 |
| Welcome to Elitist Jerks |
We're testing some new features on the site regarding OpenID registration and coordination with gamerDNA. If you experience any issues with registering an account, please take the time to fill out a report and send it to this e-mail address. We would appreciate any assistance you could provide in making sure everything is functioning as intended. Thanks!
If this is your first visit, please be sure to check out the FAQ and the forum rules. Users must register to post and new registrations are subject to a one day "mute" period to get acquainted with the community.
|
09/21/07, 11:57 AM
|
#51
|
|
Von Kaiser
Night Elf Death Knight
Frostwolf (EU)
|
Originally Posted by Chimera
not that I think it will reverse the trend in your graphs, but shouldn't the table look like this:
0-69 avoidance
70-94 block
95-100 crushing
|
yes and it doesn't 
|
|
|
|
|
|
09/21/07, 12:51 PM
|
#52
|
|
Glass Joe
Tauren Warrior
Laughing Skull
|
Originally Posted by Teez
I just gave that a try and got some ... rather odd results. Apparently, boosting my +hit rating from 88 to 150 (I was putting in some ballpark value to account for the lack of being able to add weaponskill) made my TPS go from 875 to 824 - could it be that you're using the inverse of what you intended to there? I'm having the same thing happening if I increase mob damage, keeping everything else constant. Went from 13k damage before mitigation to 23k and lost considerable amounts of TPS - something seems a little off here :/
|
It actually randomizes; so try running each scenario three to five times and average the results.
Good point. I don't accommodate weapon skill yet.
Edit: so I ran the thing five times with skill rating at 50 and skill rating at 150, and came up with the following:
50 hr
682
687
662
672
694
679 AVERAGE TPS
150 hr
743
743
796
806
751
768 AVERAGE TPS
Last edited by thugthedum : 09/21/07 at 12:57 PM.
|
|
|
|
|
|
09/21/07, 3:28 PM
|
#53
|
|
Glass Joe
Tauren Warrior
Laughing Skull
|
Now, I went back and did some modeling as you indicate comparing 13k with 23k incoming damage.
Keep in mind; this is with my gear, which is entry level gruul acceptable - he hits me harder than anything else does, and I find when actually tanking him that I'm rarely if ever rage starved.
So maybe upping the damage on a relatively high damage boss (all things considered) isn't the best way to go; it may be that even at 13k incoming our sim warrior is able to apply heroic strike to a large majority of his swings, so giving him all that extra rage isn't much of a difference. Or, it may mean my sim has a bug or something. I'm at work and don't have time to look. I did have time to run 10 more sims, 5 @ 13k and 5 @ 23k. You're right. My "incoming rage" indicator goes up, indicating that I am not completely faulty in the way I assign rage from incoming damage; but the tps increase is non-significant to non-existant.
Remember, of course, it is actually random.
13000 inc dmg
rage inc: 4715 4956 4186 4475 4330 AVG: 4532
tps out : 687 729 692 729 703 AVG: 708
26000 inc dmg
rage inc: 7491 7406 7151 8768 8683 AVG: 7899
tps out : 637 719 661 696 717 AVG: 686
Also, in reality, in this gear I am capable of reaching 1100-1300 tps; what with chamois buffs, mongoose, prayer of mending, etc. The sim doesn't take everything in to account yet.
Edit 2: when I go in and drop the boss incoming damage, tps goes down substantially.
Last edited by thugthedum : 09/21/07 at 4:57 PM.
|
|
|
|
|
|
09/21/07, 5:06 PM
|
#54
|
|
Von Kaiser
|
Originally Posted by Berthold
Should give the same results but I will think I will test it this afternoon.
For the Shahraz fight. Can some one give me a very quick glance at the fight concerning incoming damage/attack table/swing timers?
Then I will give it a try.
|
here's a wws if you want it: Mother Shahraz - WWS
she hits our tank for 7k-9k
saber lash hits the tank for about 7k-8k every 15 seconds (assuming 2 offtanks to soak)
tank has about 55% avoidance on both saber lash and melee.
mother has shadow damage beams she throws around, and since the tank should be in full tank gear (no sr) he'll take full damage from them. trouble is one run he might never get beamed, the next he might get beamed a lot. the beams can hit for anywhere from 2k to 6k shadow damage and some of them leave a 2k/tic shadow damage DoT.
my guess is that your simulation will favor stamina even without accounting for the shadow damage beam. when you account for the shadow damage beam that's just going to make it favor stamina even more.
|
|
|
|
|
|
09/23/07, 11:33 AM
|
#55
|
|
Von Kaiser
Night Elf Death Knight
Frostwolf (EU)
|
Originally Posted by Chimera
my guess is that your simulation will favor stamina even without accounting for the shadow damage beam. when you account for the shadow damage beam that's just going to make it favor stamina even more.
|
You are right!..So it seems its always the same:
hard hitting boss-> stamina.. fast hitting boss-> avo.
To support this idea I played around with some formulas to see why that is and I think I found a satisfying solution.
At first a little bit of math.(skip the section if you only want a textual explanation)
MATH START
We are going to calculate the possibility of a boss having a lucky streak with his swing timer as the varying factor.
MT has 17k base hp. 50% base avoidance. based on BT socketing and some gear choices he can decide between 12% avo and 3.6k hp (my guess)
Boss hits for 4k dps. So he will kill the MT after
(MT hitpoints/(dps*swing timer))
number of hits.(rounded up)
The possibity to hit is
(1-avo)
per hit.
So that means for our Stamina tank we get
P(dead) = (0.5)^(ceil((20.6/(4*swing))
and for the avo tank it's
P(dead) = (0.38)^(ceil(17/(4*swing)
MATH FINISHED
Now lets take a look at the graphs

As we can see the avoidance is below the stamina for fast swing timers(< 2). This is because the additional hit (5 instead of 4) that the stamina tank can take doesn't compensate for the fact that the avoidance is less likely to be hit 4 times at all. But as soon as we reach the 2-hit region for the avoidance tank the stamina tank shines, (until he again reaches his 2-hit region).
I hope this explains why "stack stamina after 490 def" is not the rule of thumb for every encounter and that fast hitting bosses should be tanked in avo gear.
Thanks for your attention 
Any comments?
p.s.yes i know it's just a matter of the static numbers at which point stamina and avoidance exchange their positions, but it should be clear enough that and why this happens
p.p.s. and yes i didn't talk about the other beneftis of a stamina heavy tank (easier to heal/ better threat generation etc. etc)
|
|
|
|
|
|
12/24/07, 4:12 PM
|
#56
|
|
Glass Joe
|
I have chosen the side of avoidance, in the way I select my gear, and my gemming. If I seem like I am waving a banner, its because I like the benefits avoidance gives, many of which have been already stated.
Anyway, here is some new perspective.
Given more or less any fight, there is a worst case scenario. That may be getting hit 5 times in a row, or 10 times in a row, or whatever. There is a general 'rule' that the worst case scenario _will_ happen, and you should select your gear based on that worst case situation, and because of that, go stam. You will survive those bursts, and the healers will be much happier.
--Argument 1--
The argument I make is this: Because I am stacking avoidance, my worst case scenario will happen less (from a little to a lot less) than a stamina tank's worst case scenario. Here's the numbers.
I am comparing a 40% avoidance, 21K hp (unbuffed) tank vs my 60% avoidance, 13.5K hp (unbuffed) tank.
I am assuming that any unbroken chain of hits is what leads to a tank's death. For a 4k (after armor) swinging mob, that is six swings on the 20K hp tank (without a heal) and four or so hits on me.
A 40% avoidance tank's chance to get hit six times in a row is .046656 (for any six swings). My chance to get hit four times in a row is .0256 (for any four swings). There are 6/4 more 'any four swings' than 'any six swings' on any given encounter. Applying some questionable math, that bumps the effective worst case rate up to .0384. Roughly a 17% less chance of eating the worst case scenario. I have not done the gear research that some others have, but these are fairly extreme, and hopefully reasonable cases. If you add the fact that most buffs add static numbers of hit points, then you can get the numbers shifted quite a bit.
21K tank becomes 26K tank. 13K tank becomes 17K tank.
6 hits becomes 7 hits. 4 hits become 5 hits.
21K has a .027 chance of eating 7 hits in a row. 17K has a 0.010 chance of eating 5 hits in a row.
17K has an effective worst case rate of 0.014, and will get his worst case 50% less than the 21K tank.
--Argument 2--
Given any two worst case scenarios do happen, there should be heals coming in eventually. It should be plain to see that any healing has a multiplicitive effect on the avoidance tank far greater than the hit point tank. So 'saving' the tank from his/her worst case scenario requires fewer life-saving heals.
--Other comments--
I believe, personally, that the hit point vs avoidance survival argument should not be computed at the single gem level, because hit points give you a very linear survival (can survive 6 hits instead of 5), and avoidance gives you exponential survival (40 to 60% avoidance is a 33% decrease in hits taken, 60-80% is a 50% decrease, etc). I think, instead, one should look at a static number of gem sockets, like 10 or 15, and check the numbers when they are all stam, or all avoidance (dodge > defense, because block is only real avoidance if you are avoiding crushing blows), or mixed.
Even better would be to compare two gear sets that favor one or the other, along with their gems.
|
|
|
|
|
|
12/24/07, 4:57 PM
|
#57
|
|
Piston Honda
|
The fast-hitting models are favoring avoidance over stamina, even as wipe prevention and not total damage taken, due to the fact that avoidance is able to get closer to its theoretical average. If a boss swings 10 times a second for 1k you're certainly going to take some huge spikes regardless of how much avoidance you have, but given a steady amount of healing on you, you are less likely to dip low enough to be killed when you stack avoidance. The problem is that in such fights where mobs are hitting fast and hard you're far more likely to shift healers to be able to heal the stamina tanks up than suggest that they stack avoidance. Fights last quite a long time these days, and you have to stack healing such that in every possible scenario the tank has enough healing to stay alive until the healers are fully recovered. If you have a lower stamina, higher avoidance tank, you can't rely on good avoidance streaks during silences and healing reductions - you just need to have more healers on the tank and/or a better method of dealing with the interruptions because at some point during the fight you're likely to get screwed. If you instead had a tank with more life that can absorb those spikes, you wouldn't have a problem.
In my opinion the "stack stamina" idea works well not as much because stamina is extremely cheap, but more because healers will always have the mana to heal the additional damage you take from not avoiding the hit. Until mana becomes an issue for healers in boss fights, stamina will be king.
|
|
|
|
|
|
12/24/07, 7:12 PM
|
#58
|
|
Bald Bull
|
Avoidance reduces average througput necessary, not just mana consumption. If stamina isn't necessary to avoid spikes, an avoidance tank might under some conditions allow you to get away with ditching a healer for a dps, which could be critical in a timed or pseudo-timed fight. I don't know of any particular fights where that's actually the case (I haven't seen T5-6 content) but it's a possibility.
Personally I think that at least one tank should build an avoidance set, just in case. For practical reasons it will either not be your main tank or not be his primary tanking set, but being able to switching tanking/healing strategies for strange circumstances (including not having enough healers online one night) is worth the effort.
|
|
|
|
|
12/24/07, 7:22 PM
|
#59
|
|
King Hippo
Ermad
Human Paladin
No WoW Account
|
Originally Posted by Aurrior
--Argument 1--
...
I am assuming that any unbroken chain of hits is what leads to a tank's death. For a 4k (after armor) swinging mob, that is six swings on the 20K hp tank (without a heal) and four or so hits on me.
|
The problem with this situation is that it is unrealistic that the tank is going to die without any heals. You didn't consider the fact that the stamina tank would die 50% slower then the avoidance tank in this situation. Which gives him much longer to receive heals to prevent the death. If he is dying to quickly that there is no time to get a considerable heal on him, stamina would help more because the 4 hit string is going to happen eventually and only stamina will prevent death then.
|
|
|
|
|
|
12/25/07, 1:26 AM
|
#60
|
|
Disharmonious
Dwarf Paladin
Lightbringer
|
Originally Posted by Endoscient
The problem with this situation is that it is unrealistic that the tank is going to die without any heals. You didn't consider the fact that the stamina tank would die 50% slower then the avoidance tank in this situation. Which gives him much longer to receive heals to prevent the death. If he is dying to quickly that there is no time to get a considerable heal on him, stamina would help more because the 4 hit string is going to happen eventually and only stamina will prevent death then.
|
In addition, most of the modeling I've seen with higher mitigation tanks usually have realisticly a greater smattering of AC, which can increase the hits-to-death significantly higher, into the 7+ range. The reason this matters is because each heal landed on the tank actually becomes -a larger effective heal-, and that's a pretty huge deal.
|
Originally Posted by bartolimu
It makes me want to hit Marge Thatcher on the nose with a rolled up newspaper.
|
|
|
|
|
12/26/07, 2:44 PM
|
#61
|
|
Glass Joe
|
Originally Posted by Endoscient
The problem with this situation is that it is unrealistic that the tank is going to die without any heals. You didn't consider the fact that the stamina tank would die 50% slower then the avoidance tank in this situation. Which gives him much longer to receive heals to prevent the death. If he is dying to quickly that there is no time to get a considerable heal on him, stamina would help more because the 4 hit string is going to happen eventually and only stamina will prevent death then.
|
In my experience, every situation where the tank can die can be compensated by lots of heals, so I am not quite sure what it is you are saying. I can think of a few situations where pure survival does matter, namely:
Gruul, around phase 7 or 9 or so, there's a ground slam/shatter followed almost immediately by a silence.
In some foolish case, the tank gets out of LOS, or range.
Dual wield mobs that can crush. (Prince?)
Now, yes, in any worst case scenario, the avoidance tank will die first. I can't argue that, because it inherantly removes everything an avoidance tank is trying to accomplish.
The point is that the avoidance tank has worst case situations happen less than a stamina tank. From another post, the only difference gems can make is about 13% avoidance, or 3.5K hp. So lets take that perspective on a buffed tank with 45% avoidance, and 14K hp.
We have a 58% avoidance tank with 17K hp buffed vs a 45% avoidance tank with 20K hp buffed. Lets take a massively hitting mob, doing 15K hits for 6K post-armor damage. The avoidance tank will get hit three times and die, the hp tank, four times and die. The chance for the avoidance tank to get hit three times in a row is 7.4%, and the hp tank four times in a row is 9.1% In either of these cases, the amount of hit points you have does not matter, it is the worst case scenario, and the case where you die.
Even given the less hit points, the avoidance tank will get hit three times in a row LESS than the hp tank will get hit four times in a row.
On top of that, it takes more healing for a stamina tank to get out of the danger zone. That was the point of argument 2. A single paladin who bubbles out of silence and spams flash of light has a much higher chance of keeping an avoidance tank alive than a stamina tank.
----
Now, the obvious argument is, well, heals will be coming in. Fine, lets say the avoidance tank has 6 seconds of healing, which would be, oh, 2 greater heals. The Hp tank would have 8 seconds, which lets pretend is 3 greater heals. We'll assume that all the other healers are otherwise occupied, because otherwise you could heal through it. And lets also assume that after that healing, the healers are silenced, with the tank having to survive at partial health during a worst case situation.
2 greater heals will heal for 8K, having the avoidance tank live through the third hit, a fourth hit, and die again on the fifth hit.
3 greater heals will heal for 12K, having the hp tank live through the fourth hit, fifth hit, and die on the sixth hit.
Now we have the situation where an avoidance tank has to be hit five times in a row vs the avoidance tank having to be hit six times in a row. Checking the math, we come up with this.
An avoidance tank will get hit 5 times in a row 1.3% of the time.
A hp tank will get hit 6 times in a row 2.7% of the time.
That is twice as often. Even if somehow the hp tank gets another heal in there, and lives through the sixth hit, they will get hit seven times in a row 1.5% of the time. That is still higher than the avoidance tank.
That is the power of the avoidance tank. Your worst dreams do not come true very often, if at all.
So to say that the four hit string will always happen, and only stamina will save you is false, because each person's worst case scenario is different, and an avoidance tank's worst case scenario will happen less than the hp tank's worst case scenario.
|
|
|
|
|
|
12/26/07, 2:57 PM
|
#62
|
|
Glass Joe
|
Originally Posted by Oggie
In addition, most of the modeling I've seen with higher mitigation tanks usually have realisticly a greater smattering of AC, which can increase the hits-to-death significantly higher, into the 7+ range. The reason this matters is because each heal landed on the tank actually becomes -a larger effective heal-, and that's a pretty huge deal.
|
Agreed. Armor can make a huge difference, but I don't like the return on investment from 60-75%. I think most druid tanks like to sport about 20K hp unbuffed and 70% armor unbuffed. Most warriors sport anywhere from 13K hp to 17K hp, and about 60% armor unbuffed.
This gives the druid a huge buffer when it comes to getting healed, as he can eat several 20K (raw) hits without dying. This is why they are so loved as hurtful strike eaters on Gruul, because you can ignore them for a couple hits. (Not to mention their threat/rage efficiency)
Armor and hit points compliment each other quite well, but I think the biggest choice you can make in armor is picking a good shield. I picked up the arena 2 shield, and it still has more armor than any shield up to about the T5 realm. The enchants are iffy in my book, but I haven't crunched the numbers.
And yea, the healing efficiency shouldn't be ignored either. Avoidance shares this trait as well, though it is not as obvious.
|
|
|
|
|
|
12/26/07, 3:16 PM
|
#63
|
|
Glass Joe
|
Originally Posted by PSGarak
Avoidance reduces average througput necessary, not just mana consumption. If stamina isn't necessary to avoid spikes, an avoidance tank might under some conditions allow you to get away with ditching a healer for a dps, which could be critical in a timed or pseudo-timed fight. I don't know of any particular fights where that's actually the case (I haven't seen T5-6 content) but it's a possibility.
Personally I think that at least one tank should build an avoidance set, just in case. For practical reasons it will either not be your main tank or not be his primary tanking set, but being able to switching tanking/healing strategies for strange circumstances (including not having enough healers online one night) is worth the effort.
|
Our guild ran Kara with 2 healers under this assumption, and cleared it faster than we ever have before. We are mostly max-Kara geared, so this isn't entirely suprising, but I figured you might want some evidence that indeed, you are correct.
|
|
|
|
|
|
12/26/07, 7:51 PM
|
#64
|
|
Glass Joe
|
I busted out the statistics book, and did some better analysis, and it turns out that at least partially, I am talking out of my ass.
But doing some more analysis, I figured out some strange stuff.
<enter math>
The chance of flipping a coin and getting heads is defined as:
def chance(x,n):
return factoral(n)/(factoral(x)*factoral(n-x))
A coin is fifty fifty. We want an independant probability (percentage), so we use:
def chanceprob(x,n,p):
return chance(x,n)*p**x*(1-p)**(n-x)
These are verifiable at the site: No Title
The hits/swings chart is a chart of the probability of how many hits will occur given how many swings, and is calculated by:
chanceprob(hits,swings,chance hit)
The swings to kill/swings chart is a probability that you will eat at least x many swings given another number of swings. Example:
The chance to eat 3 swings given 4 total swings for tank A is .17188 (the chance to get exactly 3 swings given 4 swings) plus .03112 (the chance to get exactly 4 swings given 4 swings) both for tank A. This gives us .20300, or a 20% chance for tank A to eat at least 3 swings given 4 total swings.
What we find is this.
<exit math>
The bold indicates where an avoidance tank will die where a hit point tank would have lived.
Now the interesting part is, when you look at the chances of getting killed, Tank A (avoidance) has a higher chance to get killed in fewer swings in very certain situations. This sorta makes sense, an avoidance tank can die faster than a hp tank. However, if you extend that out to more swings, a hp tank has a greater chance of dying, because hit points will only help you for so long.
A hit point tank can survive certain very specific situations better than an avoidance tank. This has largely to do with the average hit of the mob. If you can stack hit points such that you breach one of the multipliers of the mob's hit, then you have a larger chance of surviving, again, very specific situations.
That seems to be where all those graphs with all those jaggedy corners comes from.
The odd part is, those areas where hit point tanks thrive are not that uncommon. Imagine a 5 second silence. 6 seconds will have passed against a mob who's base damage is 15K (6K after armor). A hit point tank will likely live where an avoidance tank will die (or be forced to save him/herself). Imagine a 3 second heal afterwords. Hp tank will still have a better chance of living.
Interesting, anyway.
|
|
|
|
|
|
12/27/07, 7:48 AM
|
#65
|
|
Banned
Tauren Druid
Ragnaros (EU)
|
Whoever makes a reliable table with every single factor, including the skill of the other raid players versus percentages of "whops" and "yeah!" - good and bad happenings - has to either be suffering from some mental disease or be well educated.
As some might know, the benefits of attack power and critical strike chance both depend on being X and Y of eachother - they must be balanced (except when it comes to specific class mechanics regarding crit, however...) - And the same relation must be maintained regarding Avoidance and Stamina. The number of factors regarding this issue is grand.
Warriors draw a lot of advantages from Defense Rating. Once they reach Crit- and Crush immunity, they prioritize Stamina to a certain point. When a Warrior reaches that certain point he must decide whether to either go for more Defense Rating or Stamina. And when it really matters: Warriors have many abilities to help them survive during harsh moments.
Druids must rely on stamina and armor. Druids will be hit ...all the time. However, Druids need dodge as well. Druids should have between 30-40% dodge, and then stack stamina (expecting Armor to be decent). There is a point where Druids will not favor dodge, because when dodge occurs no rage is gained. Dodge must eventually be balanced with Hit rating. Dodge Miss Dodge Miss Dodge Miss WIPE ...we've probably all been there.
Edit: Statisticly, disregarding all class mechanics, avoidance costs the least mana to maintain.
If you understood anything of what you just read I believe the rest is selfsaid.
|
|
|
|
|
|
12/27/07, 5:20 PM
|
#66
|
|
Glass Joe
Night Elf Druid
Dalvengyr
|
I think that post #23 is the key post here. In any situation where you are tanking multiple mobs, a high avoidance tank has a significantly greater chance of survival. A corollary to this is that a high avoidance tank also has a better chance of holding aggro on a greater number of mobs at once because healing threat is less. Specifically, people can consider a pull in Shattered Halls where a single tank is tanking a Shattered Hand Centurion and the associated Shattered Hand Gladiators. If a group attempted that pull straight up with no CC and the tank pulled poorly (with most Gladiators at 80%+ health), a tank with high avoidance has a much better chance of surviving than a tank with high armor/stamina and low avoidance. There are other examples in 25-man instances of course, but in the SH example, you can't just 'throw more healers at the problem'.
Personally, as a druid tank, I am more interested in the rage starvation aspects of high avoidance build-outs (i.e. at what level of avoidance does your TPS take a hit) than at what point a wipe becomes a certain. In regards to the comment that you 'gear up at least 1 tank for high avoidance', I think that the protection paladin is a tank that really shines in multi-mob situations and is well suited for a high avoidance build (although, in the case of a paladin, the value of block would also need to be determined).
Outside of the multi-mob tanking situation, I think that a stamina / armor build probably does a better job of wipe prevention than an avoidance build primarily because a high stamina build dramatically increases the efficiency of hots and hots are the best tool a healer has to smooth out spike damage or protect against breaks in the healing flow due to fight dynamics (silence, knockback, forced movement, etc..)
|
|
|
|
|
|
12/27/07, 6:07 PM
|
#67
|
|
Glass Joe
|
I use a 12/6/43 threat build (9 rage devastate/9 rage HS) with 60% avoidance, and generally, threat is not a problem for me. I haven't hit T5/T6 content yet, but generally I outpace most everyone by at least 20%, up to about 100% on long fights. Shadow priests still give me a run for my money, but they have the triple threat problem.
In low threat situations, I pick up a 2.6 main hand, to gain more tpr.
On a good hitting mob, like Gruul, I generally maintain about 7-800 tps, with spikes in the 1K region.
On trash, and other mobs that are trivial... (not merely doable, but trivial), I dual wield. This guarantees about 1K tps, and scares the hell out of some people.
Hell, just devastating once per swing (soloing mobs), I maintain about 400 tps.
Other things that apparantly help:
Bind this macro to your mouse wheel:
/cast Heroic Strike(Rank 10)
/cast Devastate(Rank 3)
This alone can guarantee that you will never ever miss a GCD, or a melee swing where you have excess rage (You can even HS during both swings of a windfury), and I have only seen a couple situations where I have remained at max rage while spamming this.
But then again, devastate feels like cheating any more.
Multi mob agro, I spam T-clap and cleave, while dropping revenge and shield slam on the main target inbetween T-clap cooldowns. Rage starvation is rarely a problem here, and I usually have about 1200 TPS across 4 mobs.
I guess one problem I have is that I run with generally 1-2 priests and a paladin. They only have two HOTs between them. But what you said doesn't make much sense to me. Are the healers afraid of overhealing with HOTs? That is just silly. Armor, yes, please.. but the decision never seems to be: Armor vs HP vs Avoidance, because armor is a no brainer with plate and a good shield. The decision, at least with warriors, is Hp vs Avoidance.
With druids, this may be completely different.
More tricks include the following:
When pulling, make sure to get PoM and Earth shield. These give you threat.
Use your rage generating abilities (Bloodrage, and whatever Druids have) to bridge rage between mobs.
Find out when a mob is threat-locked on to you. I define this as: I have enough threat on this mob that nobody will be able to pull it off me before it dies. When that occurs, stop using abilites and max out your rage. Then bloodrage to make your rage last longer.
As a last resort... sit down or turn around. You'll get hit, and you'll then have enough rage to do what you need to.
Get expertise.
Use a slower weapon that scales better with instant attacks. (May be exempt for bear druids)
Tell the rest of the raid to hold back for a second, or that threat is their responsibility. It causes lots of whining, but its better than wiping. Maybe they will finally pick up those -threat trinkets, or ask for Salv.
Multi-tank some mobs instead of having them CCed. I do this religiously in heroics.
Find the threat per rage of all of your abilites. Use only the ones that are efficient.
Throw HOTs on yourself right before you pull. The ticks will give you some extra initial threat.
Misdirect during battle.
I have never had a problem causing threat, but that seems to be why people like me as a tank. Anyway, back to being on topic...
|
|
|
|
|
|
12/28/07, 6:47 PM
|
#68
|
|
Glass Joe
Night Elf Druid
Dalvengyr
|
To clarify my point regarding hots, the reason why I feel that a high armor / high stamina build increases the efficiency of hots is that after a certain point it is no longer necessary to keep the tank topped to ensure that they can survive any potential bursts. In order words, the healers can direct heal to 85% and then let the dots tick to heal the last 15%. Healing a tank with high avoidance can be a fire drill any time they take a big hit because there is a small chance that they will take successive big hits and die and as such it safer to keep them topped off at all times. Part of my point here is to say that the synergy between the tank and the tank healers is the most important part of choosing what sort of gear to pick. To a certain extent, a person can argue that using a higher avoidance build that frees up healers is 'being easy to heal', but I think that if your main objective is zero-wipe repeatability, raid compositions/gear choices that are built around the least twitch dependent healing strategies will always be the most effective.
|
|
|
|
|
|
01/06/08, 5:59 AM
|
#69
|
|
Glass Joe
|
Something not mentioned yet is the use of "clickables". The obvious downside of increased avoidance (and lower HP) is the bad streak of hits which increase your burst DPS-In. However, many of the bad rolls can be offset with Last Stand, Shield Wall, or a host of other clickies. This gives the alert tank a big advantage over a slow tank when using more avoidance, turning that 5% chance to wipe into a 1% chance. As for the higher HP tank, clickables are somewhat less of a concern because the goal is to overcome all burst damage by simply having the HP to survive, then compensating for the higher DPS-In with more healing. If someone were to create a more in depth analysis, this would just be one more factor to complicate it all.
The only other thing I'd like to mention is that I've seen "I like" used a lot. To be blunt, it doesn't matter what you like when it comes to game mechanics. The game is programmed a certain way, and as a tank your goal is to reduce the probability that you'll die (something people don't "like"). What that comes down to? You need enough HP to survive the max burst damage possible based on the healing you are receiving, then you simply increase avoidance/mitigation as much as you can to lower the DPS-in, thereby reducing the amount of HPS you require. For newer tanks, the gear available now is leagues above the old stuff (16k unbuffed w/ just 10 man/badge items). To the point where stacking HP gems is simply not needed until higher level content.
|
|
|
|
|
|
|