[7.2] Legion Shadow Priest Guide and Discussion

[Deleted]

i uploaded a better one, where i dont fail that hard at the end. i use dispersion at the start to make it easier, but you can do it without.

3rd torrent with 7600 haste.
https://youtu.be/yAWjLHVpU4E

[2Dee]

I'm in a predicament and I am hoping you all can help me, at least to some degree. My guild is on loot council and they've been pretty damn conservative about giving me loot, even though I am a full-rank raider and thus I'm not always getting the absolute best gear for my needs. At my ilevel I know I can be haste-capped, but thanks to them not tracking the fact that they've hardly given me anymore than two pieces of loot from heroic EN since the beginning, (all my other gear is from dungeons, WQs, or personal Seal rolls) I'm not geared well. Does being close, but not quite haste-capped affect our dps so much that we won't rank well at all? I am seriously getting disheartened because of raiding this expac and I really hate doubting myself, I strive to do the best, so I am humbling myself and coming to ask you guys to look at what I can improve, maybe in casting and such, idk...

Here's my armory and a parse link where you can mull through other boss pulls on other nights too.

We have very similar gear except my haste is around 10,800 and I have the legendary ring instead of a useful one like yours and I'm parsing around low to mid 80s. For Xavius, parcing high requires a bit of luck. If you don't reach the dream in the first stage -> get max corruption while maintain S2M -> S2M and corruption at phase 3 you will not be very high. Overall Xavius is a poor fight to compare to your fellow priests imo, unless you get in the dream in P1 do not bother.

I looked at your Mythic Nyth fight and one key thing I noticed is that you use PI way too early and dying way too soon. Your using it at 50ish stacks while the boss is in RoS range. You need to work on your rotation in S2M to reach at least 70-80 stacks before using PI. Unless you can reliable reach 100 stacks of VF you will not rank very high. Another comparison, your peak dmg in S2M is 880k dmg/sec, mine (less haste, irrelevant legendary, almost identical ilvl) is 1,250k/sec and that only puts me around 78 rank.

[Sxq]

I'm in a predicament and I am hoping you all can help me, at least to some degree. My guild is on loot council and they've been pretty damn conservative about giving me loot, even though I am a full-rank raider and thus I'm not always getting the absolute best gear for my needs. At my ilevel I know I can be haste-capped, but thanks to them not tracking the fact that they've hardly given me anymore than two pieces of loot from heroic EN since the beginning, (all my other gear is from dungeons, WQs, or personal Seal rolls) I'm not geared well. Does being close, but not quite haste-capped affect our dps so much that we won't rank well at all? I am seriously getting disheartened because of raiding this expac and I really hate doubting myself, I strive to do the best, so I am humbling myself and coming to ask you guys to look at what I can improve, maybe in casting and such, idk...

Here's my armory and a parse link where you can mull through other boss pulls on other nights too.

http://us.battle.net/wow/en/characte...eltress/simple

https://www.warcraftlogs.com/reports...nkings&fight=5

You've got some pretty basic issues with cast priority (opener/non voidform is wrong, not nearly enough void bolts in void form) and you're often dying extremely early or surrendering too late.

[ThePhantombox]

i uploaded a better one, where i dont fail that hard at the end. i use dispersion at the start to make it easier, but you can do it without.

3rd torrent with 7600 haste.
https://youtu.be/yAWjLHVpU4E

So, should I just take this as a testament to how good crit is? I'm sitting at about 20% crit and 30% haste now and I'm able to do three void torrents, but it's REALLY close and not that feasible in raid. I noticed you have 25% crit, does the extra 5% make that much of a difference?

[Deleted]

i uploaded a better one, where i dont fail that hard at the end. i use dispersion at the start to make it easier, but you can do it without.

3rd torrent with 7600 haste.
https://youtu.be/yAWjLHVpU4E

Okay, done it with better SW usage and a greedier PI. Still way more butt clenching that what you did here.

[Ocedic]

If you could produce a video of you reaching 3 vtor on a dummy without a bloodlust with 10500 haste, I'd like to see it. I can't reach it with 11000 haste (I suppose I use my PI too early).

Nevertheless, even with only 2 vtor he should be able to squeeze at least 50k more dps without too much effort.

Are you using double dispersion? Either way my post wasn't saying his DPS is low because of lack of tri-tor. I was more making a point that his gear was fine and that shouldn't be his main focus to improve.

[Yellowfive]

The thing is that you need to include multiple fight setups etc. to gain a better balanced statweight/gearing strat. If you only build your toon for one fight type, you will be kinda lackluster for all other fights. So it's not the accuracy of the actual sim data, but the overall utility of the weights you gain from that. A more balanced build is likely to be performing better than one that does only, say Single Target simulations. I think this needs to be emphasized for lots of gearing advice.

True -- right now we have some defaults based on a single fight type and popular talent setup. We'll be adding many more options very soon, and then our "auto" option will examine your character and choose the best fitting strategy that we have pre-calculated (or you can always put in your own). The "auto" option will work within the particular fight type you want to optimize for.

As for generalizing across different fight types, that's a non-trivial problem that we are still working on. The main problem is that nobody has the correct data to do this right now. simc is notoriously weak for modeling any kind of real fight situation. We have built our simulation from the ground up to support better boss scripting, and we are almost done with several new scripts that we will generate gearing strategies from -- scripts that realistically follow actual boss encounters.

Once that is good to go, we'll perhaps work on creating strategies that "merge" the data from different fight styles to create a "general" strategy. Nobody is doing this well right now, anywhere... unless someone has magically figured out how to generate good data for it that I have missed. If so, please point me to this data! In the short term, we'll probably tell people to gear for the specific scenario that is most interesting to them, or hardest for their team to beat. You'll still do good enough on the other/easier fights for you.

[Kilee25]

This is a subject that I have always had a strong, but seemingly lone opinion on. I actually do not like simming multiple fight types and deriving stat weights from them. My preference is to find what I am absolutely best at, and then gear myself to do that exceedingly well. If I find myself in a situation where I cannot perform at my optimum, I simply say, "Oh well," and forget about it. In my opinion, any sims that take me away from the goal of being strongest in these area/s that I want to be strongest in, is basically weakening me and diluting my performance. I do not care to max my damage in situations where I am not that good at to begin with.

In general, I run a 1-target and 2-target fight, and split my stats between them about 80/20. And I will drop the 2-target stats if I feel they are significantly and negatively impacting my ST stats. If that happens, I just build two sets of gear instead.

I also run a light movement sim just to get an idea of how my stats are affected, but I rarely incorporated these into my stats.

I have to stress that I don't do this out of laziness or a lack of understanding of what the "multiple fight" styles are trying to accomplish. Rather, I genuinely disagree with that methodology. I feel that deviating from "simple" fight styles is only useful for bleeding edge progression, not for generating generic stats for "average players."

Whenever I see TCers throwing up that whole, "My stats are better than yours because I looked at X more fight types than you," I usually just roll my eyes because I don't feel they are saying something as important as they think they are saying.

Regardless of how you feel about it, it doesn't make the body of work from AMR significantly worse compared to SimC. Certainly, not nearly enough to call it, "Crap," etc.

[Yellowfive]

In general I'd agree -- not a huge fan of trying to create gearing strategies that try to "average" fight types together. We have tried this in the past and in general it wasn't that well received, only reason we would do it though is if people demand it for some reason.

We also don't want to go the route of trying to perfectly model every single encounter in excruciating detail... that's overkill, and not really necessary to get good gearing strategies.

But we think there is a middle ground there -- model a few scenarios that are archetypes of common fight styles you will do as a player. Instead of completely manufacturing fake and unrealistic situations (e.g. a 2 or 3-target fight where those targets are present for 100% of a boss fight... very rare real boss situation), we use scripts of actual fights that are representative of different styles -- scripts that gloss over the finer details (so that they don't end up too specialized to be generally useful), but that aren't so generic as to also not be that useful.

For example, at this moment in time, if you care about single-target damage, you should probably use a short script. The major single-target DPS checks aren't that long. It just makes sense for ST theorycraft to be tuned for short fights right now. For more persistent AoE damage, the major case for it is Mythic+ we think. You should probably tune your AoE strategies a bit more towards that style of fighting right now. For cleave, it's something like Cenarius with short sections of cleave mixed with single target -- persistent cleave isn't a major thing right now, you shouldn't gear for it.

When new raids come out, I would argue that we look at the new encounters, pick a couple that are "representative", and tune gearing strategies around them. You'll perform better without having to go overboard with super-detailed boss scripting. I think the simc attempts at this have been poor at best... either adding N persistent targets or just randomizing stuff. Better would be to model it after actual patterns you will encounter in the current progression raid with enough "fuzz" that you can't just game the simulator and get bad results. That's what we are doing at AMR -- we are convinced that it is a fundamentally better approach.

I really wish we had it completely done at this moment... would make this discussion easier. When it is done, I think TCers will have a lot of fun using the AMR tools to come up with pretty solid theories.

[Caladia]

Sitting at the main stage and seeing that 7.2 will have new artifact skins based on a challenge and none of the Priest ones stand out as being any different than base to know it's even upgraded. But hey, owl mount! :P

Looking forward to seeing what the 7.1.5 changes bring later next week.

[davesignal]

That owl mount will look stupid as hell in Shadowform. All the glorious-glowing-LIGHT stuff always looks terrible in Shadowform.

[tikk]

ok i just got a ethereal ern 895 ilvl from night bane is this anygood i can replace a 855 oakheart with a socket or a 870 swarming with a gem socket can someone give me addvice

[Daedius]

Noob question;

Do you always enter void form regardless of how high lingering insanity has stacked at?

[Deleted]

ok i just got a ethereal ern 895 ilvl from night bane is this anygood i can replace a 855 oakheart with a socket or a 870 swarming with a gem socket can someone give me addvice

No. However keep it for Dragons when the fearing asshole is in the sky.

[Deja Thoris]

I'm in a predicament and I am hoping you all can help me, at least to some degree. My guild is on loot council and they've been pretty damn conservative about giving me loot, even though I am a full-rank raider and thus I'm not always getting the absolute best gear for my needs. At my ilevel I know I can be haste-capped, but thanks to them not tracking the fact that they've hardly given me anymore than two pieces of loot from heroic EN since the beginning, (all my other gear is from dungeons, WQs, or personal Seal rolls) I'm not geared well. Does being close, but not quite haste-capped affect our dps so much that we won't rank well at all? I am seriously getting disheartened because of raiding this expac and I really hate doubting myself, I strive to do the best, so I am humbling myself and coming to ask you guys to look at what I can improve, maybe in casting and such, idk...

Here's my armory and a parse link where you can mull through other boss pulls on other nights too.

http://us.battle.net/wow/en/characte...eltress/simple

https://www.warcraftlogs.com/reports...nkings&fight=5

This isn't a gear issue. It's a player issue. Seriously, when you look at your performance you need to look at whats in your control, not whats out of it. Your guild is poking into mythic and you have cheapo enchants but that minor. Lets look at your cenarius log for starters.

https://www.warcraftlogs.com/reports...-done&source=2

You died at 4:30 in due to StM, w'll get to that later. First thing of note is that you entered void form 5 times but only used void torrent once. You need to void torrent immediately because it keeps you in void form longer. (Make sure your dots aren't going to run out - if they are delay void torrent till after first void bolt)

https://www.warcraftlogs.com/reports...&view=timeline
Your opening sequence is wrong. You need to open with a MB just before the pull timer so it finishes its cast at the "start" of the fight. Follow up with VT SW:P and then flay till MB is back off cooldown. Your first MB was at 2.5 seconds. Second was at 11 seconds. Thats an 8.5 second gap for a spell with a less than 7 second CD.
Hit MB on CD is the lesson from this.

You also need to swipe more shadow word deaths. Lots of adds died but you only hit 5 in the time you were alive.

Finally, StM. I mean you died stupidly here but it still shows some fundamental flaws in how you played it. If theres anything you're going to take away from this it has to be how to StM well and I'm not going to give the full lesson (God knows I suck at it too) but some pointers and I'll repeat the *why* its important.

Mass hysteria increase dot damage by 2% for every void form stack. Combine that with faster dot ticks and you get exponentially more value out of StM for every possible second you can hold it. The value in StM comes from high stacks so you have to get there consistently. 3rd void torrent is the dream but if you're shit like me be happy with 100+ seconds :P

You pop StM and PI at pretty much the same time. This is wrong. PI is saved for later in voidform to stop you flopping over. It's haste stops you dying. Your aim with it is to save it for as late as possible. Probably 85-90 stacks but until you find your feet with it 70 is fine.

The quick version on how to do it properly is refresh dots. Enter voidform. Void torrent immediately. Disperse. Void bolt >>> rotation (theres plenty of places that explain it) Most important thing by far is to void torrent on CD. Nothing should delay this.

Finally, a good UI helps. Make sure you can track dots well on nameplates. Have WA's for important cooldowns. Theres an insanity bar WA, get that and theres also a "time to die" WA. Get that and plan to StM at 2 minutes left on a fight.

Good luck.

[beefeetr]

How do people handle PI+Void Torrent+SF in regular rotation? I feel that it's best to time PI so that it ends as soon as your Void form does. I pop PI then immediately Void Torrent and then SF when 12 seconds is left on PI. Is this optimal?

[davesignal]

Noob question;

Do you always enter void form regardless of how high lingering insanity has stacked at?

I think the answer is "yes, unless it's Xavius and you have a 100 stack Lingering Insanity".

[Djriff]

That owl mount will look stupid as hell in Shadowform. All the glorious-glowing-LIGHT stuff always looks terrible in Shadowform.

I'm still at the con but wanted to address this. It got confirmed last night that certain specs will have different colored mounts. So shadow gets a shadow version. Same thing for mages, they showed the arcane one in the video.

As for the artifact. I want something on the level of that awesome guardian Druid.

[Deleted]

The quick version on how to do it properly is refresh dots. Enter voidform. Void torrent immediately. Disperse. Void bolt >>> rotation (theres plenty of places that explain it) Most important thing by far is to void torrent on CD. Nothing should delay this.

Good luck.

Why should he Disperse right after VoiT now? It no longer builds stacks. I save it for +85 stacks area if I make mistake with SW charges and/or wait for something to come off/buy time till 3rd VoIT.

[Gorsameth]

Why should he Disperse right after VoiT now? It no longer builds stacks. I save it for +85 stacks area if I make mistake with SW charges and/or wait for something to come off/buy time till 3rd VoIT.

It buys time for Void Torrent to cool down.
meaning your at lower drain when you torrent the 2nd time. meaning you have more haste per drain on your way to the 3e torrent which = more time in voidform and makes it easier to hit that 3e one.