Jump to content


Photo

Tanking enchant discussion


  • This topic is locked This topic is locked
54 replies to this topic

#21 tenshender

tenshender

    Glass Joe

  • Members
  • 4 posts

Posted 03 November 2009 - 06:38 PM

ah im bad at theorycrafting : P, while i realize that statistically blood draining looks to be the best enchant atm, there are a few questions that need to be answered about ICc b before this can be determined. I'm no math wiz and i hate statistics, but i understand and enjoy the mechanics of the game and trying to use them to ones advantage. Maybe one of the better theorycrafters could answer some questions for me.

The dilemma mainly centers around ICc and the supposed -20% dodge debuff.

The arguement for mongoose may be strengthened, and the arguement for blade ward lessened in ICc with the addition of -20% dodge. My warrior is sitting at around 26% dodge. I'm Assuming this is pretty low or that my dodge is high because my parry rating is low, but that this is nonetheless easily achievable by a progression tank. Now, with the addition of the debuff, hed be sitting at around 6%.

The first question is how much dodge% would 120agi give after diminishing returns at around 6% dodge. Im assuming this will be much much greater than 120agi at 26% dodge but i am curious as to how much.

Since parry isnt affected in ICc, my 2nd question is: how much more, if at all, is 120agi worth than the amount of parry given by bladeward.

Since bladeward's parry is still being greatly diminished im hoping to see an ANCIENT enchant beat out something that plenty of people who don't look much into game mechanics as a godsend and the unquestionable best enchant turning the phrase " if you dont have bladeward you aren't a progression tank," which i have heard many times before and just laughed.

Finally i'd like to know if statistically(*cringe*) mongoose or blood-draining is better considering the increased amount of dodge given. I'm guessing blood-draining is probably still favored, but id like to stir up some good debate and get some gears turning in peoples heads.

Heres to hours spent pondering an ancient enchant we should have forgot about years ago, ty blizz!

#22 PantheraOnca

PantheraOnca

    Glass Joe

  • Members
  • 19 posts

Posted 03 November 2009 - 07:16 PM

i believe the -20% dodge is taken after diminishing returns.

if thats the case, you would still compare dodge/parry/agi identically as you would now on a per point basis.

in other words, the -20% dodge doesn't have an effect on the diminishing returns on dodge.

#23 tenshender

tenshender

    Glass Joe

  • Members
  • 4 posts

Posted 03 November 2009 - 08:13 PM

well if the dodge from 120agi is added to the full 26% then diminished, THEN -20% is taken then mongoose is of less value, however can we get confirmation that this happens? it seems to me that since its a constant -20% debuff that the diminishing returns on a periodic buff would be calculated separately

#24 vorpalblade

vorpalblade

    Don Flamenco

  • Members
  • 493 posts

Posted 03 November 2009 - 08:42 PM

well if the dodge from 120agi is added to the full 26% then diminished, THEN -20% is taken then mongoose is of less value, however can we get confirmation that this happens? it seems to me that since its a constant -20% debuff that the diminishing returns on a periodic buff would be calculated separately


That's correct, the amount of dodge granted by a mongoose proc will be the same at your post-radiance 6% avoidance as you would have gotten at your 26% pre-radiance level, based on the current Blue posts that radiance will be calculated after DR. That said mongoose may well still wind up being stronger, simply because its uptime and avoidance contribution is calculable in advance, knowing only your stats (your attack speed, instants, and proc rate).

Whereas part of calculating blade ward's avoidance contribution relies on knowing more about how frequently you'll get hit, since that will help determine the likelihood of an incoming attack clearing your stack, and preventing you from building it higher than a couple stacks. Even once blade warding can be accurately modeled for a given incoming-damage scenario, mongoose still gives you more things overall more reliably (Dodge, Armor, and Crit) and will probably remain the better all-around tank enchant compared to blade ward.

edit:
Here's your confirmation of that, by the way.

You see, the petty rules and regulations for the general forums don't apply here. If you're a fuckwad you will systematically be mocked and embarassed to the fullest extent of our abilities. In short, take your 12 bucks, shove it up your fucking ass, and don't come back until your IQ reaches double digits.


#25 tenshender

tenshender

    Glass Joe

  • Members
  • 4 posts

Posted 03 November 2009 - 11:56 PM

I'd like to take a moment to QQ about Blizzard's terrible foresight and the fact that players suffer with things like sunwell and ICc radiance because they cannot create a balanced game, but i won't.

By all math it looks like Blood-Draining is the best for survivability, Accuracy is best for threat, and mongoose is still the best all around tanking enchant, 4years later....go blizZ!

Thank you kethas for raising some interesting questions.

#26 Executation

Executation

    Glass Joe

  • Members
  • 19 posts

Posted 04 November 2009 - 03:20 AM

Mongoose will help us avoid (0.36% / 40%) * 40k = 360 of that, and will mitigate (40k - 360) * (0.16%) = 63 of the rest, for 423 damage avoided/mitigated.


Mongoose

Anub'arak hits for 27500 on average in 25 man.

Armor mitigation: .16% * 27500 = 44

27500 - 44 = 27456

.36% * 27456 = 98.84

98.84/45000 = .22%

In order for Mongoose to be at least as good as Blood Draining, you need to see 5 stacks of blood reserve less than 1 time out of 20 times that you are reduced to below 35%. Clearly, Mongoose is better than 26 agil to weap. However, depending on how often you get reduced to below 35%, Blood Draining isn't necessarily better than mongoose. In order for Mongoose to pull ahead of Blood Draining, you would need to be getting hit to below 35% more than once every 5 seconds. I'm assuming that between auto attack and 1 sec global cooldown on specials, you should be able to get 5 stacks of blood reserve up in 5 seconds.

Next, I'll try and apply the same approach to Blade Warding. Even if the numbers I'm crunching can be considered to be gross exaggerations of what is happening in a fight, I'm applying the same exaggerations consistently. After these calculations are finished, I will test Blood Draining to see how closely its performance matches up with my assumptions. I think that Blood Draining is really the only enchant that needs to be tested for in-game performance, because it's the only one not based on avoidance chance. It's ability operates on a completely different mechanic, which, IMO, makes it very refreshing :)

#27 bryn88

bryn88

    Glass Joe

  • Members
  • 5 posts

Posted 06 November 2009 - 05:21 AM

I'm assuming that between auto attack and 1 sec global cooldown on specials, you should be able to get 5 stacks of blood reserve up in 5 seconds.


Nope. It takes about 50 seconds since there's an internal cooldown on the stack procs of 10 seconds.

#28 Executation

Executation

    Glass Joe

  • Members
  • 19 posts

Posted 06 November 2009 - 03:25 PM

Bryn88,

You're right, I missed that, thank you.

So the question changes slightly.

In 50 seconds, will you:

A. Avoid the hit that kills you when you are below 35%

or

B. Stay above 35% for 50 second periods throughout a fight.

If you can "stay safe" for 50 seconds, then you have a free smart health pot usable about 2-3 times per fight. I'll take a look at my math and see if I can tweak it to figure out EH ratio to total health with the 10 second internal cd.

It will be interesting how these scenarios play out in ICC. -dodge but could be that could make blade warding that more favorable because it's parry.

However, I heard that they will decrease the size of the boss hits and increase their frequency, so we should actually see fewer times when we are below 35%, allowing us to accrue more stacks of blood reserve.

#29 Sulika

Sulika

    Glass Joe

  • Members
  • 17 posts

Posted 04 December 2009 - 04:18 PM

Mongoose

Anub'arak hits for 27500 on average in 25 man.

Armor mitigation: .16% * 27500 = 44

27500 - 44 = 27456

.36% * 27456 = 98.84

98.84/45000 = .22%

In order for Mongoose to be at least as good as Blood Draining, you need to see 5 stacks of blood reserve less than 1 time out of 20 times that you are reduced to below 35%. Clearly, Mongoose is better than 26 agil to weap. However, depending on how often you get reduced to below 35%, Blood Draining isn't necessarily better than mongoose. In order for Mongoose to pull ahead of Blood Draining, you would need to be getting hit to below 35% more than once every 5 seconds. I'm assuming that between auto attack and 1 sec global cooldown on specials, you should be able to get 5 stacks of blood reserve up in 5 seconds.

Next, I'll try and apply the same approach to Blade Warding. Even if the numbers I'm crunching can be considered to be gross exaggerations of what is happening in a fight, I'm applying the same exaggerations consistently. After these calculations are finished, I will test Blood Draining to see how closely its performance matches up with my assumptions. I think that Blood Draining is really the only enchant that needs to be tested for in-game performance, because it's the only one not based on avoidance chance. It's ability operates on a completely different mechanic, which, IMO, makes it very refreshing :)


If we assume that you are going below 35% at least once every 50 seconds vs Anub HC which seems pretty reasonbale then the heal from Blood Draining will be 400 per 10 seconds since at that rate every stack you get will be used. It doesn't matter if it is used at one stack or at five, as long as you never waste a proc due to being over 5 stacks it will be a total of 400 healing per ten seconds.

If your 44 number is accurate then Anub would have to be hitting you 400 / 44 = 9.1 times per ten seconds which seems quite unlikely.

But look a bit more. Mongoose is, an average of an extra 0.53% chance to dodge, on average. So if we take a situation where you tank anub for 4 minutes and he is melee attacking every 2 seconds for 27.5k on average.

In total he will attack you 120 times. Lets take a slightly low 60% total avoidance figure.

So your avoidance becomes 60.53% on average due to mongoose meaning in 120 swings you take 27,500 * 0.0053 * 120 = 17,490 less damage due to the avoidance on mongoose.
You will also take 44 less damage on 45% of the attacks that hit due to the armor increase. So that is 120 * 0.3947 * 44 = 2084 less damage due to the armor proc.

Total damage mitigated by mongoose, on average will be: 19,570 damage per fight.

In 4 minutes you will use 24 stacks of Blood Draining for 9,600 healed.

So in that particular example the mitigation you get from mongoose is about twice as much as the healing you get from Blood Draining (using the OPs 45% uptime for mongoose figure).

So the question then is how much of the mitigation you are getting from Mongoose comes at a time when you need it? That is going to be directly proportional to how much of the time you spend "in danger" however you define that. All of the blood draining comes when you are in danger, i.e. on low health. You would need to be in danger about 50% of the time for mongoose to provide better mitigation at those times. Whether you are or not only you can decide.

#30 Executation

Executation

    Glass Joe

  • Members
  • 19 posts

Posted 04 December 2009 - 06:57 PM

This is interesting because your argument seems to support Blood Draining more and more as your gear is upgraded. More stamina, mitigation, and avoidance on gear means less time spent "in danger", allowing Blood Draining a greater opportunity to reach 5 stacks.

I was arguing the opposite, comparing the benefits of Blood Draining and Mongooose as a percentage of health. Since Blood Draining does not scale with increased stamina, i.e. it's not healing for a percentage of your overall health, it becomes less meaningful with increasingly better gear. "Better" meaning higher stamina.

However, instead of comparing Blood Draining's heal to Mongoose's dodge/armor based on "overall tank health", I'm starting to think that it needs to be based on a combination of "time spent in danger" and "size of incoming hit", which previous posters have already theory-crafted on.

Something that has not been considered is:
Blood Draining's benefit is effectively working with any type of incoming damage; melee, magic, or AE. Mongoose only works on melee attacks.

#31 Ballistae

Ballistae

    Von Kaiser

  • Members
  • 40 posts

Posted 05 December 2009 - 02:56 PM

...allowing Blood Draining a greater opportunity to reach 5 stacks.

But getting the 5/5 effect once every 50 seconds isn't more hps than getting 1/5 effect every 10 seconds. It seems to me that Blood Draining only loses efficiency when you go past those 50 seconds without a need for its effect.

#32 Warstehgnome

Warstehgnome

    Von Kaiser

  • Members
  • 77 posts

Posted 07 December 2009 - 12:03 AM

The best point is though that when Blood Draining is needed, it is there. You can never rely on mongoose to be there when you really need it.
Long-time Tankspot.com member Wars

#33 Gorazin

Gorazin

    Piston Honda

  • Members
  • 106 posts

Posted 11 December 2009 - 12:23 PM

With regards to the changes in 3.3 including the debuff Chill of the Frozen Throne. The reduction in avoidance would seem to reduce the effectiveness of mongoose's agility boost. Also bosses hit comparatively less now and so tanks should be taking less spikey damage, which should mean blood draining is consumed less often and therefore provides less frequent but larger heals.

#34 Jumai

Jumai

    Von Kaiser

  • Members
  • 33 posts

Posted 23 December 2009 - 11:25 AM

This question has plagued me for some time. While I don't have a conclusive answer (wtb mathematical model of blade ward), I can share some thoughts which otheres here seem to have overlooked.

First up, in a raid-buffed situation, unless I've missed something Mongoose is worth 132 agility due to Blessing of Kings. Everything I've seen above calculates Mongoose's value from 120 agility.

Second, though it's obvious, I feel like it should be stated explicitely that when comparing Blade Ward to Mongoose, your mileage may vary. All of these damage-mitigated totals and similar stats seem to be calculated using entirely arbitrary before-proc stats. Since the two enchants are affected by seperate diminishing returns curves, this arbitrary base stat array can't be used to extrapolate generalized payoff ratios (nor. without a complete breakdown of stats used to model, can what we get be normalized for a parry/dodge split balanced for DR). Generally speaking the more parry rating you have, the worse Blade Ward is going to look compared to Mongoose, but this is a loose rule and there is plenty of room for odd cases.

Third, Icecrown so far seems to favor fights that involve tank-swapping or time-dispersed damage to the off-tank. Due to the way Blade Ward behaves when you're not being attacked this causes mayhem on the math. For example, when evaluating Blade Ward on Saurfang, the buff uptime is a useless number: because you will never parry when you're not tanking him, more procs will run down to time and more buff timers will refresh due to stacking, but this increased uptime takes place mostly while the buff does nothing. Calculating exactly how much timer refreshing extends off-period procs into the live period on average should be doable for warriors, but it's a messy probability calculation I don't have the patience for in the early AM hours. If you care, note that off-period Mongoose procs still provide the threat benefit, unlike Blade Ward.

Fourth, I'd just like to point out that the Blood Draining advocates in this thread seem to be enamoured with its being "always there when you need it", but to my eyes that is hardly more certain than Mongoose or Blade Ward. Just off the top of my head, here's some questions that aren't easily answered:
How often do you take damage > 35% of your health pool?
How often do you drop below 35%, then catch a wave of overhealing before the next hit lands?
How often do you drop below 35%, then get overkilled by 1801-2200? Blood Draining at 5 stacks will fail to save you from about half of these.
How often do you drop below 35% and catch less than 5 stacks of Blood Draining?
How often are you saved from death by Blood Draining, then killed anyway before your healers can stabilize you?


After stirring the pot so much, I can at least contribute one or two certainties. Provided I haven't forgotten anything, Mongoose proc should yield 295.68 armour (132 agi @ 2 armour per agi = 264, +10% 5/5 toughness, +2% meta gem). Mongoose proc should also yield 1.795% dodge before diminishing returns (132 agi @ 73.53 agi per 1% dodge).

I can agree that when using mongoose, I usually have within 1% of 30% uptime for tank-and-spank boss fights.

I can state that I enjoy the fact that my mongoose procs never wholly "miss" (provided I'm being attacked) due to the armour component, and suffer from less spikey behavior than blade ward's consumeable buff.



I can also suggest the basis for a strategy to employ for analysing Blade Ward (which would require a full parse to use)... please check my logic. Here's my thought:

thesis: avoidance value of blade ward = (parries caused by blade ward) / (times attacked when able to parry)

define:
(parries caused by blade ward) = (times a stack of the buff was consumed) / (percentage of your parry chance contributed by blade ward (pPC))

pPC = (average parry chance increase from blade ward at buff consumption (aPI)) / (average parry chance at buff consumption (aPC))

aPC = (("base" raid-buffed parry chance) + aPI)

aPI = (average parry chance increase from 1 stack at consumption (aPI1)) + aPI2 + aPI3 + aPI4 + aPI5

aPI* = (increase in parry chance at * stacks from parry chance at 1 less stack) * (percentage of buff consumptions which occur with at least * stacks active)



The number you get using this will obviously be a literal-value-in-this-case number, but averaging out the results from enough fights should give you a real expected value for your enchant. What I've posted is essentially psuedo-code, so if anyone familiar with combat log parsing wants to automate the algorythem it should be straight forward.

#35 Cobeathris

Cobeathris

    Glass Joe

  • Members
  • 6 posts

Posted 23 December 2009 - 10:14 PM

The thing is, Blade Ward/Mongoose do something different than Blood Draining. Its like trying to compare an Avoidance Trinket to a Stamina Trinket. Both of them provide something valuable, but they provide that value in a different context. If you want more avoidance, you should go with Mongoose or Blade Ward, if raw health is what you need, then you should use Blood Draining. I strongly doubt that there is enough difference between them to use one type when gearing according to a different philosophy.

#36 Jumai

Jumai

    Von Kaiser

  • Members
  • 33 posts

Posted 24 December 2009 - 04:33 AM

My point is that that outlook makes no sense to me. Blood draining is not "raw health," it is a chance of a variable-size heal, much like blade ward is a chance to not suffer a variable-size hit. Your example of stam vs avoidance is a different case: the reason we consider stam/ehp seperately from avoidance is that avoidance is RNG based. Blood draining can and will still fail for reasons outside of player control, so it's really more comparable to avoidance than stamina despite having the word health in its tooltip.

#37 Cobeathris

Cobeathris

    Glass Joe

  • Members
  • 6 posts

Posted 26 December 2009 - 02:58 AM

Right, it is a chance of a variable size heal, and you do not always know what the heal will be, but unless you are constantly flipping back and forth between over 35% health, and under 35% health (which will still result in small heals about every 10 seconds or so), or you go from over 35% health to dead in one shot, the chance that you will get SOME type of heal from it is nearly 100%, not having at least one stack up while attacking something just doesn't happen. To my knowledge, there is no stamina or health enchant for weapons, so if you want health on a weapon, Blood Draining is pretty well your only option. The lowest it can heal you for is 360 health. Looking at a few of our guild's parses (Heroic Northrend Beasts and Lord Marrowgar), it looks like the average heal I am getting from it is about 1200. On the low end, that makes it worth about 1 Solid Majestic Zircon, and on average, that makes it worth about 4 of them. Yes, there are cases were it will be worth nothing, but, since I can't get health on my weapon in any other way, I am willing to deal with that.

#38 Ganj420

Ganj420

    Glass Joe

  • Members
  • 16 posts

Posted 31 December 2009 - 07:41 AM

Myself, I just do not see a lot of fights were I am constantly dipping below 35%. Usually it only happens a few times per boss. The problem I have with avoidance enchants is that they "may" save you when you take a big hit, but you can't rely on them. It's not a huge heal, but it honestly doesn't have to be.

Bosses are being tweaked so that they do less damage and hit more often so that the overall DPS output increases, but you won't get two shot like on Gormuk or Anub. Tanking is still all about effective health, and right now Blood Draining is the best enchant to increase your EH. It's also works against both magical and physical damage.

That being said, with the size of the heal being so small, it's still very weak, as are most of the other tanking weapon enchants in my opinion. I doubt that using Blood Draining vs. using another weapon enchant is going to make or break most encounters.

#39 Cobeathris

Cobeathris

    Glass Joe

  • Members
  • 6 posts

Posted 31 December 2009 - 04:23 PM

That being said, with the size of the heal being so small, it's still very weak, as are most of the other tanking weapon enchants in my opinion. I doubt that using Blood Draining vs. using another weapon enchant is going to make or break most encounters.


Agreed, I wish there was something better/more interesting for tanking enchants.

#40 Jumai

Jumai

    Von Kaiser

  • Members
  • 33 posts

Posted 04 January 2010 - 07:47 AM

Well, it doesn't have a thread because it's important, it has a thread because it's confusing. You're not any more likely to wipe a raid because you chose Greater Inscription of the Whatsit over Heavy Borean Armour Kit for your shoulder enchant, but the only reason we don't bicker over that is we all feel well informed enough to decide. When it comes to weapon enchants, many of us find the choices difficult to evaluate because discovering what they actually do is computationally hideous.

RE: blood draining and the EH arguement. If you're going to play the hardline EH card, you want 26 agility. After talents and raid buffs that's going to net you ~62 armour 100% of the time. Ok, it's insignificant EH, but it's EH. Blood Draining isn't.


Is there a recount-style addon that tracks your HP totals cross-sectioned with a timeline? It would be nice to know how often I am between dead-from-average-hit-damage and dead-from-average-hit-damage-plus-3200, how long I sit there, etc. I always check my death log on boss fights, and I almost never get overkilled by less than 5k, but there's a selection bias there.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users