Combat Guide for Cata [12/01/2011] - Page 14

thirus · 01/24/11, 12:09 AM

Originally Posted by Aldriana

While that specific source might be dated, I do think its fair to say that the observed gap between Combat and Mutilate appears to be a bit larger than the 2-3% predicted. Picking out the most single-targetish, least interrupted fights in current content, and looking at the #20 DPS on WoL (i.e., high end but not some fluke run):

Your point may be right that Combat and Mutilate divide more then 3% from each other, but you have to consider two things.

1. Combat does scale pretty well so comparing 346ilvl Combat and 346ilvl Mutilate the gap is around 6%. In BiS however the gap tightens to ~3%.(numbers taken off shadowcraft-app)

2. As I said in my last post, the frequency of well geared Mutilate rogues is way above the frequency of well geared Combat rogues. The #20 Mutilate rogues you stated have an average ilvl of 360 while the #20 Combat rogues do have an average ilvl of 355.(this example isn't 100% accurate, I just looked up those #20 guys in the armory and that was the kind of difference you can expect)

Also you didn't wrote down the 25man #20 rank, but the 10man #20 - I guess this wasn't intended.
Anyways considering the above mentioned points we get closer to the expected deviations. 4-5% make up for your talent build(if Mutilate or Combat), 4-6% for the gear difference and probably 1-3% skillwise.

If we would really consider that Mutilate is 12-17% ahead of Combat, you or someone else would probably have had to rewrite the current model of Combat in the shadowcraft-application or fix some major bugs that inflate the estimated dps. But I guess this isn't the case.

Danagat · 01/24/11, 2:21 AM

I think the fact that there sometimes is a "gogo nuke it!"-Phase (Cho'gall, Ascendant Council) is one point that pushes assassination DPS in praxis.

CD's like Tol'vir Potion/Heroism are used when reaching this phase, normally <35% when assassination has an advantage over combat.

In addition, the optimal usage of killing spree/adrenaline rush is a huge factor when modeling combat dps, as far as i know they have a much bigger impact on dps than vendetta/cold blood. In praxis you will sometimes not use ks/ar when they are ready because the point when you really *need* dps to kill the boss is not reached.
it is often not *optimal* to play an "optimal" combat cycle.

mulvad · 01/24/11, 4:19 AM

Originally Posted by Aldriana

So, computer repaired, I ran some numbers and got, quite literally, the exact same answers. Which probably implies that I have the same set of changes, which are... close to right. I haven't accounted for the change in weapon enchant proc uptimes due to having more MH attacks and fewer OH attacks, but overall the DPS effect of that should be minor.

Nice, so I didn't fuck up. I do believe that I accounted for the increased MH attacks, at least in regards to poison procs. We can compare our code changes if you like to see if we interpreted the patch notes the same way.

I've been running raids as a combat rogue for some time now, and even with perfect rotation I fail to even come close to the dps of our assassination rogue (with comparable gear). My numbers was very close to what you find at State of DPS or similar sites.

This made me think. Why is the modelled combat dps so much higher than actual numbers? I came up with pretty much the same stuff that has been discussed here.

My gear was not "heroic t11 BiS" like in the model. I figured combat might scale better with gear and there for be weaker in my ilvl 355 gear.
The model is based on constantly attacking the boss which pretty much never happens (exceptions exist). The penalty for a combat rogue to move out of combat seems higher to me? Based on the fact that you more or less instantly energy cap and that auto attacks generates a higher portion of our overall dmg (both through white attacks and through combat potency). An assassination rogue can pool his energy while away from the boss, and provided that the deadly poison stack doesn't drop I think that the dps loss will be fairly minimal compared to a combat rogue. (This is based on pure speculation, I haven't theorycrafted on this).

This made me think that perhaps reforging and gemming to haste is only the best option in theory and not in praxis.

But I think we should consider the very real posibility that there simply is a bug in the combat model that makes it over-estimate the dps. Because all the things we are talking about here should not be enough to offset the damage by the 20% we are seeing from the real data.

thirus · 01/24/11, 8:59 AM

Originally Posted by mulvad

This made me think. Why is the modelled combat dps so much higher than actual numbers? I came up with pretty much the same stuff that has been discussed here.

Well I don't know why you aren't able to reproduce the numbers shadowcraft puts up so I will post two examples comparing estimated dps and actually performed dps.

1. estimated shadowcraft dps: ~20300
actually performed dps: ~20800
log: World of Logs - Real Time Raid Analysis

Fairly long fight duration(6,5 minutes with prepot)and no high deviations on the expected critrate - 1% parry, 1% block on whitehits and 2% parry on SS.

2. estimated shadowcraft dps: ~21100
actually performed dps: ~23400
log: World of Logs - Real Time Raid Analysis

Fairly short fight(3 minutes without prepot)and again no high deviations on the expected critrate - 3% parry, 2% block on whitehits and 2% parry on SS.

Regarding your second point mulvad you are somehow right. Any second a Combat rogue perfroms his cycle on a mob he constantly lowers the cooldown of his abilties and therefore a Combat rogue does lose more dps than a Mutilate rogue if both aren't able to dps a mob.

Gorathra · 01/24/11, 9:33 AM

Originally Posted by Aldriana

See, I think the straight-across "top DPS" numbers are probably pretty skewed at this point. Because Assassination is better, more people play it; because the boss that drops MH Daggers is much easier (particularly on Hard Mode) than the bosses that drop MH Swords, Assassination has an advantage there; and so on. Thus, while at a given gear and skill level, the gap might only be (say) 5%, in practice gear and skill aren't usually equal when comparing top performance on fights, so the numbers reflect a larger difference.

I think that Aldriana says it here, the best rogues are playing Assassination spec so they inevitably will produce higher DPS number than the 20/40/200 "best" (aka those rogues that refuse to play assassination) combat rogues. Now I am not saying that all of the top combat rogues are bad or anything I think its tough to get a good judgement on the actual DPS when a majority of the top tier rogues are skewed to one spec over the other.

I did Magmaw 25-man this past Tuesday as combat and pulled in a respectable DPS of 19537 World of Logs - Real Time Raid Analysis

That put me almost with the top 200 for the fight and I was only in ilvl 347 gear at the time (did not have the fluid death trinket) where as inspecting the players in the lower half of the top 200 for the fight for assassination you see almost all ilvl 359 geared players.

So I would think that the model is likely correct, it probably just cant take into account the actual skill levels of the players putting up the top 20/40/200 numbers and it assumes BiS gear (which a lot of the combat rogues in the logs don't have).

So in my opinion I think the difference on numbers is based in part on skill level and the gear differential between the two specs which is biasing the numbers that we are actually seeing on top 20/40/200 logs

Stony · 01/24/11, 11:33 AM

Originally Posted by mulvad

Why is the modelled combat dps so much higher than actual numbers?

As a combat rogue with top 10 parses on most of the bosses this tier, I would say that encounter design more than anything else affect combat rogues to a much larger degree than assassination, with the main culprit being AOE and Bandit's Guile.

Looking at the bosses fight by fight with 4.06 changes:
HMagmaw: redirect not working on exposed head really hurts combat more than assa, but with cds every expose phase combat should be competitive
HOmnitron: with RB affecting redirect and permanent BF, combat is going to be very strong here next patch
HMaloriak: I respec Mutilate here for AOE damage
HChimeron: waiting for caustic slimes to finish casting in order not to risk death delays KS quite a bit here, Caustic slime RNG also can energy cap you on AR
Atremedes: BG dropping every air phase hurts combat much more
Nef: Combat should be competitive

HHalfus: We tank 2-3 drakes on top of Halfus so while BF can pad damage here, it is only useful if you can BF the right drake to kill
HValiona: respec Subtlety
HCouncil: permanent BF will be very good on this fight p1+2, p3 assassination will edge out
Chogall: Combat cannot swap to adds at all unless you want to lose significant damage, assa can also pad meters with aoeing bloods

HWinds: I stay on wind the entire time, with imp recup and soaking the hurricane with 1 priest, AR should be up for every shield, combat is useful on this fight
Alakir: KS is pretty much unusable this fight

Overall with the 4.06 change I feel combat will be very competitive for a majority of the fights, but there will be a few problematic fights that you still should respec for.

Lokar · 01/24/11, 2:20 PM

Originally Posted by Stony

HMaloriak: I respec Mutilate here for AOE damage

So, as someone who plays Combat, I am sort of confused by this and was hoping for clarification. Combat has the +75% thrown weapon damage, as well as being able to AR during green phases. So how is Mutilate better for AOE? Is it simply the poison damage from talents and Mastery? Or is it something else I am missing?

big30head · 01/24/11, 2:30 PM

Originally Posted by Lokar

So, as someone who plays Combat, I am sort of confused by this and was hoping for clarification. Combat has the +75% thrown weapon damage, as well as being able to AR during green phases. So how is Mutilate better for AOE? Is it simply the poison damage from talents and Mastery? Or is it something else I am missing?

Assassination does most of it's AoE dmg via poisons applied with FoK, not thrown weapon damage. Since the poisons are unaffected by the AOE cap, Assassination rogues do very good damage vs large groups of mobs.

Edit: ieatpaperbag's exerpt from the Assasination guide.

Of the three rogue specs, Assassination is capable of providing the most AoE damage using FoK. Keep in mind that FoK it self hardly does any damage; it’s the poisons that are easily applied from your throwing and melee weapons (via Vile Poisons) that do all the damage. This means that it does take several FoKs before your DP stacks up to 5 so that you can proc substantial IP damage from your MH and that, most of an Assassination rogues’ AoE damage “bypass” the AoE damage cap. Just to note, it is not yet know how many targets need to be in your vicinity for FoK to be worth using.

Stony · 01/24/11, 2:52 PM

On HMaloriak there is an additional black phase, where you must aoe 6 adds with 6+ mil hp each over 1 min. Mutilate has much better sustained aoe.

Verain · 01/24/11, 8:47 PM

Regarding your second point mulvad you are somehow right. Any second a Combat rogue perfroms his cycle on a mob he constantly lowers the cooldown of his abilties and therefore a Combat rogue does lose more dps than a Mutilate rogue if both aren't able to dps a mob.

I think this might depend on how long you have to spend away from the boss. Back in the olden days of Ulduar, while waiting for Mimiron to do stuff between phases, combat would be spinning blade flurry, killing spree, and adrenaline rush. Combat still gets to spin killing spree and adrenaline rush while mutilate gets to spin cold blood and vendetta. The difference is smaller now, but it intuitively feels like a long period off of a boss would favor combat. I do agree with your arguments as regards the bosses in this tier, however, and suspect that combat is punished more.

Sarvius · 01/25/11, 1:08 AM

Originally Posted by big30head

Since the poisons are unaffected by the AOE cap, Assassination rogues do very good damage vs large groups of mobs.

I don't think there's any current raid situation in which rogues are hitting aoe cap. Cap isn't based on the number of mobs (directly), it's a total amount of damage that can be done in one cast of FoK. So say FoK's aoe cap is 100,000 damage. We could aoe cap on 2 mobs if our fok did 50,001 damage on each target. Your point was obviously that vs large groups of mobs you're more likely to aoe cap - which is true - but it's never actually going to happen so it's not a contributing factor to assass being the aoe spec.

Assassination simply does better aoe damage because of poisons. On anything greater than 2-3 mobs it's going to be superior aoe - though not necessarily superior dps overall; blade flurry may still give combat that honor until x number of mobs are being hit.

SirWilliam92 · 01/25/11, 3:41 AM

Originally Posted by Sarvius

Cap isn't based on the number of mobs (directly), it's a total amount of damage that can be done in one cast of FoK.

I believe this is no longer accurate. They may have changed it back, but I recall them changing the AoE cap to be 10*single target damage. For example, if FoK did 1k per target, it would have an AoE cap of 10k. Above 10 targets, it would start splitting the 10k amoung them, but below 10 targets each would get hit for 1k.

Assassination therefore "bypasses" the cap because the poisons are considered separate to the FoK and therefore deal full damage to every target well beyond the 10 target "limit".

Cerevantes · 01/25/11, 11:22 AM

Originally Posted by SirWilliam92

I believe this is no longer accurate. They may have changed it back, but I recall them changing the AoE cap to be 10*single target damage. For example, if FoK did 1k per target, it would have an AoE cap of 10k. Above 10 targets, it would start splitting the 10k amoung them, but below 10 targets each would get hit for 1k.

Assassination therefore "bypasses" the cap because the poisons are considered separate to the FoK and therefore deal full damage to every target well beyond the 10 target "limit".

Here is the quote from Wowwiki about 3.3 and beyond aoe caps

3.3 will significantly alter the manner in which AoE damage cps are calculated. Instead of having a hard cap on the damage done, the maximum damage will be equal to the damage that would be dealt to 10 mobs hit by the ability. For example, if a spell does 1000 damage per target, it would deal 10000 total damage against 10 targets, and that would the maximum damage it could do against any number of targets (before critical effects are taken into account). Against 20 targets, the spell would do 500 damage per target, while against 40 targets the spell would do 250 damage per target. Against 10 or fewer targets, the spell would continue to do 1000 damage per target. Thus unlike pre-3.3 mechanics, the damage cap now scales with gear, the maximum number of targets that can receive full damage no longer scales inversely with gear, and the damage done once beyond target cap now also scales with gear.

I believe what Lokar is trying to say is that there are no situations in current raiding where you dps 10+ targets, and was wondering if 75% bonus throwing damage would equate 75% of Mutilate's damage being poison damage. Just from converting WoL numbers from our rogues we should show a 45% dps loss from non Mut AoE on Abberations (the only currently FoK'd raid encounter mobs)

Verain · 01/25/11, 1:51 PM

I believe there can be more than 10 lava parasites and more than 10 of Maloriak's "green phase" adds up at once.

Technically ALL rogue specs get around the aoe cap if they put deadly on their thrown, but I don't think there are any situations where this is advisable for combat or subtlety. But if we were in 50 whelps-for-30-seconds scenario, I think you'd do it (or actually just be mutilate to do it better).

The real thing seems to be this: Mutilate pressing fan gets to apply the thrown weapon poison, the offhand weapon poison, and the mainhand weapon poison, at rates specified by the weapon speed (IP or WP, the ppm ones) or a flat percent dictated by spec and poison (deadly, mind numbing, cripple). That just turns out to be much much better. My mostly 359ish (with a couple blues, no 372 yet as we can't do HMs yet) setup gives me like 75% extra poison damage from mastery, which is a rather lot. Mutilate's top uncapped stat directly boosts their aoe damage- other rogue specs just don't scale like that on aoe.

As combat, it's hard to judge the number of mobs where going from flurry and punch into fanspam is correct. Without mathing it, 4 seems like it could work and 6 definitely seems to benefit fan.

Cerevantes · 01/25/11, 2:03 PM

Originally Posted by Verain

I believe there can be more than 10 lava parasites and more than 10 of Maloriak's "green phase" adds up at once.

Technically ALL rogue specs get around the aoe cap if they put deadly on their thrown, but I don't think there are any situations where this is advisable for combat or subtlety. But if we were in 50 whelps-for-30-seconds scenario, I think you'd do it (or actually just be mutilate to do it better).

The real thing seems to be this: Mutilate pressing fan gets to apply the thrown weapon poison, the offhand weapon poison, and the mainhand weapon poison, at rates specified by the weapon speed (IP or WP, the ppm ones) or a flat percent dictated by spec and poison (deadly, mind numbing, cripple). That just turns out to be much much better. My mostly 359ish (with a couple blues, no 372 yet as we can't do HMs yet) setup gives me like 75% extra poison damage from mastery, which is a rather lot. Mutilate's top uncapped stat directly boosts their aoe damage- other rogue specs just don't scale like that on aoe.

As combat, it's hard to judge the number of mobs where going from flurry and punch into fanspam is correct. Without mathing it, 4 seems like it could work and 6 definitely seems to benefit fan.

Although you seem to understand that Mutilate has vastly superior aoe damage in all cases except low target cleaves, I have three brief corrections. One is that you don't dps the parasites on HM, making them a non target. Secondly is that there are 18 abberations on Maloriak, which are killed in packs of 9. In both these situations you won't be reaching the aoe cap, but Mutilate will still out perform.

Lastly, Fan of Knives does not apply the poison on your throwing weapon. Only the Mutilate talent Vile Poisons applies poisons on Fan of Knives, and those are the mainhand and offhand poisons. Sofar as I know, the only abilities that apply the poison on throwing weapon are Deadly Throw, and the base Throw ability.