galzohar

Even though the variance in ISB uptime is bigger than the actual difference you'd get with increasing your crit, you still should be looking at what the *average* icreased by. If the ISB uptime increased by 5%, that's a 1% DPS increase to all your shadow users on average, no matter what the variance is. Wether that is or isn't worth getting those extra 10% crit or not depends how good 10% crit are for your personal dps compared to the stats you lose to gain that crit and how much dps the other shadow users are doing.

The only thing high variance means is that you cannot go to WWS etc and see the difference, just like if the fight is short enough you can't measure your crit chance via WWS (remember people reporting 20% crit on wws when they have 30%? or reporting 40%? doesn't mean their average crit rate isn't 30%...). Generally the best thing to do is to increase your average dps, as it will increase both the minimum reasonable dps you may roll and the maximum.

 

Spline

10% is a lot of crit. Stacking that up at the expense of spell damage, then going to WWS and not being able to see the difference is disappointing at best.

 

Evidicus

As a raiding Destruction Warlock, I understand the importance of keeping ISB up as much as possible. However, I think gearing for crit at the expense of damage is a mistake, regardless of a person's good intentions to help their raid. Anytime you do this, you sacrifice guaranteed damage for a chance to crit, and even a 40% chance to crit is still a 60% chance *not* to. I am no mathematician, but I've seen first hand that even with a high crit rate you can string along several non-crits in a row on a regular basis. Each Shadow Bolt is a new spin of the wheel with the exact same crit chance as the last one (unless you have Darkmoon: Wrath I suppose), so even Warlocks with the same crit rate can see variances in how many crits they do in the same fight.

I have socketed for more crit in the past only to see a drop in my personal DPS and no cooresponding boost in raid DPS or ISB up time, so I feel your pain. The end result was a complete resocket back to as much pure damage as possible. Based on my personal raiding experience and observations, I gear for around 25% or so crit (after being hit capped for bosses, of course) and then everything else gets put into damage. This setup has been the most effective for me personally by far, and I still put up my fair share of ISB debuffs.

If you can pick up more crit by getting a spot in a Boomkin or Elemental Shaman group, by all means do so, (free crit is always a great thing), but other than that I wouldn't go overboard stacking it. It just isn't worth it, in my opinion.

 

Arelenda

10% crit gives 5% increased uptime? That's a whole lot. It's less than 1% per Shadow User, though.

Example:
5 Shadow users, each at 1000 dps (WITHOUT ISB)
ISB uptime: 50%

ISB bonus = 20% * 50% = 10%
so real output = 5x1100 dps, total 5500dps

Now increase one warlock's crit by 10%. His personal damage goes up by roughly 8%: when assuming 20% crit rate upped to 30% the result is roughly (70+2*30) / (80+2*20). ISB uptime goes to 55% according to the model.

our critlock goes to 1080dps (without ISB)
the others remain at 1000 dps (without ISB)

ISB bonus = 20% * 55% = 11%
critlock goes to 1198dps
others go to 1110 dps (0.9% increase)
real output = 5638dps

Gain = 138dps raidwide or 12.6% of personal warlock dps.

For comparison: statwise 10crit = 221 crit rating. 221 regular spellpower would result in approximately 11% gain.


As a rule of thumb, I'd equate crit rating with spellpower as destro.

 

Evidicus

What many people will tell you is that Immolate and Corruption have little to no place in a Destruction build. As far as pure DPS theory goes, this is correct. Your hands on experience may vary slightly.

For example, there are several encounters where threat is a huge concern. If I am watching my Omen meter like a hawk (which I normally do given this build's tendency for threat spikes), then I can determine those times I need to Soul Shatter or "throttle back" a bit. One way I continue to DPS while still lowering my threat gen is to cast an Immolate or Corruption (provided the debuff slots are available, of course). That 2 second Corruption cast will lower my damage per second, but it beats pulling aggro, blowing a Soul Shatter too early, or just standing around with my thumb up my butt doing zero damage and just waiting to cast another Shadow Bolt.

Sometimes, regardless of how skilled your tanks are, something goes wrong. The boss could get a string of parries and hit like a truck, forcing your tank to worry more about mitigation abilities than threat gen abilities for a short while. The tank could dodge a lot of the incoming attacks and generate less rage than normal. These are also scenarios where pumping in 8k Shadow Bolt crits right away may not be the best idea.

Another example could be a fight that involves a lot of movement, like Archimonde. There have been several times where I've put up Immolate and Corruption prior to running away from Doomfire, just because some DPS is better than zero DPS.

Personally, while I greatly appreciate all the theorycrafting and spreadsheets I find here and on similar forums, the truth is that they are all tailored towards the best case scenario. The problem with taking it all too literally is that raiding doesn't happen in a bubble where random variables and unexpected situations never happen. The data and opinions in this thread are all extremely valuable guidelines, but that is all they are -- guidelines. What I would advise, Amandil, is take everything with a grain of salt and just be ready to think on your feet. After all, if anyone had all the answers out there, these threads wouldn't need to be as long as they are.

 

DaveA50

If you need to watch threat, I don't understand why you would want to use corruption. With a 0/21/40 build, you'd have Destructive reach for -10% threat from destruction spells, but no improved drain soul for -10% threat from affliction spells. However, for fights like archimonde where you know there will be a time when you can't dps, such as 5 seconds before the next fear, using both corruption and immolate would allow for some dps while incapacitated.

 

Spline

For a 0/21/40 build with Destructive Reach, Corruption produces more threat per point of damage than Shadow Bolt or Immolate. You shouldn't ever cast Corruption if you're worried about threat. Get that thumb under there.

 

Evidicus

It's not spike threat though, so my thinking was that it's easier to manage and easier to watch myself "coast" to a safe area on the threat meter so I can ramp up another Shadow Bolt. I'll try the thumb theory though, just to see.

 

Krazen

I don't think gearing for crit is that much of a concern, because in any item slot there are only a handful of pieces with set discrete values for each slot to compare with each other (ie Vestments/Rhonin/Malefic/Corruptor for chestpiece).

I think the bigger question is whether or not to Spinel a given piece, given your guild's constraints on gems, and whether to pursue a socket bonus or not. For instance, if Spinels are limited, is it really fair to grab 2 for your Belt of Blasting?

 

galzohar

Now the real question is if 10% crit, in fact, gives 5% increased uptime in a normal raid as stated above?

 

Arelenda

Even so, all spells cast before you have to Soulshatter are at 50% threat. I can see the reasoning behind the theory "if my next bolts crits, I might draw aggro. I'll cast Corruption/Immolate for no immediate threat, then go back to bolting. If both are up, I'm best off using Soulshatter". I prefer using Life Tap/Immolate over Corruption in that scenario myself, since it has no range or threat issues.

Either way, one late Soulshatter usually is enough for me. Maybe if my gear improves I'll get into this scenario more often.

I think it is quite obvious. Whatever spell is highest on the spreadsheet, common sense should prevail. If stuff doesn't live long enough, don't use dots. If threat is an immediate issue, don't use potential 8k Shadowbolts. If you need to cast something fast so you can move early, go for Immolate.

 

galzohar

Rarely would you actually be able to tell that you have 1.5s to cast but 2.5s will take too long. Even bossmods timers don't seem to be accurate to the level of <1s.

When corruption is less DPS than shadowbolt, casting it is just silly. Even if you have to move you'd still be doing more dps spending that time on shadowbolt, and also get less threat per dmg done. If threat is an issue no reason to cast a spell that's less threat efficient. Also some would claim that with certain spells they can "ride the tank's threat" closer than with others, but in reality it really doesn't matter if you give the tank a 2k or 4k threat buffer, as in the long run it'll be a negligible amount of dmg relative to the total dmg you did in the fight. If threat is an issue you need to slow your dps, and while there are many ways to lower your dps and it really doesn't matter how you do it, just don't use abilities that actually generate more threat per point of damage done...

 

Kobal

This way of thinking is a common fallacy.

The main reason that Corruption and Immolate are not used in a "regular" DPS rotation for a 0/21/40 build is that their damage per cast time is lower than that of shadow bolt.

However if their damage per cast time is lower, then it also makes no sense to cast them right before running. The better strategy would be to cast some shadow bolts instead. Yes, if you cast some SBs in those 5 seconds, you will do no damage while running, but the additional damage you did before being on the run more than compensates for this.

 

PSGarak

Has anybody done any closed-form solution modeling for [Darkmoon Card: Wrath]? I know there have been X-thousand-cast simulations at various crit levels, but what I'm really interested in is an explicit formula for how it affects ISB. It increases uptime more than the same amount of crit because it redistributes the crits favorably. I think it's important to know, and I would just like some warning if I'm about to reinvent the wheel before I go off and model it on my own.

http://sig.gamerdna.com/quizzes/INFL...erks/Garak.png

 

Krazen

What would the point be? Even under the best analysis, there's no way the trinket compares to Icon/Crusade/Hexhead/Skull. Plus the fact that it scales negatively with +crit on better gear...

 

• GokieKS

This is precisely the situation I faced. Our guild's supply of Crimson Spinels were always short, to the point where people have to be on a waiting list, so I ended up using [Potent Pyrestone] and [Glowing Shadowsong Amethyst] on any piece that had a good socket bonus, with care to only have two of the Purple gems (for the CSD meta requirement). Basically I went with socket bonus colors on everything except the T6 chest, where the difference between 3 [Runed Crimson Spinel] and 2 Yellow/1 Blue + Socket Bonus would've been too great. It was also a consideration I had when I decided to pick up the T6 pants (which is really a pretty mediocre piece compared to T5 and the other two T6 level cloth DPS legs) instead of waiting on [Leggings of Channeled Elements] - the Kaz'rogal pants would need 3 +12s to shine at its best, and we simply didn't have enough of those to go around and I don't even know when I would've gotten them.

And I really don't consider Belt of Blasting to be good enough of an item to warrant re-gemming for when you gain access to the BT/MH gems - there's a ton of +hit available from all the rest of the T6 level pieces, so all it really is is just a decent stat-less belt - both the belts from Anetheron and Supremus are upgrades, and neither boss is particularly hard. So barring some really bad luck with drops, both are very viable options.

 

Shuko

Darkmoon Card: Wrath - WoWWiki - Your guide to the World of Warcraft

 

Shuko

I have to disagree with that. Blasting is actually the best/on par with Noose. Both have about 1.4% crit, so its 5dmg vs 23hit. With sta from T6, the stats are a nonfactors. With that hit you can use a different wand or cloak for example for more crit/dmg. Last time i checked, Supremus belt didnt even come close to those two, because missing 2 sockets just hurts too much.

 

Xelo

As a affliction lock.. Is it even worth casting Immolate unless scorch debuffs are up etc.? Was just curious I've seen so many people say different opinions. I took it out recently on bosses, unless I'm moving around a bunch or on trash.

Thanks.

 

Arelenda

I suggest you read the compendium. It has cookies. Also, a section called "Immolate vs Shadow Bolt".

 

Afflictor

has anyone worked on what a best in slot set would be with a 13% hit cap? (elemental shaman)

 

Eph

Posted a little ways back, http://elitistjerks.com/f31/t17008-w...tml#post573114. Posts before and after discuss other gear sets as well.

 

Pruka

Edit: Found the thread that answered my question.

 

KuruQan

OK,can somebody answer me,why the OP thinks CoA is better than CoD?CoA has 120% spell coefficient,while CoD has 200%.CoD is 70DPS,while CoA is 56.5DPS+you lose a gcd for reapplying CoA.I dislike all the locks,who use CoA in raids,so have I been wrong all this time?Is there something Im missing?

 

Krazen

Well, a pyrestone is ~11 spellpower equivalent for personal dps, so you lose more by putting them in 2 socket items rather than 3.

For example, on any YYB item (both legs, t6 robe, vashj robe): A choice of 36 spellpower from Spinels, or 11+11+6+5 = 33 spellpower from the alternative. Each spinel is a 1 spellpower gain.

Whereas on any YB item(belt of blasting, t6 shoulder): A choice of 24 spellpower from Spinels, or 11+6+4 = 21 spellpower from the alternative. Each spinel is a 1.5 spellpower gain.

On a single socket Y item (t6 gloves), the obvious gain is 1 spellpower, and on a single socket blue (carved wand), the gain is 4 spellpower.

I already socketed my slippers of the seacaller for the bonus, but in hindsight I kind of wish I didn't. But I think the priority should be those 2 socket YB's first, once metagem requirements are changed on the CSD.