Hypatia, looks quite interesting. I'm looking forward to seeing more specifics when you're happy with how your app has been crafted. :)
(Although I'm a little unclear on the graph and axis stuff...!)
My only critique of the test is that given the numbers I've run on the average 264 tank, it seems like you may have created an edge case where Stamina would have a very distinct advantage.
As far as I can tell, in a worst-case string of high-end attacks combined with the correct offset of heals would result in a near 100% death scenario: with 2 primary hits, 1 large periodic, and 1 small hits within the same period as 2-3 periodic small heals and one average-sized heal. In this edge case, the margin of death is actually fairly small, and even minor amounts of Stamina would cause the tank to survive until the next large heal internal.
So, at first glance, it would seem as if this test would favor Stamina naturally. This is not to say it is an invalid test (because such scenarios do happen) but it may be a scenario where the end result is somewhat predetermined.
If looking to simulate something like this, I would consider (and may actually spend some time, now that you've put the idea out there!) modeling a more normal boss scenario. In general, the parameters of the boss incoming damage are fairly constant, and the chance of large-scale alignment of hits is fairly uncommon--Blizzard usually purposfully removes such situations, e.g. Gormok and Thorim adjustments.
Instead, I would say the typical cause of tank death is large sources of damage or even normal damage where healers make mistakes or get interrupted.
So, I would actually propose a simulation where the incoming damage does not predetermine death simply due to a worst-case burst scenario--but a scenario instead where typical healing and interruption patterns were roughly modeled. Presuming the tank already has enough Stamina to survive decently (as that is at least the minimum criteria if you know what you are doing) provided nobody makes any major errors.
e.g. Paladin with average size heals using Holy Light spam with average crit rate percentage. Throw in some basic modeling for Divine Plea and refreshing of Jugements, and perhaps a random 'don't stand in bad stuff' movement interruption at a maximum of every X seconds for Y seconds.
While this technically may result in a similar result, I don't think it would be quite so skewed towards an edge case where having only a small amount more health would significantly decrease the chance of worst-case alignment issues. (That said, I'm only presuming that knowing what you said--but given the damage figures and average ilevel you mentioned, it seems likely that it could be the case.)
I may try to conjure up something over the weekend if I am feeling industrious. :)
(I do find it interesting, though, that even with this potential edge case in the scenario, 9 Stamina socket bonuses seem to cause the tradeoff to roughly equal out for Warriors. That actually doesn't seem amazingly far off from Burst Time style assumptions.)
You're absolutely right. The approach consisting in an interruption of heal causing the death of the MT seems more realistic from gameplay experience. I'm going to model that.
Note that I'm still looking for good advices. :D There are so much parameters, and yes the results I've obtained are only concerning a specific hypothetic boss...
Yeah, definitely. Simulation is tough business, just because of all the factors involved. I'm happy that Rawr is primarily formulation at times! :P
I ran several simulations on a 264 prot warrior. I estimate the value of 1 gem of stam versus 1 gem of dodge by dividing coefficients of the curves I used before. For exemple in the 120 70 (which are the 2 main boss attacks), the equivalence would be 1 stam gem = 3.1 dodge gems with 0% failed heals, 2.3 with 10% failed heals...
It clearly shows that under pressure, I mean with failed heals, avoidance becomes more and more important although still inferior to stamina. I must work on statistical errors now for better precision.
0 10 20 30 (% of failed heals)
120 70 3.1 || 2.3 || 1.97 || 1.69 [Big boss attacks]
100 60 5 || 3.17 || 2.34 || 1.86
80 50 13|| 3.12 || 2.46 || 2.07 [small boss attacks]
Note that current values are still not correctly verified but it gives an idea.
(On a side note, let me know if you want me to open a new thread, as I might deviate a little)
While I definitely don't mind this in my thread because it is very fascinating and quite relavent, more people might see it if you made a new thread. Up to you really.
Great work so far, far better than I could ever do.
Btw, that's Hypathia not Hypatia, our ever loving author of the effectiveness of avoidance.
I know this is a very loaded comment for some people, but it is not intended as a "troll". It's simply in the interest of contributing to the accuracy of the information out there, because people read this stuff and turn it into "conventional wisdom":
Isn't this analysis really specific to Warriors? Doesn't the presence of Ardent Defender radically alter the calculation for Paladins? If the point of single-mindedly stacking EH is to avoid the worst case scenario of getting insta-gibbed in a short window of high damage, it may be more of an issue for some classes than others.
If so, don't the relative values of armor and dodge, which do the exact same thing -- reduce incoming physical damage -- become much more of a close call?
I agree wholeheartedly that tanking does not lend itself to mathematical analysis nearly as easily as DPS does. It is impossible to model all of the different scenarios and considerations you have to take into account as a tank. That's why I read this kind of theorycrafting and take it into account, but when it comes to making gear changes I go by "feel". Over the course of enough fights, if I feel like I'm lacking in one area, I'll shift a little in that direction until I'm comfortable. If I'm taking the biggest hits the current bosses have to offer and not dying from them, to me it's safe to start adding some avoidance.
k im gonna ask a few simple questions.
Do you dodge EVERY physical attack?
Does the damage reduction from armor apply to EVERY physical attack?
Right...but I think you're leaving out the fact that when you dodge an attack, you dodge 100% of the damage from that attack.
Originally Posted by Davih
The two stats have an identical % benefit over the course of time. 1% of dodge and 1% of armor both reduce 1% of incoming physical damage.
The distinction being made here is that in worst a case scenario you don't want to rely on dodge, which has an element of randomness. But the more talents/trinkets/CDs you have to deal with that scenario, the less likely it is to happen, and the less important that distinction becomes.
And you dont dodge every attack. With 34799 armor (my armor with imp devotion aura, JV normal and CSK as trinkets) i reduce all physical damage taken by 69.55% (gonna round it up to 70% because its easier to type)
Would you rather reduce all damage taken by 70%, or reduce some damage by 100%?
If you had 70% avoidance, you would reduce 70% of all damage by 100%.
Originally Posted by Davih
PM him, i really dont know how to explain it well and he has much better knowledge and has already explained it to several people before making this thread
Originally Posted by Aggathon
Well, I'm not really arguing against what he's saying. I'm just saying it may be incomplete. If the thread was titled "Why Warrior tanks should stack EH to avoid being two-shotted" I wouldn't question it at all.
Originally Posted by Davih
Oooohh no they don't. Dodge lowers the average physical damage taken over a given period of time, by taking say 1 out of 10 hits out. Armor reduces damage taken per hit, and over that period of time you may get hit all 10 times, but each time you only take 90% of the damage you otherwise would have taken.
Originally Posted by Thaurissan
The thing with effective health is that it interacts with healing. If bosses hit for 10% of your total life per hit, then dodge would be an awesome stat because there isnt one particular moment during a certain length of time where you are at risk of dying. This unfortunately is not the case and bosses tend to hit for just over half a tank's health. This means that we get hit once, we then expect to get healed up. Worst case would be that you only have a few hots etc on you and you don't receive any direct heals and you therefore (assuming the boss hits just over half our health for arguement's sake) get healed only a small amount, and you take the next hit and we only just survive dying. The thing that helped us survive was a combination of armor to reduce the damage, and stamina to prevent the hit from taking us to 0 hit points.
Look at the scenario where we replaced armor for avoidance. Even if we gained a substanial amount of avoidance, in this worst case scenario, not avoiding the 2 hits, we would die because we didn't have the armor or health to save us.
We're are not concerned currently with taking less overall damage in a fight, but we are concerned with surviving the fight and surviving these worst case scenarios. Currently to survive these worst case scenarios we have to gear totally for effective health. In future content there may well be a level at which we can assume to have enough effect health to survive encounters and can start stacking more avoidance or threat even.
If you looked at ANY major guild's tank during ToGC, you would see that they were gemming for most bonuses with dodge/stam or defense/stam. In ulduar we gemmed more for threat, with things like expertise/stam and hit/stam. The worst case scenario was survivable with gemming in such a way. In ICC it just isn't.
But sure, the 30% buff is coming pretty soon and we'll all over gear content and you can gear for whatever the hell you like.
EDIT: In before "but Ardent Defender!!". Worst case scenario is that AD has already proc'd and it is on cooldown and so are your other cooldowns. But please stop derailing this thread and if you don't agree with what Agg wrote, PM him to discuss it.
Ardent Defender doesn't just save you from a killing blow, it reduces all damage which would take you below 35% by 20%. So if I have 60k HP and I get hit for 30k unmitigated twice in a row, the second hit only does 24k. If the proc goes off, it usually will put me back up around 35%. If the proc is on CD, that means I've been in the worst case scenario twice, or for a very long time. It seems to me it would have to alter the calculation. My point is not that what has been said here is wrong, but that it probably only completely applies to Warriors, since they seem to be the only class that's being looked at.
Originally Posted by Passive
On a deeper level though, the question isn't just whether "EH is better than avoidance", but how much better it is. If we're talking about real world application, which usually means choosing between two pieces of gear, or a particular combination of gems/enchants, we need to know at what point dodge becomes better than armor. I'm not sure it's even possible to accurately model that, which means the whole topic gets oversimplified into:
"LAWL! EH is better than avoidance N00B!"
Faulty logic is faulty. I get it, you'd rather have guaranteed reduction on every hit than take zero damage from some, but that alone is not sufficient to say avoidance is somehow subpar or worthless.
Originally Posted by Davih
There are ways to do that, but you aren't doing them, and it isn't the point of this thread.
That might be right, but by reaching 70% avoidance(block not included) you have sacrificed a lot of EHP through gems/enchants/gear choices. This would leave you in a very dangerous situation if you took a full streak of unlucky hits and the avoidance didn't save you on any of them.
Originally Posted by Thaurissan
There was a great post on this thread about the chances you could take an unlucky streak of hits when gearing towards avoidance without having your "life saving" cooldowns available.
I really don't see how AD changes these calculations at all, I simply see AD as an equivalent cooldown to say shield block. I think the math and other testimonials speak for themselves. Look into the work Hypathia has been doing, so far it seems to indicate that 1 30 stam gem is over twice as effective as a +20 dodge gem for all classes. How good varies between classes, but it's still better.
Edit: also now that I've had a couple months of playing a healer class under my belt, I can tell you without a doubt tanks that stack EHP are easier to heal for 2 reasons:
1) Less EHP seems to translate to more overhealing. When I cast a pennance against a low HP tank, usually only 2 ticks heal it. For that same cast i can heal a tank with more stam for more hit points and I can spend extra globals shielding the raid since the tank can take a couple more hits before I need to drop a big heal on him.
2) I've had tanks that stack avoidance, and they're fine and dandy then all of a sudden they seem to just drop 100% or near 100% to dead before I can even shield or FoL them. That almost never happens with tanks that stack EHP. EHP tanks are sooo much easier to heal, at least from my perspective.
The funny thing is, with an avoidance tank in ICC, you would probably spam heal them even more than you would an effective health tank because of that *just in case* or worst case scenario. This is counter-intuitive to the avoidance tank taking less damage over a fight and therefore requiring less healing.
We really do need to clear up this distinction, because Satorri had a post about it that I really need to get around to linking on the main page, but basically it comes down to this:
Avoidance is NOT BAD. At all, avoidance is quite good.
Sacrificing large amounts of EHP for avoidance IS bad in most scenarios, especially in ICC25. If it wasn't for CotT then avoidance would be even better (as is evident by the graphs) however we cannot get far enough down the graph in ICC25, even if we stack avoidance, for it to be worth it.