Anyone know the number differential betwen the races for Spriest? Whats best

[kamelpewpew]

Like the title says does anyone know the number differential betwen Gnome, BE and troll race for Shadow priest?

[anon5123]

Someone simmed the results, and the difference in DPS between races was very minor. (less than 1%)

Play whatever you want.

[Xirrohon]

Like the title says does anyone know the number differential betwen Gnome, BE and troll race for Shadow priest?

Sims with maxed artifact and 880 ilvl gear.

Goblin (377,375 dps): http://beta.askmrrobot.com/wow/simul...085de863e31857

Blood Elf (374,826 dps): http://beta.askmrrobot.com/wow/simul...89af4eb26fbffb

Troll (373,350 dps): http://beta.askmrrobot.com/wow/simul...a2c2e1472f0d76

Tauren (372,379 dps): http://beta.askmrrobot.com/wow/simul...58b93cc8790899

Pandaren (372,308 dps): http://beta.askmrrobot.com/wow/simul...74c37f94e9734f

Undead (372,285 dps): http://beta.askmrrobot.com/wow/simul...4782fab940a105

Makes sense. 1% haste always is pretty strong. 15% haste 5% of the fight isn't particularly amazing for a class that is all about sustained damage. Goblin also has the added utility of rocket jump which is pretty nice tbh.

Blood elf might be inflated by the part that I ran it with AS since it's also running STM.

Top to bottom (undead vs goblin) is just over a 1% difference at this. Up to you if that's worth playing a race that may not look incredibly appealing to you.

EDIT: Just to sate my curiosity, I ran a set with lower gear (around 840 ilvl and no set bonuses). Top to bottom:

Goblin (226,620 dps): http://beta.askmrrobot.com/wow/simul...0b768735505264

Pandaren (226,193 dps): http://beta.askmrrobot.com/wow/simul...cb25b77295b300

Troll (225,689 dps): http://beta.askmrrobot.com/wow/simul...73490f086ef7d9

Undead (224,972 dps): http://beta.askmrrobot.com/wow/simul...9d4e389bcc476b

Blood elf (224,501 dps): http://beta.askmrrobot.com/wow/simul...e031a3bf5a1ce8

Tauren (224,230 dps): http://beta.askmrrobot.com/wow/simul...c2b1befe109cff

Unsurprisingly, pandaren does better in lower gear levels when you have the best food (which appears to be 375 stats). You can also see how goblin will end up scaling harder than the others just due to what 1% haste does to our spec.

Also on the topic of troll racial. 15% haste 5.5% of the fight is only about 0.83% haste effectively. Since we're not a burst class like arcane, it makes sense that a static buff giving more effective stats over the entire fight, will be better than a burst buff.

How about goblins, they have a flat 1 % increase in haste, isn't that better?

Yes. Flat stats are typically better now since they nerfed the on-use racials after MoP. doesn't help that they also removed snapshotting.

This was from the SPreist Legion guide. Goblin is best but like the other poster said, its all very close.

[GT4]

Sims with maxed artifact and 880 ilvl gear.

Goblin (377,375 dps): http://beta.askmrrobot.com/wow/simul...085de863e31857

Blood Elf (374,826 dps): http://beta.askmrrobot.com/wow/simul...89af4eb26fbffb

Troll (373,350 dps): http://beta.askmrrobot.com/wow/simul...a2c2e1472f0d76

Tauren (372,379 dps): http://beta.askmrrobot.com/wow/simul...58b93cc8790899

Pandaren (372,308 dps): http://beta.askmrrobot.com/wow/simul...74c37f94e9734f

Undead (372,285 dps): http://beta.askmrrobot.com/wow/simul...4782fab940a105

Makes sense. 1% haste always is pretty strong. 15% haste 5% of the fight isn't particularly amazing for a class that is all about sustained damage. Goblin also has the added utility of rocket jump which is pretty nice tbh.

Blood elf might be inflated by the part that I ran it with AS since it's also running STM.

Top to bottom (undead vs goblin) is just over a 1% difference at this. Up to you if that's worth playing a race that may not look incredibly appealing to you.

EDIT: Just to sate my curiosity, I ran a set with lower gear (around 840 ilvl and no set bonuses). Top to bottom:

Goblin (226,620 dps): http://beta.askmrrobot.com/wow/simul...0b768735505264

Pandaren (226,193 dps): http://beta.askmrrobot.com/wow/simul...cb25b77295b300

Troll (225,689 dps): http://beta.askmrrobot.com/wow/simul...73490f086ef7d9

Undead (224,972 dps): http://beta.askmrrobot.com/wow/simul...9d4e389bcc476b

Blood elf (224,501 dps): http://beta.askmrrobot.com/wow/simul...e031a3bf5a1ce8

Tauren (224,230 dps): http://beta.askmrrobot.com/wow/simul...c2b1befe109cff

Unsurprisingly, pandaren does better in lower gear levels when you have the best food (which appears to be 375 stats). You can also see how goblin will end up scaling harder than the others just due to what 1% haste does to our spec.

Also on the topic of troll racial. 15% haste 5.5% of the fight is only about 0.83% haste effectively. Since we're not a burst class like arcane, it makes sense that a static buff giving more effective stats over the entire fight, will be better than a burst buff.



Yes. Flat stats are typically better now since they nerfed the on-use racials after MoP. doesn't help that they also removed snapshotting.

This was from the SPreist Legion guide. Goblin is best but like the other poster said, its all very close.

All your links are broken for me.

I highly doubt Troll is using Berserking properly here. For the StM profile with proper Berserking usage, Simc puts troll conveniently (>1% difference) above all other options as is to be expected since it directly affects your StM uptime. For the non-StM profile, proper Berserking usage still puts Troll ~0.3% above Goblin.

Edit: Did a quick check and the default AskMrRobot rotation indeed uses Berserking as soon as Void Form is popped, aka completely wastes it.

[britishbubba]

All your links are broken for me.

I highly doubt Troll is using Berserking properly here. For the StM profile with proper Berserking usage, Simc puts troll conveniently above all other options as is to be expected since it directly affects your StM uptime.

The links are broken because he just copied my post from the guide which doesn't have the full links since it shortens them when posted.

http://www.mmo-champion.com/threads/...2#post41747262

All the links should work there. And iirc, I had it just using berserking on CD as when I simmed using berserking "properly" (using it at the end of STM instead of the beginning since you only get one use) it was barely a dps gain for just that specific void form.

EDIT: If I wasn't lazy, I'd run it with using troll racial during your last void torrent btw. *should* give the largest gain for troll but I imagine it still only makes it even with goblin once the set bonuses are in since the 2pc scales so well with flat haste for an entire fight's duration.

These were very quickly done sims. If you want 100% definitive proof of what racial is best, you'd have to ask twintop. He probably already has those sims sitting around someplace.

-snip-

In the future, you're better off just linking directly to the post so you can avoid links not breaking. Also allows people to more easily find any follow up discussion related to the post you're quoting.

[GT4]

The links are broken because he just copied my post from the guide which doesn't have the full links since it shortens them when posted.

http://www.mmo-champion.com/threads/...2#post41747262

All the links should work there. And iirc, I had it just using berserking on CD as when I simmed using berserking "properly" (using it at the end of STM instead of the beginning since you only get one use) it was barely a dps gain for just that specific void form.

EDIT: If I wasn't lazy, I'd run it with using troll racial during your last void torrent btw. *should* give the largest gain for troll but I imagine it still only makes it even with goblin once the set bonuses are in since the 2pc scales so well with flat haste for an entire fight's duration.

These were very quickly done sims. If you want 100% definitive proof of what racial is best, you'd have to ask twintop. He probably already has those sims sitting around someplace.

I already gave you the relative numbers that come from the current simc version (pulled from github). I don't need to ask anybody for that. Twintop's default profiles (which are supposed to be optimized given certain conditions) are also exclusively using troll.

If you don't believe me, I can just as well upload the results.

[britishbubba]

I already gave you the results that come from the current simc version (pulled from github). I don't need to ask anybody for that. Twintop's default profiles (which are supposed to be optimized given certain conditions) are also exclusively using troll.

If you don't believe me, I can just as well upload the results.

Twintop's default profiles last I checked are using troll and specifically not using their racial so that they ignore racial bias.

EDIT: Twintop's post from H2P discussing his sims. Second point for restrictions:

https://howtopriest.com/viewtopic.php?f=75&t=8640

"Troll for race, but NO berserking used (as a baseline).

[GT4]

Twintop's default profiles last I checked are using troll and specifically not using their racial so that they ignore racial bias.

EDIT: Twintop's post from H2P discussing his sims. Second point for restrictions:

https://howtopriest.com/viewtopic.php?f=75&t=8640

"Troll for race, but NO berserking used (as a baseline).

Those are the profiles he uses to compare different talents, trinkets, etc. as not to skew results with specific racials.

The default profiles which maximize DPS (for cross-class comparison) use Berserking and have it conveniently win over other racials. In case you're interested when Berserking is used:

(Note: actions.vf is the rotation during void form outside of StM)
actions.vf+=/berserking,if=buff.voidform.stack>=10&buff.insanity_drain_stacks.stack<=20&!talent.surrend er_to_madness.enabled
actions.vf+=/berserking,if=buff.voidform.stack>=10&talent.surrender_to_madness.enabled&target.time_to_d ie>variable.s2mcheck-(buff.insanity_drain_stacks.stack)+70

(Note: actions.s2m is the rotation during void form inside of StM)
actions.s2m+=/berserking,if=buff.voidform.stack>=90

[britishbubba]

Those are the profiles he uses to compare different talents, trinkets, etc. as not to skew results with specific racials.

The default profiles which maximize DPS (for cross-class comparison) use Berserking and have it conveniently win over other racials. In case you're interested when Berserking is used:

(Note: actions.vf is the rotation during void form outside of StM)
actions.vf+=/berserking,if=buff.voidform.stack>=10&buff.insanity_drain_stacks.stack<=20&!talent.surrend er_to_madness.enabled
actions.vf+=/berserking,if=buff.voidform.stack>=10&talent.surrender_to_madness.enabled&target.time_to_d ie>variable.s2mcheck-(buff.insanity_drain_stacks.stack)+70

(Note: actions.s2m is the rotation during void form inside of StM)
actions.s2m+=/berserking,if=buff.voidform.stack>=90

Fair enough. I'd still be very surprised if that holds true in the case of multi target since the 2pc will scale well with the 1% haste from goblin. Can you link to the results you're referencing?

[GT4]

Fair enough. I'd still be very surprised if that holds true in the case of multi target since the 2pc will scale well with the 1% haste from goblin. Can you link to the results you're referencing?

Gonna take a few minutes. gonna rerun them in a more easily readable format and upload them.

Edit: It seems to depend on the fight duration (possibly bloodlust usage?). My initial simulations were using a custom duration that I had used for my own tests. When reverting to the default 450 +-20%, troll is now only marginally ahead for S2M and goblin now wins for non-S2M. I've had some work come up in the mean time so I'll report back with some uploaded reports after checking what's going on here (possibly adjusting the APL) and doing some simulations with T19P gear as well.

Edit2: Found the issue. Berserking was used suboptimally during the first bloodlust. I'll adjust the APL, rerun the simulations and upload them afterwards.

Edit3: I'm stupid and accidentally ran the simulations for T19P instead of T19M in the S2M case. Rerunning them again. As of now, troll is winning in all cases by the way.

Edit4: Finally done... for now.
Guess I have two results:
1) Troll is the best for all profiles when Berserking is used properly.
2) My simple APL adjustment helped a lot when playing without S2M but seemed to have negative consequences for the S2M case so you'll probably want to look into S2M Berserking usage again. At worst, you can just return to the old APL for S2M Berserking usage where troll is already significantly ahead of other races.
All races (adjusted APL for trolls): http://hidden.bplaced.net/PriestSims/SP_HordeRaces.html
Troll vs. Goblin T19M_S2M (where goblin was winning with adjusted troll APL) with adjustment reverted: http://hidden.bplaced.net/PriestSims/SP_S2M_OldAPL.html

The latter is what I was referencing to earlier (with other races also included but not as easily readable).

TL;DR: Troll wins in all scenarios if you use Berserking ideally (for single target as of gear available in the first raid tier - can't say anything for any other case yet).

[britishbubba]

Gonna take a few minutes. gonna rerun them in a more easily readable format and upload them.

Edit: It seems to depend on the fight duration (possibly bloodlust usage?). My initial simulations were using a custom duration that I had used for my own tests. When reverting to the default 450 +-20%, troll is now only marginally ahead for S2M and goblin now wins for non-S2M. I've had some work come up in the mean time so I'll report back with some uploaded reports after checking what's going on here (possibly adjusting the APL) and doing some simulations with T19P gear as well.

Edit2: Found the issue. Berserking was used suboptimally during the first bloodlust. I'll adjust the APL, rerun the simulations and upload them afterwards.

Edit3: I'm stupid and accidentally ran the simulations for T19P instead of T19M in the S2M case. Rerunning them again. As of now, troll is winning in all cases by the way.

Edit4: Finally done... for now.
Guess I have two results:
1) Troll is the best for all profiles when Berserking is used properly.
2) My simple APL adjustment helped a lot when playing without S2M but seemed to have negative consequences for the S2M case so you'll probably want to look into S2M Berserking usage again. At worst, you can just return to the old APL for S2M Berserking usage where troll is already significantly ahead of other races.
All races (adjusted APL for trolls): http://hidden.bplaced.net/PriestSims/SP_HordeRaces.html
Troll vs. Goblin T19M_S2M (where goblin was winning with adjusted troll APL) with adjustment reverted: http://hidden.bplaced.net/PriestSims/SP_S2M_OldAPL.html

The latter is what I was referencing to earlier (with other races also included but not as easily readable).

TL;DR: Troll wins in all scenarios if you use Berserking ideally (for single target as of gear available in the first raid tier - can't say anything for any other case yet).

These sims aren't using tier... Mine were. Goblin should benefit from the 2pc more than troll will since it's just 1% more DoT ticks which is a large amount of insanity over the fight. That's just going to become a larger margin when there's more targets to DoT up. At least it should since goblin averages out to more haste over the fight duration (1% vs 0.84%)

[GT4]

These sims aren't using tier... Mine were. Goblin should benefit from the 2pc more than troll will since it's just 1% more DoT ticks which is a large amount of insanity over the fight. That's just going to become a larger margin when there's more targets to DoT up.

Why would you assume tier sets now when tier sets only come with the third raid at the beginning of 2017?

Not to mention, you can expect Blizzard to rebalance and rework tier sets before their release as many (all?) haven't been updated since their initial beta implementation and some even improve abilities that no longer exist (or improve ones that have since been reworked, making the bonuses either useless or way too strong).

Additionally, you posted a second set that wasn't using the tier bonus and even there you got troll as only third. Since that can be improved a lot by better Berserking usage (see the T19P sims I posted), it should be obvious the same can be done for your other simulations which were using the same sub-optimal Berserking usage, improving troll's DPS to some extent.

Then you're making assumptions without any basis. E.g. "Goblin should benefit from the 2pc more than troll will since it's just 1% more DoT ticks which is a large amount of insanity over the fight." isn't a logical conclusion at all since you could use the same argument without the 2pc (which you basically did in your first post when comparing troll's averaged out haste to that of goblin) and you'd be wrong (see: the sims I posted). The exact same would be true for this conclusion: "That's just going to become a larger margin when there's more targets to DoT up." All of those conclusions don't make sense because shadows don't do even damage all fight long and insanity isn't always worth the same. At the start of void form, for example, your passive haste does nothing for insanity generation because you frequently overcap anyway. On the other hand, troll can be used explicitly to stay longer in void form while at the same time multiplying the high damage you have at high stacks, exponentially scaling your DPS. That's just in addition to being able to squeeze out more S2M time by using Berserking once you cannot keep up with the drain any more (and once again multiplying the extreme DPS you're doing at the end of S2M).

Lastly, from what I've seen the priority list of AskMrRobot is highly unoptimized compared to that used in simc which is only logical given that each simc module is maintained by at least one person specifically responsible for that class. That might not make a big difference for other, more simple classes but for SPs this has a huge impact.

This isn't supposed to be an attack on you but I just don't see how you can draw the conclusions you did in your first post. Basically all you showed so far is that troll is worse than goblin and blood elf if you don't use berserking correctly while not playing your rotation optimally either. Everything else is just speculation on your side (of which some has already been debunked using simulations that have proper berserking usage).

Edit: I just wanted to add that your calculation regarding troll's average haste is also practically wrong. 15%*10/180 is correct for an infinitely long fight with 100% damage uptime. For a real fight, you'd be looking at a way higher berserking uptime if you were to use it on cooldown (e.g. extreme examples being 100% for a 10 seconds fight or 20/190 for a 3:10 fight), resulting in way higher average haste as well.