Legion Theorycrafting [7.3, All specs]

[Azortharion]

Are you just selecting two talents here? Instead of a full combination of dps talents? If the latter, what other talents were you taking?

Stomp, Bestial Fury, Crows.

- - - Updated - - -

When I said the APL depends on the talents, I was saying that how play changes based on talent balance. If talents are likely to wildly change (such that simming one vs another isn't worth the time), why bother coming up with a priority list that might be wrong when the numbers pass hits?



Or did I misinterpret this? I assumed you were saying that because talents and numbers are in flux, it's not a good use of time to sim them.

Here's how theorycrafting before an expansion/patch works:

Implement the talents into SimCraft.
Make APL for the spec as a basis.
Make/add APL components that will maximize how each talent is used. This isn't always straightforward (but it mostly is). Do this a lot.
Do your comparison sims, or whatever.

We're at like, step 2/3 of that process.

Balance changes to talents rarely change how their APL is implemented. It sure does not warrant skipping that step altogether.

[Effinhunter]

Our process for getting the perfect APL is going to be a bit different because of artifacts, relics, and traits adding another layer of complication. Speaking of which.

BM Artifact Relics

Gear level: 825, 858 weapon
Simcraft: 450s, 20% dev, 10k iterations, single target, source


Essentially, you'll want to overload Pack Leader or Jaws of Thunder, both of which affect Kill Command. Another option would be Furious Swipes for Beast Cleave, but since this was a single target sim, you're not going to see any benefit in these sims. I'll do AoE sims at a later date. Still need to look into SV.

[silverstarzs]

[MENTION=1027319]

@silverstarzs The simulator is actually very accurate.
. . . . bug free
. . . .APLs (rotations) are good, but not perfect.

Understood - thanks - nice tool.

Here is BM vs MM comparison with 875 gear sets and 900 weapons. As stated above neither APL will be perfect . . . but the size of the gap is interesting.

[Effinhunter]

That's not that surprising, really. Iskar has tightly clumped adds and BM is absolutely absurd on AoE if adds are clumped (as it has always been with beastcleave).

SV Artifact Relics

Gear level: 825, 858 weapon
Simcraft: 450s, 20% dev, 10k iterations, single target, source


RC is showing no benefit because Raptor Strike is not being used in the current APL on single target (other things use up resources before you get to this). Sharpened Fang augments Mongoose Bite, a core rotation ability, and Explosive Force increases damage from Explosive Trap. That's why those two are so powerful in single target.

First post should be updated to here

[silverstarzs]

Single target (Zakuun) for comparison - specs essentially identical

[Shanoa]

Why don't you make this comparison with SV also? Or it isn't working right now? I'm really in love with SV (i love melee) so i'd enjoy some more info and comparison like this one above with SV too

[Verko]

Is the AOE gap really that bad? I enjoy MM by far the most, would it be still viable for most fight or are guilds going to want people to play BM for the AOE?

On current Pre-patch MM has great AOE, i'm assuming this doesn't continue into legion then? Or is BM just really good at a niche AOE situation?

[Aldare]

Is the AOE gap really that bad? I enjoy MM by far the most, would it be still viable for most fight or are guilds going to want people to play BM for the AOE?

On current Pre-patch MM has great AOE, i'm assuming this doesn't continue into legion then? Or is BM just really good at a niche AOE situation?

MM still has good AoE in Legion, it's just that BM has insanely good AoE due to Beast Cleave being doubled by Hati when the artifact is levelled up.

[eathdemon]

I know this is a haerd question to ask, but how does bm scale? given the artifact weapon switching specs will be hard. if bm starts out better, but cant keep up it doesnt matter. ps, if bm if good, I would rather go bm I am just afraid of scaling.

[Effinhunter]

I'll get to those kinds of sims soon, but I need to take a more serious look at the BM and SV APLs first.

[silverstarzs]

how does bm scale? .

Crude analysis here.

Comparing performance at 810 gear with 3 crappy relics with 875 with 3 good relics.

BM: ST (Zakuum) 188% multi-target (Iskar) 254%
MM: ST (Zakuum) 142% multi-target (Iskar) 202%

That makes it look like BM scales better than MM.

Would also be interesting to see for other classes - to know how hunter specs scale relative to other classes.

[Freerunz]

That makes it look like BM scales better than MM.

Can you show your APL's

[silverstarzs]

^^ default AMR

[Freerunz]

^^ default AMR

Well, than this numbers is nothing, so stop misinformating people.

[silverstarzs]

explain in detail - why default APL's produce not even directional correct results? The results so far seem consistent with experience.
Perhaps you can help them improve their APL's You can read/comment on their wiki, where various people are working on the APL's.
What do your preferred APL's produce as a "scaling" result?

Scaling calculations are somewhat less sensitive to many 'apl improvement opportunities which cause DPS increases' because they are washed out of both the numerator and denominator. But they are more sensitive to a few particular types.

[bouchbagfett]

They had the AMR team on the Hunting Party Podcast last week and Swol made it pretty apparent he is trying his best to have good APL's and make the sim tool as accurate as possible. He was very thoughtful and it was easy to see he is actually playing a hunter. I wouldn't reject AMR sim based on APL's. I think the issue is more that tuning isn't done yet so it is hard to draw conclusions from a product that isn't going to be the same as the final one.

[silverstarzs]

^^ also to be fair - in addition to Bliz re-balancing (which you can never wait to be 'finished' because it is never finished) AMR is a new 'beta' tool. Which means that not all features are yet implemented and the users (eg me for example) probably dont yet know how to use it to best advantage. But people (including me) are curious were we stand right now and it probably gives directionally useful results to coarse questions.

[Zoopercat]

@Freerunz - if you have something specific to report about the AMR tool, please share. All bugs are fixed in a day Also, since we have all of the mechanics open sourced on a wiki, a lot of people have been checking them out to make sure they are implemented correctly (in addition to us doing our own QA too). Here's an example wiki page.

The mechanics on the AMR sim are very solid. We've done 3 passes ourselves now, and have several experts who've been playing with it, and reporting any bugs found. It's really, really thoroughly tested

As for APLs, those are always in development. But they are quite good. In fact, our survival APL was better than the SimC one, so we shared our approach. And like all good TC, we put our heads together (us and some simc folks) to make an even better combined one. Just one example of how thorough we are

Hope some of that info helps. Again, if you found a bug, let me know. Otherwise, the hunter stuff currently has no open bugs. (IT doesn't mean there are no bugs though).

To run your own sims, go to beta.askmrrobot.com. For those new to it, here are some resources:
-- 9 min intro video, covers a little of everything.
-- How to work with APLs, 5 min video
-- 25-min video on comparing 2 rotations to make them better
-- Changing gear and artifacts (it's so simple the demo in the video is 30 seconds long)

And here's an example of some of the things we investigate on a per-class basis. If you have specific hunter questions, I can set up something fun for you all as well.

[silverstarzs]

^^ Zoo, something I would find helpful in the simulations is if it could provide something like a BiS gear set and BiS talents for a given iLvL/boss. I did not go to the effort to pick best budget gear, nor to modify the currently understood 'standard' talent selection. I think you are working on that, but that is probably where my simulations are weak(est).

Effinhunter - sorry to butt into your thread - Probably we need an AMR thread - because you were doing a terrific community service with your work and we should not distract from it. I will butt out, and please continue

[Azortharion]

You can't simulate the best talents for a fight, are you serious.