Well, I've been gone for a week (AD-EU real life meet) and now this.


First of all: haste calculations

To get more reasonable results, you should incorporate some kind of mana consumption rate. Life Tap causes global cooldowns and therefore lowers your dps. E.g. If you need to tap once per 4 bolts, then you need to add 1.5s per 4 bolts to your casting time when calculating sustained damage.

total_casting_time = base_casting_time / (1 + haste_rating / 1577) + 1.5s / shadowbolts_per_tap

Note that how much mana you get depends on your spell power (80% coefficient), and Shadow Bolt cost can be talented to a 5% reduction.


Second:

I've had more than enough of the spec jyhad.

Some people like affliction. Some people like demonology. Some people like the 30/21/10 spec. Each and every one of these specs can top a raid's damage meter, and they can be fun to play and require skill to perform optimally. These are all facts.

Debating without solid evidence is very counterproductive and against the spirit of this board. Due to the nature of these specs (all have a different playstyle), objective results are almost impossible to get.

I've already went as far as to list some specs that I don't particularly think are raid viable (destro fire, for example) to avoid bruising egos. The compendium mentions the relevant debate. I will leave it at that.



Third:

To the warlock that claims not to use Life Tap: I'm very interested in seeing how you can keep up good dps without using it. Can you elaborate? I tend to go out of mana in less than two minutes. Mana potions are spent in less than half a minute. Am I missing something?


(was edited shortly after posting for clarity, resulting in a small crossquote)

Indeed. I'm one of those who saw an increase at the tail end of the T4 gear level.

Of course there is a huge YMMV tag to this whole thing. However, I was merely curious if the additional scaling of destro would eventually overtake the raid advantage provided by malediction. Obviously this depends entirely on your raid composition, since you need to estimate the benefits that 3% shadow damage has to all shadow damage-using raiders. (I was secretly hoping someone would hit it with a math bat so that I could calculate the break even point with some raid composition inputs, but alas.)

At what gear level are affliction locks around 85% the DPS of destruction? No Touch of Shadow is a huge DPS loss.

Leuiler tells me I'm spending 15% of my time life tapping, so this is somewhat true. However, the fact is that there are much better DPM places to put that spriest mana to use: hunters are fairly mana starved and their spells don't cost much, so the difference an spriest makes is significantly larger; mages can convert pretty much any quantity of mana into damage, given threat ceiling. The main reason I liked spriest is so I could "dip" from VE healing, so healers wouldn't have to top me off from LT. But that's not enough to justify me retaining the spriest over others. Much much much much more important is WoA. Spelldamage is almost always a warlock's best scaling stat, and a static 100 more is a pretty huge buff.

Also, do not discount the extra crit. One of detro's primary goals is increasing ISB uptime. Crit helps you do that. Checking assessment post most raids, crits make up about 40-50% of my damage done with SB.

Also, if you can guarantee the ele shaman, you can pick up gear w/o hit (or dump gems with it) to grab more damage, a very good trade-off. Assuming perfect itemization, that 4% hit you drop could be another 50-60 damage or another 2-3% crit.

I would hope everyone is discounting CoD/CoA here. If you are running any of those, you already are in a luxurious 4+ lock raid and likely will have debuff problems anyway. Some T5 raids might ditch CoE if enough of their mages go arcane, but I'm assuming that's going away now due to the recent mage buffs WRT fire/frost. A raid without 4-5 of (rogue, hunter, dps feral, dps warrior, lolret, enh, tank) is not going to be very common.

This is true, and likely the reason why T4 locks report higher than the "spreadsheet" says they should, because of the artificial scaling of the group buffs. This is why you should always tinker your simulations assuming full normal raid buffs (AI crit, DS damage, flask, food, oil, etc.) and come up with a couple of difference scenarios involving possible raid buffs (shaman, spriest, moonkin, etc.) Always do apples-to-apples comparisons with numbers if at all possible.

This still doesn't change the fact that destro does scale better and that past a certain point, is doing more damage. Even when you give the affliction locks all these benefits, they cannot catch up, as they just don't scale as well. Since you have limited buff slots, you use them where they benefit the raid the most.

This is a sensitivity for destro, as has been noted. Do not discount Soul Siphon (I know, I laughed too when I first heard it), however. Glancing over a few healing intensive encounters, I find I'm healing myself with that to the tune of 2/3rds what the affliction locks are healing themselves with SL. Will it entirely make up for it? No. But good healers, smart LT timing, and that talent make up for a lot more than you're granting here.

The only situations I've run into where this was a major liability (i.e., holding up the encounter) was when I was tanking T5 stuff. The extra healing stress does matter for Leo or Capernian.

Wait, 80% of destro? How many dots are you juggling again? That's a 68% cast ratio of SB with 4 dots to maintain as well as a LT/DP cycle to keep up with. The best I'm seeing with UA builds in the spreadsheet is a 55% cast ratio of SB with no haste gear and no immolate. That's a 13% difference.

Even with perfect dot rotations and equivalent gear, I'm still seeing spreadsheet numbers on the order of 100-150dps ahead by the time you're kitted out in T5. That is not the very high end. The earlier performance (in T4, for example) is most likely due to what you're claiming. However, the scaling does win out shortly thereafter: affliction scales with damage, destruction scales with damage, crit, and haste.

How is "tweaking your build" more important than talents? Seems to be the same thing to me. :-P

EDIT:
Well said. If someone has hard numbers to share, I'm all for them. Comparing on stuff other than data is what intarweb flame wars are made of. :-P

How are you not Life Tapping? Are you lucky enough to always have a Shadow Priest + pots? Do you sac fel pup?

Even with an Ele Shm and Spriest at some point your going to have to use LT to maintain top DPS. Other than stacking Int and saccing a lot of everything else I see no way for this to be possible and top meters.

If you paid attention to what I wrote you'd see, good spriest + RESTO shaman + mana pots (acquired and required addiction for this). Sometimes, if spriest gets nice dps I can even skip a mana pot. In a 5 min fight, no Life Taps, no stat stacking, sacced succu ofc. Ele shaman ... never tbh, resto seems way better as support for locks.
And we have a weird-ish setup that started with our all out on haste shaman. He always gets spriest. Since then caster groups (2x) are spriest + resto shaman + casters, rarely a healer among those or on some, few fights, healers get in. So yeah, on most of the fights, if I'm in range of the Mana Tide/Mana Spring, I don't need to LT.

And all this is pretty much dependant on how good your shadow priests' dps is, the better they do the less LT you need.

You may be getting by without Tapping using an abnormal group and possibly suboptimal gear with regen stats, but you would do more damage sacing Succy and Life Tapping.

Dakalro, please post some parses were you aren't life tapping. Like the above poster, I can't fathom how it's possible to maintain high DPS without Life Tapping. Lately I've been having a Resto Shaman and a T6 geared SP in my party (the ideal mana regen you cite in your posts), but I still run out of mana on everything but RoS (only because each phase is so short). Granted, I run with 145 Haste Rating (increasing the rate at which I consume mana). Still, though, I don't see how it's possible to avoid Life Tapping without multiple Shadow Priests stacked in your group.

Rough Calculations for Mp5:
Gear: 0 Mp5
Consumables: 0 Mp5 (no DPS consumables)
Demonic Sac: 0 Mp5 (Succubus sacrificed)
SP: 325 Mp5 (assumes 1,300 DPS SP)
Mana Spring Totem: 72.5 Mp5 (assumes 5/5 Restorative Totems and T4 bonus: ticks for 29 every 2 sec)
BoW: 49.2 Mp5 (assumes 2/2 Imp BoW)
Mana Tide: 40 Mp5 (assumes 10,000 raid buffed mana pool)
Super Mana Potion: 100 Mp5 (assumes average mana return)
Total: 586.7 Mp5
Shadow Bolt: -798 Mp5 (assumes 5/5 Cataclysm and 0 haste rating)
Net: -211.3 Mp5

With the assumed 10,000 raid buffed mana pool, your total time spamming solely Shadow Bolt is a paltry 236.6 seconds of casting (94 Shadow Bolts). [Note: With 0 Mp5 and a 10k mana pool, you will be oom from SB spam after 63 seconds.]

Dakalro, initially I was skeptical that you can DPS without Life Tap, but after doing (and re-doing) some basic mana regen math, four minutes without a Life Tap is feasible, although it does group compositions that may be sub-optimal for overall raid DPS.


Edit: In the time it took me to post this you posted two clarification posts. Still, your clarifications do not change any assumptions from this post (and I even grant you a larger raid buffed mana pool than you likely have). In one of your clarification posts you claim no life tapping over a 5 minute fight. If that's true, then you are not continuously casting (as the Mp5 math above proves).

Edit 2: Fixed SP Mp5 (can't believe I messed that up) and adjusted other values based on the replies from Dakalro. Conclusions changed drastically. Still, though, shows that Life Tapping is required on all but the shortest fights.

Unless I'm completely blind I didn't LT here . Ignore the rest of the spells cast, it was fooling around, an almost wipe and I died only a couple of seconds before boss, forgot to CoD after the 2nd one and overall not so awesome when all you get is "Aggro from Shadowy Construct" with 4-5 around you . True, kind of short fight, will try and find an older one if I still kept the log.

Shadow priest returns 5% of damage, 1200 DPS * 5 seconds * 0.05 = 300 mp5, not 60.

It doesn't have to put you on 0 mana usage, just enough to keep you going for the length of the fight though. Then there's this where I ended up with way more mana than I needed, around 60% or so after 3 min.

So... to go without LT (with 2.5s bolt and no curses...):
base_mana + mp5*fight_length/5 = fight_length*399/2.5

mp5 necessary = 798 - 5 * base_mana / fight_length
or
fight_length possible with no LT = 5 * base_mana / (798 - mp5)

How much extra damage is one charge of Shadow Vulnerability worth?

If I'm out of mana and everything is on cooldown, I can either wand or lie down (as a mage or whoever really). Now the last option doesn't feel very satisfying, especially when a boss is on very low health.

So, how does wanding (1.5s speed, 177 base DPS wand, shadow damage) compare to just standing around?
How much is one charge of shadow vulnerabilty worth in a typical raid setting?
Bear in mind that all endgamge wands are 1.5s shadow wands, non-shadow wands are from arena (1.9s, fire) or under 100 DPS, and the slower wands regenerate less mana from JoW. Not wanding also would regain no mana from JoW.

So, you see it's all pretty complicated, so I'd love to obtain a figure on the value of one Shadow Vulnerability charge to work with.

It is downright impossible to tell for sure. Spreadsheets can't come close due to horrible ISB hidden mechanics discussed in previous posts. ShadowSeer tries but can be horribly off if you're out of range to receive debuff messages. There are variables one can change to increase range on damage messages, and most mods do. It's been reported that debuff messages aren't affected by these. ShadowSeer has been put on hold until 2.4 for this reason, with me hoping that the much needed cleanup will fix these things. Currently its ISB data is 100% reliable only when staying within 30 yards of a boss at all times.

It also depends heavily on the fight and amount of shadow users. The kind of fight where you go oom is typically going to be a boss fight, so let's assume that.

In my experience, ShadowSeer has recorded values of about 1500-2000 added damage per crit with around 6-7 shadow users fighting a boss, on the rare occasion where I could get what I consider reliable figures. One could very roughly estimate that a charge is worth 400 on average in that raid.

Even if we halved this, wanding with a shadow wand seems to be very ineffective. Very low shadow damage hits every 1.5seconds are the Warlock's worst nightmare, and the reason I recommend not using Lash of Pain on a Succubus. Intuitively, I'd shy away from anything that reduces Shadow Priest dps. Almost 25% (=5x5%) of their damage done goes to mana starved people in a long fight.

For optimal results, make a "swap to spirit weapon and non-shadow wand and start wanding" macro. This would allow you to benefit from spirit, non-shadow damage, and JoW. The only wand that has high dps and doesn't do shadow damage is [Merciless Gladiator's Touch of Defeat] and it's cheaper younger brother [Gladiator's Touch of Defeat], according to wowhead. You might still have a pre-2.3 Evocation staff lying around.

One should be realistic, though. We're talking low numbers in a small portion of a long fight, it'd be very hard to measure. Obtaining a pvp wand for the sole purpose of increasing your raid dps contribution when wanding is pushing it quite far. But then again, that's what these forums are for.



If you really want to do your own research, have a melee person (like a MT) run ShadowSeer, they'll be producing reliable data about ISB damage contribution per crit in your raid. I haven't tested this in a while myself, but it ought to still work in 2.3.

It's not very standard from raid to raid because it varies by composition, but it's relatively easy to model: recalculate the ISB uptime adding in another "caster" with an effective 1.5s cast time and 0% ISB proc rate. Use the drop in ISB uptime to see how your shadow caster dps changes. My guess is, it will tank. Like, really bad. 1.5s is fast compared to cast speeds so you're basically casting as much as two destro warlocks, essentially halving their crit rate for ISB uptime models. If you have 5 shadow casters doing 1000 pre-ISB damage each, and knock the ISB uptime by 20% absolute (not unreasonable), you just reduced the raid damage by about 200 dps, or 300 per wand hit. I have yet to see any 200 dps wands. Note that this is dps, not per-charge; in a 60%-uptime model, each hit has a 3/5 chance of using a charge so each that does use a charge is 5/3 the damage of the average case, meaning 500 damage per wand hit.

In a bizzare counterintuitive twist of fate, this may actually be worse for your personal regen. Shadow preists doing less damage means you get less mana. I don't think the actual gain is comparable to JoW in any way shape or form but your group members will be angry with you if you made them go out of mana faster.

I recomend getting a non-shadow wand for actual JoW wanding. Preferably one as fast as you can find. Avoid nature damage wands too if you have enhancement shamans, for the same reason: 1k Lightning bolt > 50 wand hit.

Also, never ever ever wand in threat-capped fights ever, ever. Every non-subtletied point of damage that you do is 1.1 to 1.4 points of subtletied damage you could have done instead to reach the same threat-cap, meaning you just shot yourself in the foot by 10-40% depending on spec.

Thanks a lot for the info.

I know my fair share about things, and that especially shadow damage multipliers have a high impact on raid performance due to shadow priest mechanics.

I rarely have to resort to wanding, and always grab my old Evocation staff for that. Yet, there are situations when SPs go to healers, Evocation gets interrupted by damage after 1 tick (yay for 1.5k instead of 9k mana back) or we have to AoE or the fights are long.

Also, I have at least 10 different wands that deal shadow damage, with 1.5s speed or slower, with up to 184 DPS. And one 1.3s wand with nature damage (stormstrike + poison/earth shock) and 62 DPS.
I could get a 167 to 198 DPS fire wand at 1.9s speed from arenas, and just saw the 1.3s 96 DPS fire wand from Sporeggar.

Oh, and 200 DPS wands sound quite realistic. Take 184 DPS base (BT wands), -5% from partial resists, -5% from full resists, +10% from 20% crit (wands are affected by spell hit/crit on gear, has been tested) and add 5% misery, 10% shadow weaving, 10% curse of shadows and you're at 233 DPS before ISB debuffs.

Fire mages have an effective efficiency of ~8 DPM, meaning that 1 extra mana when OOM translates to 8 extra damage. (The mana efficiency as damage/manacost is higher, ~14 DPM, but that's not what matters here.)
So, the mana from JoW yields about 280 damage, compared to standing still and not doing anything.


Back to the drawing board, thanks for your help!

Speaking of Affliction and its utility - has anyone come up with the threshold that makes a Affliction/Ruin build out-DPS a standard UA build for personal DPS? Then we would factor in that a Affliction/Ruin build with a decent amount of crit (which one would have at the BT/Hyjal level) would be increasing ISB uptime significantly more than a UA build would.

All in all, without any real evidence, it seems to me that such a build at higher gear levels could be quite a good substitute for the "token UA Warlock" - it does provide all the utility (SE, BP and Mal) while increasing ISB uptime and at higher gear level possibly doing more DPS?

However, as I said, I haven't really done any maths on that, and as such I was wondering if anyone else was curious enough as to do a bit of calculation on the subject.

It depends on your crit and damage if Ruin would beat UA for overall damage dealt. Certainly if you have 4 Tier 6 you have passed the mark.

Either way, once you have some Tier 6 you likely will at least break even.

What would this build look like? Talent Calculator - World of Warcraft ?

Why would you have Imp CoA and Malediction in the same build assuming you're going to put up CoS in a raid?

Or two more points in Fel Conc? I think I CoA more than I DL, but they are both pretty minor improvements.

I guess the main point of my question was to see if taking two points out of Contagion was the best place in order to get SE, Mal, and Ruin.

It's a good spec, in my opinion. The compendium already lists a build similar to this.

imp CoA is not useless, as you spam Agony on trash. Not a big deal, though.

With that spec you are pretty much only using CoE/CoS, Corruption and Shadow Bolt. SL is ok in fights where the healing helps, but otherwise just casting sbolt and keeping Cor up wins. I also removed points out of Contagion and there's no other option really. All the higher tier stuff is locked in. You need 7 points in grim reach/nightfall/emp corruption, then 6 in SL/SE and 5 in SM. That leaves you with 7 points for the last two reachable tiers. Three in Malediction are a must (only reason for this spec, 21/40 is just more dps). So you can go with 4 in Contagion and 0 in Dark Pact or 3/1. I'd never give up Dark Pact for 1% Corruption/Seed damage. Maybe if there's a boss fight where seed dmg actually matters, but right now there isn't.

In the end it is still a debuff bitch spec, you cannot compete with 21/40 in any way. However with a certain amount of gear it's better then full affliction and after all one warlock should have SE/Malediction for any non-farm stuff and I'd much rather be malediction/Ruin then full affliction. Other then that with a certain amount of crit your dps is just better, you also buff raid dps more by casting more sbolts and use up less debuff slots.
Plus I really like having both destructive reach and Intensity. Spell pushbacks suck.
While using that spec I had 4/5 in Fel Concentration instead of imp agony/amplify curse as I was using CoS all the time anyway.