Greedy Goblin

Friday, July 8, 2011

Burn phase

The PuG update: I found an extremely good Hyjal Daily guide!


Many bosses have "burn phases", when there is nothing else to do, just Patchwerk like stand and go all out under Heroism/Bloodlust. The first 3 Firelands bosses all have it. Beth P2, Rhyolith P2, Shannox after both dogs down are all stand and burn. These phases are there to test the damage dealers, if they are not good enough, the healers go out of mana and the team wipes. If the boss died, it's a clear proof that the raid has good DDs.

Except it's a big fat lie. A purposefully created illusion by Blizzard to make the majority of the playerbase feel good about themselves without ground.

To see why it's a lie, let's first see the Rhyolith-Shannox type burn, where the boss does constant, huge damage causing the raid HP to constantly drop despite healer effort. Let's see the variables:
  • B: boss DPS to the whole raid (for example Rhyolith does 8K every second to everyone, so B=80k)
  • H: healer HPS in the burn phase, usually meaning AoE heals
  • h: number of healers
  • D: DPS of the damage dealers (for simplicity, I assume that a tank does D/2 and there are 2 tanks out of 10)
  • d: number of damage dealers
  • 1400: the HP of the raid in k (everyone has 140K)
Every second the boss does B damage and the healers heal H*h, so the raid loses (B-H*h) HP every second. The raid will wipe after 1400/(B-H*h) seconds. As the tanks do half DD damage, and there are 2, the raid DPS is 2*D/2+d*D = (d+1)*D. As there are 10 people, d = 10-2-h, so the raid DPS is (9-h)*D. Until the moment the raid wipes, the raid is doing 1400/(B-H*h)*(9-h)*D damage. If it's larger than the boss HP, the boss dies before wipe. This equation seems to prove that burn phase test the DDs, as 1% increase in DPS makes 1% higher damage before wipe.

However let's check the other two variables, H and especially h. For the calcualtion, let's use the B = 80K raid DPS of Rhyolith, and D = 15K for damage dealers:


As you can see having one more healer always increases the raid damage, because the extra healer increases time to live more than he decreases DPS. Several curves ran out of the chart, it means that the healers are outhealing the boss, the raid doesn't lose HP.

But the really nasty stuff comes from differentiating these curves. The chart below tells how much the raid damage increase if the DDs or the healers increase their performance by 1%:


1% DPS increase always gives 1% damage done. However 1% healing increase give much more damage to the raid before wipe. So exceptional performance by the healers is much more important than by the DDs.

Now let's see gradual burn phase where the boss becomes stronger and stronger, like Beth'litac. She does 20K damage to all every 6 seconds and gets frenzy effect, 5% damage bonus every 4 secs. For the simplicity of the calculation, I align the two effect, making frenzy stack for 7.5% damage every 6 seconds. The calculation is much more complicated, so I can't make nice charts, just calculated at some points:


As you can see both for Beth and her stronger version, who pulsate for 30K, the highest raid damage is somewhere between 4 and 5 healers. The difference here is that there is an optimum, adding further healers make the raid worse. However the optimum is still above the "standard" raid composition. 

I'm not saying that in raids DDs are mediocre or don't have tasks. There are several mechanics that demand everyone to react and also tasks for some DDs. The point is simply that the "soft enrage" mechanics are all healer-testing and not DD-testing. There is a well-known DD-testing mechanic: hard enrage. It gives a simple, easily calculatable DPS barrier between good and fail. It's quite telling that this mechanic is no longer used. Why? They told it quite clearly: "On any given encounter, we tend to give responsibility to a few DPS players instead of all of them, and we think that’s ultimately a good thing. Not every player wants a ton of responsibility and we don’t think it would be good for them, or the game, for us to force those players into high-pressure situations. It is a game after all – it’s supposed to be fun. If challenging is what’s fun for you, well, that’s what Heroic modes are for. We think most players understand that taking on the healing or tanking roles is going to come with more responsibility, and those roles in turn tend to attract players comfortable or interested in having more responsibility."

The "soft enrage" is cleverly designed to create the illusion in the mediocre DD that he is actually good, and the boss died because of his awesome performance, while he didn't do anything a trained monkey couldn't do, and the raid would be better off replacing him with a healer (even with a barely OK one). Soft enrages are made to hide the fact of his "low stress" (read: easy) position from himself. I mean in Rhyolith P1, it's obvious that there is little job to a DD than not standing in the lava. The healers must keep everyone alive against the stacks while the pilot tries to stop them coming by destroying volcanoes. So if Rhyolith would be P1 only, the DD would notice that he did not matter. P2 fills him with the sense of importance and awesomeness, without any ground. Of course Blizzard could (and did) create fights where DD really mattered (hi there, Brutallus), but then majority of the raiders would have to notice that he actually suck.

There is only one question left: what will happen if top guilds figure out that stacking healers help? I'm not sure it it's true, as hard modes may have hard enrages, like Atra breaking shields. However if the standard setup would be 4 healers, the DD bunch would not just be outraged, they would be without guild as the shift from 2:3:5 to 2:4:4 puts 40% of them out of business. And this is the ultimate point: you can't make it hard for some and easy for the others, they will simply kick one "easy" and replace it with one "hard" to even the load for better results.

32 comments:

Sean said...

I don't agree with this post at all and it illustrates how numbers and analysis without linking to actual situations can be misleading. What a soft enrage does is test both healers and damage dealers. This is opposed to a hard enrage where it tested only DPS. A soft enrage allows healers who were conservative in their healing to last through the phase and thus give more "DPS time".

The flaw with the analysis lies in this comment: The raid will wipe after 1400/(B-H*h) seconds.

That is not how it works since HPS has rarely anything to do with it. In all fights, it is entirely possible to keep everyone up indefinitely with infinite mana. Rather, the raid will wipe when the healers run out of mana. Thus, a 1% increase in DPS is actually quite big especially when the boss gets stronger and stronger (e.g. Beth'litac).

A soft enrage is a superior mechanic since it allows the whole raid to have influence on the outcome. Contrast this with the hard enrage where the healers sit there twiddling their thumbs until the DPS manages to squeeze out that extra bit of damage.

Anonymous said...

Interesting Discussion, as far as Rhyolith goes, however, you need the DPS in Phase 1 to burn down the different types of adds in time, while still steering and dpsing the legs Beth has a similar mechanic in her phase 1.

What Azzur said is also true, you can't just look at raw output, less dps also means a longer Phase 1 (and more healers to heal it) and that means healers might be lower on mana (or higher) than normally.

Finally, wether your calculations are true or not entirely depends on the values you chose for hps, dps and boss damage. 15k is an incredibly low dps average as it is. Considering you gain both hero and the effect of a potion, you should be expecting something around twice that much dps.

Anonymous said...

So, Gevlon, are you saying that in most casts you'd be better off dropping your worst performing DPSer and replacing with a 4th healer, and just attempt to out-last the soft-enrage?

Alrenous said...

Huh. So healing is probably innately overpowered, due to scaling issues.

Take a 2v2 between a healer/damage duo and a double DD.

Say it takes 100 seconds for the damager to die. To double that time, the healer has to gear/skill up to heal half of the remaining damage. If the healer was healing 50% of the damage, then they need to improve by 50%. If they were healing 75% already, they only need an additional 17%.

To cut a fight's time in half, DPS always has to double. (Say it would take 200 seconds to kill both opposing DDs, and the healer goes OOM at 100.)

Put together, each point of DPS is worth less than the one before it, while each point of HPS is worth more than the one before it.

Conversely, if they were healing less than 50% they would suffer the reverse scaling issue...but for a healer to noticeably affect a fight, they basically have to be in the latter half of the curve, past the point it crosses the DPS curve.

And indeed, to balance PvP at roughly the 3:7 ratio, the healers have to heal nearly half the incoming damage from ten DDs. (And run out of mana only near the end of the fight, so it doesn't end up being one half-dead DD v 3 healers.)

These pvp balance calculations translate directly to raid balance. Healers scale harder than DD, both with gear and with population.

Armor suffers harsh diminishing returns, and healers would need the same sort of thing to get balanced.

Alternatively, you could have the boss heal themselves but have a small health pool, just like a player. If you can't out-dps their healing, you have to run them OOM. But if you can edge your dps over their healing, they'll go down relatively fast. (Think about the boss healing in bursts - it works out well to test DD skill.) A solution like this would provide a competing interest in stacking DDs.

I don't think tanking is well designed either. (First boss that realizes sunder isn't that threatening would wipe every raid.) I think the whole premise of both is pretty dumb.

Leeho said...

You have a very bad dps and good enough healers. Actually on soft enrage with hero up, cds up, execution abilities used - D value should be around 30k. I've checked our logs, it's 25-37k dps on burn phases for FL bosses. Tanks are going like 7-8k. And at the same time you pick up the right numbers for heal, on our attempts it's really like 20k, up to 30k for trees due to their insane hps cds. And we have really good healers.

So your numbers are just biased - you take hps of hc guild healers and dps of random pug strangers. For sure good healers are better than so bad dps (:

Maybe it just shows up that aoe healing is easier than dpsing, or less gear demanding. If we assume that your data is taken from equally skilled people, we get that when dps in casual guild is underperforming by 50% compared to hc guilds, healers are performing for 66-100%.

chewy said...

"what will happen if top guilds figure out that stacking healers help?

By this question I assume you mean what would happen if the extra healer becomes the usual raid make up ? Quite simply Blizzard would change the mechanic again.

Their objective is to make money and you don't make money by upsetting customers. If the game requires a bias towards healers and a majority of people don't want to play healers it's hardly going to be a smart move.

Grim said...

Hard enrages are still here. Shannox has one, for example. I've only downed 3 bosses in Firelands so I'm not sure if there are others, but throwing one in the face of anyone who shows up to a raid hardly constitutes not using them anymore. (also, there were hard enrages on almost all bosses in 4.0)

Also, Rhyolith P1 is all about dps and all of them not just the one steering. Its even harder than the usual "kill it `till it dies" dpsing, because it requires pacing and managing cooldowns for when it is necessary to turn him faster, rather then just burning everything on cd.

Beth is easy mode, enough said.

Haven't killed any other bosses yet, but it seems to me that DPS matter a lot.

Tanking, though, is mostly a snorefest so far (except Riplimb tank).

Kahalm said...

just out of my head: at the first kills of Lich King hero they reduced the numbers of healers to meet the dps required, or heroic Baleroc was 4 healed in 25m mode

Anonymous said...

I concur with Azzur. Additionally, you don't consider tank damage at all in your calculations, which makes the whole hypothesis flawed as well.

The damage then tank(s) get in those encounters isn't trivial. You can't just have all healers go into "AoE mode".

Gevlon said...

@Leeho: that's kind of the point. Except on the very top level, when 1% DPS increase means world first, the guilds are having much worse DD than healers and tanks. Most probably it's true for our guild too.

@Chewy: they just can't. The customers want to LOOK awesome while being mediocre. If they change the mechanic to hard enrage, they prevent the 4-healing idea, but most guilds couldn't get a single boss down.

@Grim: hard enrages are extremely lenient today. I mean you need several below tank-DDs to hit it.

@all about "15K DPS is low": I just picked a number. DPS values just increase the numbers but don't change the optimums or ratios.

Ayros said...

Gevlon, you have overlooked a posibility of tank being one hitted.

Also I don't think that no one in top guilds haven't already thought about using more healers. The thing is, that they can get by with only few healers, because they are players of exceptional skill. And less healers means more DPS, and faster kills. And you need that speed when going for first kills.

maxim said...

@Gevlon

Beth's damage stacking mechanic amounts to the raid getting 15% extra base tick damage every 6 seconds, on top of everything else that was accumulated.

Consider this:
Tick 1 (0%) and Tick 2 (5%) do 2 ticks and 5% extra together.
Tick 3 (15%) and Tick 4 (20%) do 2 ticks and 35% extra together. That's 30% more than first pair.
Tick 5 (30%) and Tick 6 (35%) do 65% extra together. That's 30% more than previous pair.
etc.

Since damage actually stacks much faster than your model assumes, dps is much more important.

The other problem i have with the post is that it seems to either skip over or hide a lot of required calculations.

E.g. how much healer hps is taken into account? How is their mana expected to behave as damage ramps up? How much damage do dpsers actually need to deal for Beth to hit the bucket?

That's specifics, though. I doubt they will change the picture to much. All in all, i'm not surprised that normal modes favour healer stacking. Once you have more dps than enough to beat the encounter you are better off replacing additional dps with hps for progression fights. This is something that has been known since vanilla.

And since Fireland normals seem to be tuned in such a way that ~353 ilvl is enough, this results in a lot of room for dps > hps substitution.

But i disagree about proper dps being irrelevant. On a fight like Shannox normal, maybe, but not on others.

Leeho said...

I actually don't assume that difference between an excellent and a good healer lies in aoe hps, maybe i wasn't accurate enough in my first comment.

I would say that the difference is in ability to save a person from dieing in a tight situation. Sometimes tanks make mistakes with taunting rotation or cds and take too much damage in a short time. Or, say, on Ragnaros raid gets fire circle at the same time with adds spawn from fire eggs, so some dpsers get hit while raid is running out. Excellent healer may have even less hps than good one, but he is able to choose the right target and right spell so fast that he will save person in a situation where less experienced healer could make a mistake or spend one second more, and so this person would die. These saves are the thing that will turn a sequence of wipes into a clean kill in two tries when experienced healer replaces less experienced one.

But you can't calculate this moments as easily as you can calculate flat hps.

High dps is somewhat like this kind of healing, you need to be really fast on reacting to things happening around and at the same time you need to choose proper ability and proper target every second or two.

Inquisitor said...

Don't they tune the DPS requirement on many hardmodes with a hard enrage, and one tight enough that you need the DPS of 5 bleeding-edge DPS to meet it?

Because there's no question that people can and will stack raids for world first kills, and given some of the crazy stuff they do, merely stacking healers would be pretty tame, and it's silly to assume they've not thought of it.

Anonymous said...

I agree with Azzur, you didn't consider mana, even with 5 healers, which would make Rhyolith's 80k incoming raid damage healable, the healers would go OOM before the boss died frequently.

You could apply the calculations if you expand them further and include some time to OOM for healers, which depends on the amount they are healing, and also their level of gear.

Then you have a realistic simulation which considers all factors, if boss dies too slow healers may be OOM, but that is counterbalanced by the fact that if you add more healers they can use less mana-intensive healing.

Also you have a simplistic assumption "The raid will wipe after 1400/(B-H*h) seconds". This is not true as there is the option to let a DPS die at any time by stopping healing him, which will let the healers available mana and GCDs hold the rest of the group up for longer. So while the 10-man group with 2 tank, 3 healer, 5 dps may be 'dying', once the poor enhancement shaman is sacrificed the group may be okay to survive.

'Top' guilds are perfectly good at theorycrafting and if healer stacking was optimal they would do so. There are surely some fights its a good idea for, I even remember 9 healing IC in TBC times, but it is not a blanket solution.

Adnade said...

Whilst you theory might hold relevance on normal modes, it's absolutely inaccurate for heroic ones.
For example:

Rhyolith - You need to be able to force the transition at around 5 minutes mark, when he starts stacking the Superheated buff. If you don't have enough dps to make it before the "soft enrage", p2 will wipe you, regardless of how many healers you bring.

Beth'tilac - You need to bring her to 75-80% in p1, if you want to kill her before her p2 stacks/damage become unhealable, which is no trivial task, considering the amount of dps you need downstairs to keep up with the adds. No amount of healers will keep you up when the aoe pulse 1 shots the raid.

Baleroc - You cannot beat him without 400k~ dps, end of story.

You normally have good insight in regards to raiding Gevlon, but when it comes to this, you need a bit more knowledge and experience before giving an informed opinion.

Frostys said...

I don't se anywhere how long the healer can actually hold on with the AoE heal spam required in such situation and it's a big point. Unless you have enough healer to go from AoE heal to single target heal, your efficiency won't raise all that much so in the end, you burn through mana like crazy.

THe reason top guild don't bring more healer to fight like this is the mana wall at the end of the tunnel. Unless the damage is low enough for them to have regen phase while other are in healing mode, they will go OOM sooner or later. Thats where good dps make a difference.

Alrenous said...

@ Azzur

Your argument holds only if the raid holds to the 2:3:5 ratio. Gevlon's point is that you can make it easier on healers and DD alike by going 2:4:4 or even 2:5:3.

I include both the OOM and DD-test issues you raise.

@ Frostnys, Adnade,

Change the total raid hp calculation ( 1400/(B-H*h) ) to total raid mana (in health equivalent) if you'll run out of mana before health. Don't forget to try running low on health AND mana, to get better efficiency.

Each healer still increases the time to live, and therefore increases the amount of mana regenerated of themselves and all other healers.

Increasing DD actually decreases the total amount of damage each other DD will do.

Again, heals and DDs scale in opposite directions. Balance issues force the heals to be past the crossing point at all times.

@ Kahalm

Healing in WoLK was different. Infinite mana meant that as long as the healers had time to cast on everyone who needed it, any additional healing was just a waste of a raid slot. Now, each extra healer, among other things, lets all other healers use more efficient heal casts.

Arrowkeys said...

I've never heard the theory that soft enrages were a DPS test instead of a healer test in the first place.

Soft enrages are a healer test. Add phases like P2 Cho'gall are the DPS test.

Sean said...

@Alrenous
Nope, I don't buy it that guilds use the classic 2:3:5 because of social reasons. Take Halfus HM, for example - once people figured out the 3-tank strategy, there were even PUGs that were using that method and successfully doing the kill.

Rather, the classic 2:3:5 is used because it is usually the optimal raid configuration. There is no doubt that if a guild figures out that healer stacking is optimal, they would use that strat and many others will follow suit.

Gevlon makes the point with charts and numbers, but does not stand up to practical testing (i.e. healer mana will run out if there's insufficent DPS). I'll be amazed if a guild found that healer stacking will actually help for those bosses.

I stand by my original statement - soft enrages is a superior mechanic because it allows the whole raid to contribute towards the result. This is in contrast with a hard enrage where the kill is purely in the domain of the DPS.

chewy said...

I don't often (if ever) come to Gevlon's defense but in respect of those discussing being out of mana I think you've missed the point.

Gevlon isn't doing absolute theory crafting here he's using some assumptions to illustrate the effect of changing the ratio between healers and dps. The first assumption is that the healers won't run out of mana, if they do then it doesn't prove anything, so he implicitly starts with "let's assume they don't" which is reasonable with respect to the way he's set the other parameters in his point.

Anonymous said...

Azzur:

Soft enrage also means the DPS can claim it was the healers who needed to do more.

Lyxi said...

@chewy

And the calculations hold true for spherical cows in vacuum.

This is not helping. These are not just minor concerns that can be ignored and brushed aside. There are argument-breaking points and can't just ignore for the sake of the argument.

because we then end up with an academic discussion with no immediate practicality.

Ihodael of Darnassus said...

"Also I don't think that no one in top guilds haven't already thought about using more healers. The thing is, that they can get by with only few healers, because they are players of exceptional skill. And less healers means more DPS, and faster kills. And you need that speed when going for first kills." (Ayros)

I feel this is the essence of the thing: for top guilds, they minimize healer and tanks and maximize dps since they want kills as fast as possible.

The interesting point is that everyone then follows their guidance without further thought. Like you said earlier about Cata HC (now no one remember but these used to be the "omg... they are so hard dungeons): just take one healer + off-healer and most bosses were easy mode (longer but easy mode).

One of the interesting points I see in The PuG raids is that you adapt strategies to group composition with success. Unfortunately most guilds never try it... they just insist in doing it the "PARAGON way".

chewy said...

@Lyxi

I do appreciate the point you're making but this isn't a cow in a vacuum calculation even if we don't consider the potential for the healers to run out of mana.

Gevlon makes the point that it is more effective to take an extra healer over an extra dps even if the healer is mediocre. The mana depletion point doesn't invalidate this argument since adding an extra healer increases the overall mana pool. Which is why I think it's reasonable to ignore the mana with respect to his point.

Having said this, if someone were willing to do the calculations on the mana I'd love to see them. Neither of us actually know (or have demonstrated) that it makes a difference one way or the other - it's merely opinion.

Squishalot said...

@ chewy: I think you picked the wrong point to come to Gevlon's defence on.

The mana is important, because of one key variable changing - time. Adding a healer is only more efficient if you assume that healers can last indefinitely. The damage being done is a function of time, and raid survival time is actually MIN(1400/(B-H*h),HM/mana usage) seconds. Essentially, healer mana pools provide a cap on the total time that you can take.

This, moreso than stacking damage, is the biggest reason why the optimum number of healers is likely to be 3-4 depending on precise circumstances, rather than 4-5.

Arguably, if you have enough healers, you can change healing rotations and increase HPM output, but if we are assuming a raid-wide damage scenario with AoE healing, and we assume that each healer can heal H HPS at a fixed mana usage, then each additional healer brought on to extend raid survival time will also be burning that fixed mana usage to extend the survival time.

Once you put a constraint on healer mana, you now have a hard-enrage style target for DPS to reach, putting the onus squarely back on the damage dealers.

I don't think the analysis is bad, it's just unrealistic and can be significantly improved by considering healer mana. I think that once it's taken into account, the 3 healer paradigm is likely to come out on top.

Anonymous said...

Interesting analysis which could form basis for theory, but it fails on various factors not taken into account: overhealing being one of them.

Another factor is the lack of disc priest and any class which puts a AoE absorption on targets healed (hpala does, rsham does, and especially disco priest does). These will prevent the AoE dmg of the soft enrage one shotting people when they are 100% healed. AoE healing + stacking, or stacking 2 resto druids helps as well, until you get to the part where people get 1-shot by AoE from soft enrage.

I think the point however is: 3 healers on 10m is not defacto the best solution. This coincides with my experience in The PuG, as well as in 25m raiding. Sometimes 4 healers was enough in 25m, sometimes 7 was not.

Chimaeron is a DPS test in P2 (plus kiting, and a CD test right before it so people are top healed and healers and tanks have highest aggro). The less healers you have, the better you will do.

Occu'thar is a DPS test, too. If DPS fail to kill the eyes, you will wipe. The healing is manageable even for 2 healers. There is not much dmg going on. If you 2-heal this fight, you have an extra DPS on the adds.

Alrenous said...

@Azzur,

Yes, soft enrage is the superior mechanic.

However, the current incarnation's balance is still optimized with stacked healers.

I don't HC raid so I can't say for sure what kind of forces might keep their healer population down. I can think of several, though. The most credible is that they rely on skill far more than gear, which keeps healers nearer the below-DD scaling part of the curve.

This doesn't change the fact that, as Gevlon has proved both by math and by doing, non-HC raids benefit by stacking healers. They don't simply because HC raiders don't.

2nd example: Perhaps HC raiders don't actually ever wipe from healer issues, but only from various encounter mechanics. If so, they'd always use the absolute minimum healer complement.

My point being that the fact that HC raiders don't stack healers is necessary but far from sufficient as evidence.

Conversely, Gevlon's d(healing)/d(gear) and d(healing)/d(population) curves speak for themselves.

Trooze said...

Gevlon's model assumes the boss does constant aoe damage, which is not the case. All of the bosses with a burn phase have a stacking buff to boss. This increases the boss damage linearly, which means the total healing requirement is quadratic over time. Thus the soft enrage timer is proportional to the square root of the number of healers. Adding healers just doesn't scale as well as adding dd players, so you can't fix low dd with extra healers. The total raid damage would be proportional to {sqrt (num_healers) * num_dd}. The optimal number of healers for a 10-man raid would 3 in this case.

Frostys said...

There is no real "efficient" way to AoE heal and thats where the assumption of more healer make you heal more efficiently gets killed in AoE situation. Beside raid CD, most AoE heals will cap out at 5 target. Ground effect heal will lose power per tick if there are too many target.

Unless a single AoE heal can give you more enough life abck for 2 tick of damage, you will end up continually healing no matter what and at some point, burn through your mana. When you get the message "I need more mana" when you try to cast a heal, it mean the DPS were too slow.

A goo0d example of soft enrage proving good healer can nullify the need for better dps is grobulus abck in Naxx. Thsi damage didn't really go up but more and more plague were distributed to player wich also created more and more "bad" zone on the ground. No matter how good the healer and how many you have, if you kill him slowly, you get more and more trouble.

Adding more healer is not a definite solution to all soft enrage problem. It definately help on some but the fight has to be a "healer fight" or at the very least have a "healer phase". Dreamwalker is an example of such fight.

Sean said...

A lot of the discussion has become side-tracked. Gevlon's main argument is that the soft-enrage is there to make the DPS feel better about themselves when in reality a soft-enrage is simply a superior mechanic. As for making the DPS feel good, there is always the DPS meter to prove whether someone is performing or not.

As for the optimal raid configuration of 2:3:5 - this is usually the best setup for "most" bosses. There will be encounters where healer stacking is useful, whilst DPS stacking (e.g. Atra/Chogall) is beneficial in others. In Halfus HM, tank stacking (3:3:4) was the "best" strat.

For healer stacking my guild was at Majordomo (10-man) yesterday and we first attempted it with 3 healers. We tried 4 but still didn't get the kill but I felt the 4-heals strat was doable.

To recap:
- Soft-enrage is a superior mechanic because it allows the whole raid to influence the result.
- Illusions of feeling good are easily dispelled by the DPS meter.
- The optimal raid configuration is usually 2:3:5 but in reality, it depends on the bosses.

Anonymous said...

"Gevlon's model assumes the boss does constant aoe damage, which is not the case. All of the bosses with a burn phase have a stacking buff to boss."

No, some put a debuff on player similar to Halls of Origination gauntlet boss. A rogue or mage can cloak/ice block that debuff off themselves. If it is AoE damage, I wonder if stacking Sub rogue would help more than stacking 1 extra healer. After all, the Sub rogue does good damage and need far less healing to burn phase AoE.

As for the person who commented on LK heroic first kill using less healers: Blizzard was aware of the RNG of the Valks on that fight. If the Valks picked up a healer, healing would get tough. If they picked up a DPS, DPS would get tough. You need the DPS on the Valks, so instead of an extra healer they simply wipe due to RNG when healer is picked up yet when DPS is picked up it goes as planned. Of course later with the Wrynn buff and better gear (+ Halion released, + patch 4.0.1 buffs) it all became easier.