PVE Raiding Compendium - Page 2

Arveene · 10/12/07, 2:02 PM

Originally Posted by Cohren

There is definitely more favorable DoT rotations than others. I think the most common or most widely agreed upon rotations are:

CoS/CoE/CoR:

CoS/CoE/CoR-->UA-->Corr-->Immo-->SL

CoA:

UA-->CoA-->Corr-->Immo-->SL

The problem with a DoT rotation like Curse --> UA --> Corr --> Immo --> SL is your UA, Corr, and Immo timer will sync up every time you refresh them (assuming no resist).
Ex. UA applied at the same time as Corr (1.5 sec GCD) Immo (1.5 cast). When Immo is applied both UA / Corr have 15 seconds on their durations.

A better rotation would be something like Curse --> UA --> Corr --> SL --> Immo.

Silverstorm · 10/12/07, 2:38 PM

I would say the rotation to avoid is UA -> Corr -> Immo. However, that does provide ease of maintenance because all three end, so you know you have to refresh all three at the same time.

Switching the rotations around works for a a little while, but given the 18 seconds of UA/Corr vs. 15 seconds of Immo, they are going to line up again anyway, and you'll lose uptime on Immo, most likely.

I'm playing around with dropping Immolate from my rotation (43/0/18 spec) because the DPCT is roughly equal to DPCT of Shadowbolt in our raids (no Fire mages right now, all Arcane). So I have an Immolate that only gets Misery to buff it, compared to a Shadowbolt with Shadow Mastery, Shadow Weaving, Maledicted CoS (from me), Misery, and potential ISB debuff (and maybe something else....I can't remember right now).

Arelenda · 10/12/07, 5:09 PM

Originally Posted by Chronitia

Frozen shadowweave provides all the stats you need for affliction, to an almost absurd extent. It is possibly the most min-maxed itemization strategy I've ever seen on a set of gear. Socketed appropriately, it provides more spell damage than tier 5, and is available the moment you ding 70.

While it gives a good leg up to any spec due to it's early availablility, it is only the affliction playstyle that it truely fuels. As said before me, it does not give you the hit (and crit) you need so badly with 21/40.

No, but it is so absurdly good that you can compensate for the lack of hit on your other gear, or get the socket bonus for +9 to hit on the set. +damage isn't exactly useless to a destruction warlock either, they get even more out of it than an affliction warlock does.

About ISB theory:

Unfortunately, I don't believe the ISB can be accurate measured by any spreadsheet, as the debuff application lags a LOT after the crit. When I wrote ShadowSeer to monitor it, I logged delays of 0.7s, sometimes up to 1.2 seconds. Good luck measuring that with a spreadsheet.

The debuff removal can also lag, resulting in extra hits being amped.

It also can't take into account how much dots are on the target (dots get amped too). I'll trust ShadowSeer over any spreadsheet, given how much the game mechanics differ from what they're supposed to do in theory. It's still beta, but I think the ISB data it gives me is reliable enough for analysis.

djmoccasins · 10/12/07, 5:17 PM

I really think you should put a 7/43/11 build up there its probably much better damage than the 0/41/20 and less down time and is good all around for PVE and PVP.

Arelenda · 10/12/07, 5:35 PM

Originally Posted by djmoccasins

I really think you should put a 7/43/11 build up there its probably much better damage than the 0/41/20 and less down time and is good all around for PVE and PVP.

Can you be a bit more specific? I don't mind adding specs I'm not familiar with myself, but I'll need a lot more information than "7/43/11".

Also, I'm modeling raiding specs. PVP viability isn't really considered.

djmoccasins · 10/12/07, 5:39 PM

Oh I'm sorry Talent Calculator - World of Warcraft is probably the most popular maybe switching a few points. It provides great damage IF you can keep the pet alive and the instant corruption makes up for the lost crit from the 0/41/20 build

Arelenda · 10/12/07, 5:47 PM

Originally Posted by djmoccasins

Oh I'm sorry Talent Calculator - World of Warcraft is probably the most popular maybe switching a few points. It provides great damage IF you can keep the pet alive and the instant corruption makes up for the lost crit from the 0/41/20 build

I think I got the idea.

I don't see how gaining 0.5 seconds on a spell that you cast once every 18 seconds will increase your damage over 5% crit, provided you have ISB. I'll add 1/40/19 to the list, though.

Instant corruptions is nice when pvping, but for raid scenario's, I doubt it's 5 talent points in a tree you're not specialising in.

Antoine · 10/12/07, 5:50 PM

The real gain is in Improved Lifetap. Improved Corruption helps too, but not as much as imp LT.

Arelenda · 10/12/07, 5:58 PM

Good point. I can't tell whether it's worth it on the fly. Added.

• frmorrison · 10/12/07, 6:10 PM

I think you should add the PvP/PvE hybrid Demo spec, 7/44/10 Talent Calculator - World of Warcraft
The gain of imp Life Tap is very high, since it means 20% fewer Life Taps.

Also add the note that 1 point in mana feed is enough to allow the Pet to spam Cleave for over 5 minutes.

For 21/40, you should add a comment about Fire destro Talent Calculator - World of Warcraft, saying it is a good spec if you don't have a shadow priest in the raid, and it is more mana efficient (although losing damage over shadow bolt even without a shadow priest around, due to Incinerate only getting 91.4% +damage added to it).

Trickykid · 10/12/07, 6:35 PM

I would point out that for a single shadow lock ISB uptime is very easy to estimate using 1-(1-c)^4.

Expanding that out for a whole raid can be done relatively simply by weighting the crit rate of each member by his share of nukes on the target. Using this weighted crit rate in teh above expression will yield a pretty accurate reading for expected ISB uptime in raids and helps to estimate the gains of crit rating on items outside of personal DPS.

EDIT:
I would also add that more than most other classes (puuuure speculation) warlocks have a larger variance in their DPCT available to them at any point in time. For example, an affliction lock's DPCT over time varies depending on which DoTs he has active (and what procs/buffs/debuffs are going). Once all his dots are up, his DPCT drops since he's forced to use his lowest DPCT spell, shadow bolt. Saving LT/DP for those low-DPCT valleys, using pots when you have no mana but are in a high DPCT peak, and similar decisions will increase DPS quite a bit over time.

KnThrak · 10/12/07, 7:26 PM

Originally Posted by Arelenda

Curse selection

(...)

Curse of Agony - Ideal for spamming on tanked trash, regardless of spec. Can also be used as an alternative to Curse of Doom, if that one is not an option. CoA is advisable on bosses that reset aggro (unless you can time CoD well), or on anything that will live 24 seconds but not 60. It can outdamage Curse of Doom on a strict dps basis, but you'll need to refresh it more often, resulting in a lower Damage Per Casting Time.

(Editor's note : for all reasonable specs and gear quality I looked at, Curse of Doom outdamaged Curse of Agony if you replace the lost casting time with Shadow Bolt spam. This seems to go against the affliction Warlock love for CoA. If I'm wrong about this, please correct me)

(...)

This is something I've put a lot of thought into while watching myself raid, and I've come to the conclusion that the biggest factor of evaluating CoA vs CoD for a damage-only basis (ignoring special stuff like say Curator) is the bossfight itself.

Due to the lost time, as you say, CoD beats CoA on it's own. However, that is assuming one is always casting when Curse of Agony needs a refresh.

Depending on bossfight, it can very well happen that at say 50% of the times when you'd need to cast CoA again, you're moving anyways, not casting shadow bolts. In those situations CoA's superior damage (and superior scaling I think, although this is off top of my head, no calculator at hand) can outdo CoD due to not actually losing much casting time at all.

Arelenda · 10/12/07, 9:50 PM

Originally Posted by KnThrak

This is something I've put a lot of thought into while watching myself raid, and I've come to the conclusion that the biggest factor of evaluating CoA vs CoD for a damage-only basis (ignoring special stuff like say Curator) is the bossfight itself.

Due to the lost time, as you say, CoD beats CoA on it's own. However, that is assuming one is always casting when Curse of Agony needs a refresh.

Good point. I hadn't thought of that.

However
- The type of fight is rather rare
- Affliction locks probably have better things to cast
- You can also life tap (or dark pact) during movements instead. This is in fact what I usually do, as 0/21/40.
- And if you need to be moving that much, you might want to stick with CoD since it requires less attention, allowing you to focus on what you're running from in the first place
- the difference is marginal at best. Over a two minute duration you hit 2 CoD (8400+4*sp) or 5xCOA (5x1356+6*sp). For affliction warlocks with imp CoA it probably is better, since CoA gets 10% from Shadow Mastery and 10% from imp CoA.

I'll mention it in the guide.

PSGarak · 10/13/07, 12:21 AM

Specific notes for common raiding thingies:
Darkmoon Card Crusade: Any spec, but especially affliction, has a really easy time keeping this stacked, and getting it stacked while the tank builds threat, because we have spammable low-cost low-agro long-range curses. It's often worth the time to debuff a target to keep the stack rolling.

Clicky Trinkets and affliction builds: Clicky trinkets should be saved for when all of your DoTs are going to be refreshed in a small window. If at all possible, on 20-second clickies, try to double-tip corruption and unstable affliction. It's worth saving the trinket for a few seconds, it probably won't cost you an extra activation.

Procs and DoTs: Recasting a DoT and overriding a previous one is never, ever worth it, no matter how awesome your proc is. You lose half a tick on average, and the damage gain from the amped DoT is not worth that half a tick. For example, 2pcT4 proc is 135 damage. Over the course of a corruption this comes out to 170ish before debuffs. The lost half-tick of corruption is 250 at that gear level. Ignore your procs, and keep casting as if nothing happened.
In a similar vein: while you want as close to 100% DoT uptime as you can manage, you want to absolutely avoid clipping DoTs, and also avoid waiting for DoTs finish. If a DoT is half a second out, cast a shadowbolt and refresh it late.
The exception is UA, and untalented corruption: it has a cast time. Remember this.

Finally, affliction builds actually do scale really well, it's just that they only scale with damage since crit is marginal and we hit-cap easily. Higher-end gear tends to spend a lot of itemization points on crit and haste, which help destruction far more. It's not exactly a matter of scaling, it's a matter of scaling with the itemization that we will actually encounter in higher raid settings.

And yeah, I hear enough about 0/40/21 that it should probably be mentioned. It plays essentially like a felguard build with a sexybus instead of a felguard, but it gears more for crit. Example spec(?):
WorldofWarcraft.com -> Info -> Classes -> Warlock -> Talent Calculator

As far as affliction builds go, I would list this:
WorldofWarcraft.com -> Info -> Classes -> Warlock -> Talent Calculator
as the absolute bare-bones affliction build, with most of the rest being determine by gear, playstyle, and raid group composition. You will note it's not a viable build: it requires fleshing out. I don't know how to make wowhead's talents do that.

Arelenda · 10/13/07, 2:00 AM

After doing tons of test and theorycrafting and writing a complete addon to measure my damage, I've discovered that I've been downranking Immolate for about a year and a half now.

I must have forgotten to update the spell when I learned the AQ20 book. And because it wasn't top rank, it didn't get autoupdated when I learned the new max rank in TBC.

No wonder I always hated fire so much.

So uh... always use top rank of your dps spells.

Thanks for the useful posts, I'll merge them with the guide as soon as possible.