In-Depth Mutilate Cycle Analysis - Page 17

Akraat · 12/28/09, 11:25 AM

Indeed this is true. None of the aforementioned classes capable of causing bleeds actually cause a certain bleed as soon as the fight begins (unless you are raiding with an MS warrior, though I see very little around).

What I want to touch on is whether garrote is worth the energy when vanishing. As some of you may know, it is entirely possible to hit vanish, but also hit your envenom keybind, resulting in the acquisition of the overkill buff, however you also use envenom.

Envenom costs 35 energy, garrote costs 50 energy. I use my vanish straight after my first overkill, and 4+ combo points are on the target for an envenom.

I'd say that envenom vanishing is an increase of dps over garroting after vanishing, not just because envenom would do more damage, but the delay that you enter stealth for garrote to become usable is quite large, compared to hitting envenom and vanish.

Edit: You must include a /startattack macro in your envenom, as envenom vanishing turns off your attack.

Killme888 · 12/28/09, 2:42 PM

Chances of a fury warrior not critting in the opening autos/ww is pretty low, especially if you run with more than one. And even then, you can simply put up SnD first instead of HfB.

Garrote damage is way too weak without Blood Spatter & Opportunity, you shouldn't be using it at all unless you're in a 10 man where your raid comp has a hard time putting up a bleed at the start.

Akraat · 12/28/09, 2:51 PM

Originally Posted by Killme888

Chances of a fury warrior not critting in the opening autos/ww is pretty low, especially if you run with more than one. And even then, you can simply put up SnD first instead of HfB.

Garrote damage is way too weak without Blood Spatter & Opportunity, you shouldn't be using it at all unless you're in a 10 man where your raid comp has a hard time putting up a bleed at the start.

Yes of course, however bear in mind in situations where you can start the boss from behind, it is certainly no waste to garrote for an opener to ensure an immediate HfB.

Also an optimal start-up for rogues is using as little CPs as possible for SnD - I normally go with garrote, HfB, use the CP on SnD, and follow my normal 4n+ rotation (even if you have to muti twice you should have enough time to pull off an envenom).

Killme888 · 12/28/09, 3:49 PM

So you'd rather use an attack that does less DPE, just so you can feel better because you only spent 1 cp on SnD?

It doesn't matter how many cp you spent on the first SnD if they all came from one attack, you're not losing anything. In fact, if you get horrendous luck and deadly doesn't apply for a few swings, then your envenom gets dodged once or twice letting your 1cp SnD fall off. That hasn't happened to me in 3.3 yet but it certainly has before this patch where I was still using garrote opener and capping expertise wasn't the best for dps.

Sculduggery · 12/28/09, 4:08 PM

Originally Posted by Akraat

Yes of course, however bear in mind in situations where you can start the boss from behind, it is certainly no waste to garrote for an opener to ensure an immediate HfB.

Also an optimal start-up for rogues is using as little CPs as possible for SnD - I normally go with garrote, HfB, use the CP on SnD, and follow my normal 4n+ rotation (even if you have to muti twice you should have enough time to pull off an envenom).

The most optimal start-up for rogues, whether mutilate or combat, is going to be to turn auto attack on from range, even before the boss is engaged (where it is an option) and then automatically start attacking as soon as you're in range whether from stealth (Mut) or not (Combat). Without Opportunity or Blood Splatter, it's simply not feasible to ever use Garrote. Since we are referring to a most optimal situation, the raid will include a warrior, and the boss will be bleeding as soon as combat starts, sometimes even before you're in range.

The follow-up from there is entirely preference, but I'm starting with Mutilate (since it will proc Deadly Poison most of the time) then SND and once HFB is up entering Envenom rotations. After that, I use Vanish right before ToTT comes back up to coincide with the buff and blow trinkets. If we've got Bloodlust at this point, Speed Pot too on fights where you can't plan it against the actual start time.

Waiting to use any move, whether it's Garrote, Shiv, or Mutilate instead of opening with auto-attacks as soon as you're in range is going to result in a loss to overall damage done. There is just no getting around the human error/reaction time unless you start the fight in melee range (Gormokk, Jaraxxis, Saurfang) and are able to spam your opener of choice as the flavour text timer reaches zero.

Mavanas · 12/28/09, 5:59 PM

I would probably agree with you on the human factor and that by opening with a garrote you are going to waste some autoattack uptime. However you need to take into account that you are also losing HfB uptime, which probably needs to be quantified and compared to the autoattack damage you gain.

Out of the attacks that are going to be affected by HfB at the start, we have deadly poison, instant poison and autoattacks. DP will not be fully stacked and IP will not be proccing from DP refreshes. Suppose it takes 4 seconds of the fight to proc HfB. During that period I would do 3 main hand attacks and 4 offhand attacks (for haste I am using my stats: t9 +mix of 245 and 258 gear, above average, but not top of the line by any means). For reasons of comparison with garrote case, where you won't have HfB for the first autoattack anyway, I am only going to consider 2 mainhand attacks and 3 offhand attacks. In the first 4 seconds, excluding 1 mainhand and 1 offhand autoattack, I expect to do 2099 MH damage, 2367 OH damage, 778 IP damage, and 1 tick of DP (assumed 2 stacks) of 2640. If I had HfB up (by applying garrote first), I would gain 1420 damage. In the garrote case, I also apply a garrote without HfB, and in autoattack case, I would apply one mutilate without HfB buff, so the marginal effect of no HfB on those two abilities is about cancelled.

So the lack of HfB is worth about 1420 damage, which is just shy of 1 second autoattack dps at the start without any buffs. If you think you lose about 0.9 seconds of dps due to human factor by not being able to apply garrote instantly, then yes autoattacking is probably the way to go. Otherwise, I'd probably go with garrote - HfB - mutilate - SND, etc.

xcyteZ · 12/28/09, 9:31 PM

Originally Posted by Mavanas

I would probably agree with you on the human factor and that by opening with a garrote you are going to waste some autoattack uptime. However you need to take into account that you are also losing HfB uptime, which probably needs to be quantified and compared to the autoattack damage you gain.

Out of the attacks that are going to be affected by HfB at the start, we have deadly poison, instant poison and autoattacks. DP will not be fully stacked and IP will not be proccing from DP refreshes. Suppose it takes 4 seconds of the fight to proc HfB. During that period I would do 3 main hand attacks and 4 offhand attacks (for haste I am using my stats: t9 +mix of 245 and 258 gear, above average, but not top of the line by any means). For reasons of comparison with garrote case, where you won't have HfB for the first autoattack anyway, I am only going to consider 2 mainhand attacks and 3 offhand attacks. In the first 4 seconds, excluding 1 mainhand and 1 offhand autoattack, I expect to do 2099 MH damage, 2367 OH damage, 778 IP damage, and 1 tick of DP (assumed 2 stacks) of 2640. If I had HfB up (by applying garrote first), I would gain 1420 damage. In the garrote case, I also apply a garrote without HfB, and in autoattack case, I would apply one mutilate without HfB buff, so the marginal effect of no HfB on those two abilities is about cancelled.

So the lack of HfB is worth about 1420 damage, which is just shy of 1 second autoattack dps at the start without any buffs. If you think you lose about 0.9 seconds of dps due to human factor by not being able to apply garrote instantly, then yes autoattacking is probably the way to go. Otherwise, I'd probably go with garrote - HfB - mutilate - SND, etc.

There is no doubt in my mind that opening with garrote in the very beginning of the fight is most likely the most benefitial way to start your rotation. I only mentioned the use of shiv to start your rotation to be used depending on a very specific situation. There are only going to be a few times when a bleed will be up by the time you are "ready" to open.

Also going back to my original question of 60+ pooling vs. 100% Envenom buff uptime, could someone answer this for me?

Mavanas · 12/28/09, 10:21 PM

Simsheet models the pooling. It also models techniques you can use in real combat to increase your envenom uptime, such as lowering the number of combo points at which you use envenom. You can lower it in general by switching to a 3+ cycle, or you can do it selectively by doing 3+ if envenom buff is down and 4+ otherwise. If you need help understanding the DPS Strat settings, pm me on the forums.

From what I have seen in my own tests, overriding the energy pooling rule if envenom buff is about to drop is a strong dps increase. However overriding the combo point rule (such as using envenom at 3+ if envenom buff is not up) is a slight dps decrease. Thus the most optimal dps rule for mutilate is a standard 4+ cycle, pooling to 60 energy if envenom buff is up, but ignoring that rule if envenom buff is down. Only envenom at 0.2 seconds remaining on envenom buff if you can time it that well, as long as you have the 4+ combo points for it.

PDInfluence · 12/29/09, 5:29 AM

Thank you for the replies:
But I am still curious as to what you do (Manava) when you Vanish for Overkill: do you Mutilate w/ Auto-attack out, or do you Garrote? The reason I am so keen on Garrote is because throughout all my progression and learning to play maximizing my dps, bleeds have been minimal to non-existent. Garrote after Vanish when HfB is at >5 secs with 3-4 combo points seemed ideal, but now I know it is not. I lose 40 seconds on my Vanish up-time, GCD on my auto-attacks waiting on Garrote to be useable, and sometimes, the time it takes to position myself since sometimes the boss is moved slightly, thus resulting me having to move just that much.

Also, I'm assuming a third glyph (discussed in other threads/posts many times, I'm aware) is pretty optional. I've been using Garrote just because I use it so often, but to open with Garrote and never use it again would mean even Vigor would be a better glyph, and FoK/ToTT glyphs don't appeal to me whatsoever. FoK is rarely used in ICC, and ToTT is rDPS rather than personal dps (sorry to sound selfish). During Marrowgar, Deathwhisper, and Gunship, my energy caps out often, thus 1 point into Vigor (not apart of the discussion and not apart of my spec) w/ Glyph of Vigor would be 10 (wo/ talent) or 20 (w/ talent) more energy each Bladestorm, new set of Adds, and Mage deep freeze ship hopping.

Akraat · 12/29/09, 9:14 AM

Originally Posted by Sculduggery

The most optimal start-up for rogues, whether mutilate or combat, is going to be to turn auto attack on from range, even before the boss is engaged (where it is an option) and then automatically start attacking as soon as you're in range whether from stealth (Mut) or not (Combat). Without Opportunity or Blood Splatter, it's simply not feasible to ever use Garrote. Since we are referring to a most optimal situation, the raid will include a warrior, and the boss will be bleeding as soon as combat starts, sometimes even before you're in range.

The follow-up from there is entirely preference, but I'm starting with Mutilate (since it will proc Deadly Poison most of the time) then SND and once HFB is up entering Envenom rotations. After that, I use Vanish right before ToTT comes back up to coincide with the buff and blow trinkets. If we've got Bloodlust at this point, Speed Pot too on fights where you can't plan it against the actual start time.

Waiting to use any move, whether it's Garrote, Shiv, or Mutilate instead of opening with auto-attacks as soon as you're in range is going to result in a loss to overall damage done. There is just no getting around the human error/reaction time unless you start the fight in melee range (Gormokk, Jaraxxis, Saurfang) and are able to spam your opener of choice as the flavour text timer reaches zero.

Couple flaws here. Firstly, if the warrior gets in range before you, assuming the boss has the same targetability functions as Gormok, you're doing something wrong. You must always begin combat immediately/as soon as possbile. Second, although it's already been mentioned, if you do start immediately, a garrote would be the best opener for HfB uptime which results in an overall dps increase. Nothing is stopping you from spamming Garrote as soon as you are in range / as soon as the boss is attackable. Even if there was a delay between garrote and auto attacks, it would be so minimal that the uptime from HfB would definitely outweigh it.

Killme888 · 12/29/09, 1:07 PM

Originally Posted by Mavanas

I would probably agree with you on the human factor and that by opening with a garrote you are going to waste some autoattack uptime. However you need to take into account that you are also losing HfB uptime, which probably needs to be quantified and compared to the autoattack damage you gain.

Out of the attacks that are going to be affected by HfB at the start, we have deadly poison, instant poison and autoattacks. DP will not be fully stacked and IP will not be proccing from DP refreshes. Suppose it takes 4 seconds of the fight to proc HfB. During that period I would do 3 main hand attacks and 4 offhand attacks (for haste I am using my stats: t9 +mix of 245 and 258 gear, above average, but not top of the line by any means). For reasons of comparison with garrote case, where you won't have HfB for the first autoattack anyway, I am only going to consider 2 mainhand attacks and 3 offhand attacks. In the first 4 seconds, excluding 1 mainhand and 1 offhand autoattack, I expect to do 2099 MH damage, 2367 OH damage, 778 IP damage, and 1 tick of DP (assumed 2 stacks) of 2640. If I had HfB up (by applying garrote first), I would gain 1420 damage. In the garrote case, I also apply a garrote without HfB, and in autoattack case, I would apply one mutilate without HfB buff, so the marginal effect of no HfB on those two abilities is about cancelled.

So the lack of HfB is worth about 1420 damage, which is just shy of 1 second autoattack dps at the start without any buffs. If you think you lose about 0.9 seconds of dps due to human factor by not being able to apply garrote instantly, then yes autoattacking is probably the way to go. Otherwise, I'd probably go with garrote - HfB - mutilate - SND, etc.

Couple points you're missing:

1. Opening with mut could proc deadly poison, therefore lowering the initial ramp up time.

2. You can start with SnD instead of HfB, at this point with our damage spread I'm pretty sure its on par or at least very close to starting with HfB, also lowers the ramp up time as well.

3. You get more cp towards the initial envenom.

I've always been doing Mut -> SnD -> HfB -> Envneom, which has always worked out better for me than the starting rotation you've described.

Akraat · 12/29/09, 2:11 PM

How long does it take before you can apply HfB? Rough estimate in seconds would suffice.

This is because I'd like to compare what you said to Garrote>HfB>SnD.

With the above startup you pretty much get both buffs up immediately, although with muti>SnD you must wait for a bleed from another raid member. Also bear in mind that warriors must wait a fair bit before dpsing, since their abilties cause quite a bit of burst aggro, and hunters and druids don't start off with bleeds either.

Sculduggery · 12/29/09, 2:59 PM

Originally Posted by Akraat

if the warrior gets in range before you .. you're doing something wrong.

Flaw here. Warriors charge to get to the target before rogues. The Gormokk exception was mentioned in the section you quoted.

Originally Posted by Akraat

Nothing is stopping you from spamming Garrote as soon as you are in range / as soon as the boss is attackable. Even if there was a delay between garrote and auto attacks, it would be so minimal that the uptime from HfB would definitely outweigh it.

Except that not all bosses start in melee range or facing away from you. Having auto attack up already and starting with mutilate would undeniably be a damage increase in those situations. Current examples would be Marrowgar, Deathwhisper, Faction Champions, Twin Valkyrs, and Anub'Arak. Also, this:

Originally Posted by Mavanas

If you think you lose about 0.9 seconds of dps due to human factor by not being able to apply garrote instantly, then yes autoattacking is probably the way to go.

I realise that this is based on napkin math, but it does support going straight into the attack cycle and not waiting for positioning to open with Garrote.

Originally Posted by Akraat

with muti>SnD you must wait for a bleed from another raid member. Also bear in mind that warriors must wait a fair bit before dpsing, since their abilties cause quite a bit of burst aggro, and hunters and druids don't start off with bleeds either.

In my raids, the wait time to apply HFB is usually less than two GCDs. In other words, in the time it takes to Mutilate and then SND, someone else already has a bleed on the target and next you can cast HFB. Warriors generate bleeds from auto attacks via Deep Wounds, so the alleged aggro of their abilities is irrelevant. Plus, generally in raids tanks start out with a pretty huge aggro lead thanks to Tricks of the Trade and Misdirection from multiple sources. Also, why wouldn't a druid be able to Mangle in that time frame? Wouldn't it be in his best interest to get his DoTs up as soon as possible?

There are just too many other sources of bleeds in a raid to waste damage time on the positioning and stealth requirements from Garrote. The fact that it's also missing the bonuses from Opportunity, Blood Spatter, Glyph of Garrote, HFB, or anything else that increases it's damage/duration really makes it lackluster.

As Killme888 and I have both mentioned, Mutilate is also going to give another chance to apply offhand poison. Garrote does not.

Jibini · 12/29/09, 3:52 PM

Originally Posted by Sculduggery

Also, why wouldn't a druid be able to Mangle in that time frame? Wouldn't it be in his best interest to get his DoTs up as soon as possible?

Mangle is not a bleed, although it does buff bleed damage. I've read people asking if it works as a bleed for the purposes of HfB, but I don't believe I've ever actually seen anyone answer in the affirmative to it.

That said, if you have both a feral tank and dps, the tank is liable to use Mangle almost immediately. The dps should then be using Rake as their initial attack, since it's their highest damage per energy CP generator, and that is a bleed.

ShadowEric · 12/29/09, 4:44 PM

Originally Posted by Jibini

Mangle is not a bleed, although it does buff bleed damage. I've read people asking if it works as a bleed for the purposes of HfB, but I don't believe I've ever actually seen anyone answer in the affirmative to it.

I mentioned about 15 posts ago that pre-3.3, Mangle was enough to allow the use of HfB. A confirmation of that is needed post-3.3, as I personally haven't specced Mutilate since then.