resistance/Sith Defiance. Procs such as Shadow Protection/Dark Protection and Guardian Slash/Crushing Blow also increase internal resist as they in fact increase damage reduction vs all damage and attack types.

 

3) I want to test that sets for the reduction of internal damage (also after that, with elemental dmg) and how do you suggest for me to do these tests? As I said I'm fairly new to advanced tanking.

 

The easiest way is to get someone who can buy the internal proc damage relic and the elemental proc damage relic. Have them hit you with their basic attack (white damage) and look at the yellow damage numbers. Note that one of them will crit more than the other due to the difference between force and tech crit. You can also take advantage of differences in abilities. For example, Shrap Bomb/Corrosive Grenade does internal damage, while Incendiary Grenade/Plasma Probe does elemental damage.

 

So, would you suggest taking 6% strength over 2 points of Accuracy for a Guardian who wants to increase his survivability without losing dps/threat in Defense Tree? Considering I'm mostly PvP Tank.

 

If you're doing PvP, then accuracy is largely worthless. Take the strength talent.

 

In PvE, I think the accuracy talent is worthwhile if only for the opener. Outside of the opener, the strength talent is superior.

Link to comment
Share on other sites

You could use an accuracy adrenal for opener. I've played around with using a surge adrenal for my VG's opener (to coincide with Battle Focus and Shoulder Cannon), which boosts my crit multiplier to around 75%.

 

That…is actually a really really good idea. The accuracy adrenal that is.

Link to comment
Share on other sites

×
×
  • Create New...