Can I ask You about... Ask mr. Robot?

[Deleted]

Hello!
I once tried using simcraft, but I didn't read any guides, so I ended up feeling "Duh, I'm too stupid for that".
I am not a masochist and I didn't want to try every possible built and item on a dummy, so I was using other utilities to buff up my gameplay and I got to like Ask mr. Robot. It was quite accurate if it's going about items, but I never tried to make a simulations available there to check my talents.
Until today. And what I discovered? That on my char - http://eu.battle.net/wow/en/characte.../Pombar/simple - the best build is:
BF-LS-Temp-CS...
Is that even possible? Unfortunately I won't be able to test it in game for a about a week.
Is my gear that much "specific"? Or just simulations on Amr.R are shit?
Second and third results, are with same builds but with:
2nd - HS (instead of LS)
3rd - AS

Ask Mr. Robot is quite covenient so I would like to know, if I can count on that.

[killor318]

Gear upgrades parse pretty well. However, trinkets I would look into other resources.

[Deleted]

Well, I didn't have many problems with gear, but what I am asking about is mostly the accuracy of the simulations, considering different talent builds.
In every sim, BF-LS-T-CS or BF-HS-T-CS has best results which seems very strange for my. Because according to what others are saying, those two builds, should be at the very bottom of the chart.

[Zoopercat]

@Kashalot and @killor318: reguarding trinkets: we actually rank all items with procs directly from simulations for Legion, making it a very customizable and accurate way to rank them A big improvement over WoD.

As for the simulations: they are very accurate* (keeping in mind that we might be fixing bugs or missed changes for a couple weeks after a big patch like this). Why can I make the accuracy claim - I'm from AMR afterall, of course I would say that, right? Well here's some info on our process:

1. We check all spell data to in-game tests. We do this because often in-game tooltips are wrong or changes are undocumented
2. We check spell data against actual logs from WCL. This is REALLY handy because I can match someone's gear and character setup from a log to our simulator. Then I can look and see that maybe every single spell is doing the right amount of regular and crit damage... except one. Maybe that one is 25% too low. That let's me take a look at the spell and see if there's something that should be buffing it, but isn't.
3. The spell wiki makes all of the data open source, and a lot of people view those pages, verifying our information (or pinging us if something is wrong).

TLDR: We veryify spell data from multiple sources, do dedicated testing, and the open source wiki is an accuracy source as well.

[Creekz]

Some spells didn't have the right priority in the rotations, but it should be updated soon as you can see here: http://forums.askmrrobot.com/index.php?topic=12471.105

I made some updates today that make better use of Overcharge and Fury of Air. When those get posted later, you should see that talent setup simulate higher than before.

I have the incoming hotfixes implemented as well. In the absence of specific legendary items or set bonuses, I'm still seeing boulderfist, hailstorm, tempest, crashing storm, landslide do the most single target DPS. The HH/AS/OC/FoA build is close in damage, but not quite as high.

[Deleted]

Thanks for the answers guys!
It's great, that ppl are keeping an eye on all those functions, even those which are free.
But well, I still need to check those rotations in game.
It seems pretty weird that old build is still best, even after huge nerfs and significant buffs on other talents.
I'm wondering if that really is true, or the simulated rotation still needs some fixes.

[Creekz]

Thanks for the answers guys!
It's great, that ppl are keeping an eye on all those functions, even those which are free.
But well, I still need to check those rotations in game.
It seems pretty weird that old build is still best, even after huge nerfs and significant buffs on other talents.
I'm wondering if that really is true, or the simulated rotation still needs some fixes.

Some of the old talents got buffed http://www.mmo-champion.com/threads/...Hotfix-Changes

[fathom81]

@Kashalot and @killor318: reguarding trinkets: we actually rank all items with procs directly from simulations for Legion, making it a very customizable and accurate way to rank them A big improvement over WoD.

As for the simulations: they are very accurate* (keeping in mind that we might be fixing bugs or missed changes for a couple weeks after a big patch like this). Why can I make the accuracy claim - I'm from AMR afterall, of course I would say that, right? Well here's some info on our process:

1. We check all spell data to in-game tests. We do this because often in-game tooltips are wrong or changes are undocumented
2. We check spell data against actual logs from WCL. This is REALLY handy because I can match someone's gear and character setup from a log to our simulator. Then I can look and see that maybe every single spell is doing the right amount of regular and crit damage... except one. Maybe that one is 25% too low. That let's me take a look at the spell and see if there's something that should be buffing it, but isn't.
3. The spell wiki makes all of the data open source, and a lot of people view those pages, verifying our information (or pinging us if something is wrong).

TLDR: We veryify spell data from multiple sources, do dedicated testing, and the open source wiki is an accuracy source as well.

The m+ stat weight options is based on +10 and random afix?
Im curious to why the stat weight difer so much from icyveins or noxxic, any reason why?

[Darleth]

Well, I didn't have many problems with gear, but what I am asking about is mostly the accuracy of the simulations, considering different talent builds.
In every sim, BF-LS-T-CS or BF-HS-T-CS has best results which seems very strange for my. Because according to what others are saying, those two builds, should be at the very bottom of the chart.

You're misreading the results. BF_AS_T_CS will be the go to talent with 4 piece set bonus. Before that, Tempest for example is only really good on cleavefights.
HS shouldn't be ever good currently; not even with both Akainus and Twisting Nether its a pretty bad talent and actually somewhat works against our 4 piece, as you use ressources on other things than lava lash, which has a good chance of proccing stormbringer.
HH/WS_AS_FoA_OC is the build to go currently for pure singletarget fights.

The m+ stat weight options is based on +10 and random afix?
Im curious to why the stat weight difer so much from icyveins or noxxic, any reason why?

Noxxic... seriously?? Nobody takes NOXXIC serious. Dont go there.
To why stat weights differ "so much": That is because of certain gearsetups. Especially trinkets play a huge role into that. Before the patch, Haste was better than mastery because of Spontaneous Appendages. I wouldn't be too surprised if certain trinkets (and probably even legendaries) are still taken into consideration for this. Also, depending on your gearlevel and current stat weights, things might get pushed around a lot. Crit currently is better than versatility, because we just have less of it than before thanks to BF not granting the extra crit chance anymore and us not using it anymore of course, and also because of the overall crit-reduce.

If anything: sim your stat weights yourself to be a 100% sure about what is an upgrade and what not. I dont know exactly how AMR simulates it when considering certain items; you gotta ask Zoopercat or Swol on this.

[Deleted]

Some of the old talents got buffed http://www.mmo-champion.com/threads/...Hotfix-Changes

Yup, I noticed that, but still, AMR's sims are using pure 7.1.5.

You're misreading the results. BF_AS_T_CS will be the go to talent with 4 piece set bonus. Before that, Tempest for example is only really good on cleavefights.
HS shouldn't be ever good currently; not even with both Akainus and Twisting Nether its a pretty bad talent and actually somewhat works against our 4 piece, as you use ressources on other things than lava lash, which has a good chance of proccing stormbringer.
HH/WS_AS_FoA_OC is the build to go currently for pure singletarget fights.

Well, I'm not misreading anything, because I'm doing sims with my current gear. And as You expect, I don't have 4p set bonus. The only thing that I am changing in that sims, are talents. And I ended up, with BF-HS/LS-T-CS being best for me. Maybe I will be able to check that today, so I'll report my in-game results.

Edit:
I've changed my gear to best now availble (with head and trinket as legedaries) and sims are showing, that "standard" (WS/HH-AS/LS-OC-FoA) build is on top, but 4th result, losing only 3k dps, is BF-HS-T-CS. I'm curious, if AMR's devs nerfed BF, HS and CS properly.

[Darleth]

Yup, I noticed that, but still, AMR's sims are using pure 7.1.5.



Well, I'm not misreading anything, because I'm doing sims with my current gear. And as You expect, I don't have 4p set bonus. The only thing that I am changing in that sims, are talents. And I ended up, with BF-HS/LS-T-CS being best for me. Maybe I will be able to check that today, so I'll report my in-game results.

Edit:
I've changed my gear to best now availble (with head and trinket as legedaries) and sims are showing, that "standard" (WS/HH-AS/LS-OC-FoA) build is on top, but 4th result, losing only 3k dps, is BF-HS-T-CS. I'm curious, if AMR's devs nerfed BF, HS and CS properly.

Are you simming with simcraft? If so, you probably got the wrong client. The downloadable version on simulationcraft.org is still on 7.1.
http://downloads.simulationcraft.org/?C=M;O=D you can download the correct version here. I dont know how accurate AMR currently is, as I haven't used it yet. Also, the shaman discord has some good information about current sims.

I can guarantee you, that most stuff wont be accurate ingame.

[Zoopercat]

@fathom81 - I don't think the weights on Icy Veins have a mythic+ version, or do they and I missed it? For DPS, the mythic+ level doesn't matter (but it does for tanks). As for stat weight differences:

1. Stat weights from AMR's simulator look at 1000s of data points (instead of just 2 per stat like stats generated from SimC)
2. AMR also takes stat interactions into consideration (SimC does them in isolation, missing the interactions. However, if you look at SimC reforge plots, that will give you info on how the stats interact)
3. We also have our machine learning method, which is better than stat weights (but, it's in beta, so it's better* with an asterisk until we have all of the kinks worked out).
@Kashalot - are you referring to the changes that will be going live on Tuesday?

[Deleted]

Are you simming with simcraft? If so, you probably got the wrong client. The downloadable version on simulationcraft.org is still on 7.1.
http://downloads.simulationcraft.org/?C=M;O=D you can download the correct version here. I dont know how accurate AMR currently is, as I haven't used it yet. Also, the shaman discord has some good information about current sims.

I can guarantee you, that most stuff wont be accurate ingame.

Nah, I was only using simulations on AMR, but I will surely try simcraft in future.

@Kashalot - are you referring to the changes that will be going live on Tuesday?

I am aware of those changes, but surely they don't affect current AMR's sims right?

I've checked some build in-game and unfortunately AMR got it wrong. Don't know what is its mistake, perhaps rotation is wrong? Because That build suggested by AMR's sims (BF-HS-T-CS) turned up to be 30-40k worse than HH-AS-T-FoA. So it is as I expected. HS is unplayable, BF and CS might be useful in some situations. But I won't be checking it very hard because of the next Tuesday's hotfix.

[Creekz]

Nah, I was only using simulations on AMR, but I will surely try simcraft in future.



I am aware of those changes, but surely they don't affect current AMR's sims right?

I've checked some build in-game and unfortunately AMR got it wrong. Don't know what is its mistake, perhaps rotation is wrong? Because That build suggested by AMR's sims (BF-HS-T-CS) turned up to be 30-40k worse than HH-AS-T-FoA. So it is as I expected. HS is unplayable, BF and CS might be useful in some situations. But I won't be checking it very hard because of the next Tuesday's hotfix.

Actually

I have the incoming hotfixes implemented as well. In the absence of specific legendary items or set bonuses, I'm still seeing boulderfist, hailstorm, tempest, crashing storm, landslide do the most single target DPS. The HH/AS/OC/FoA build is close in damage, but not quite as high.

http://forums.askmrrobot.com/index.php?topic=12471.105

[Deleted]

I've read that too, but in my case (and I even have legendary ring which should improve Hailstorm's performance) that is simply not true. In-game BF-HS-T-CS is 30-40k worse than recommended build.

[Creekz]

I've read that too, but in my case (and I even have legendary ring which should improve Hailstorm's performance) that is simply not true. In-game BF-HS-T-CS is 30-40k worse than recommended build.

Well, the hotfixes aren't live yet. AMR's sim is using the hotfixes that will come tomorrow for US and wednesday on EU.

I guess it could also depend on your current trinkets & stats. With my current gear BF-HS-T-CS sims the highest, with the recommended BIS items I would sim higher with HH-AS-OC-FoA.

[Deleted]

I tried simming my character on Ask Mr Robot today

Surprisingly with bracers and eye ring, I had the best single target result using talents Hothand + Hailstorm+ Tempest+ Fury . Even though people say hailstorm not enough with legendaries to outperform AS, I had the highest sim from this combination? Any ideas why?

[Jaese]

I tried simming my character on Ask Mr Robot today

Surprisingly with bracers and eye ring, I had the best single target result using talents Hothand + Hailstorm+ Tempest+ Fury . Even though people say hailstorm not enough with legendaries to outperform AS, I had the highest sim from this combination? Any ideas why?

Because the hotfix is buffing HS? I mean, the answer is in this thread.

[Deleted]

Because the hotfix is buffing HS? I mean, the answer is in this thread.

My sim was using current build not after hotfix. I had the best single target result using talents Hothand + Hailstorm+ Tempest+ Fury . Even though people say hailstorm not enough with legendaries to outperform AS

[Deleted]

Well, the hotfixes aren't live yet. AMR's sim is using the hotfixes that will come tomorrow for US and wednesday on EU.

I guess it could also depend on your current trinkets & stats. With my current gear BF-HS-T-CS sims the highest, with the recommended BIS items I would sim higher with HH-AS-OC-FoA.

If that's true, I mean, if AMR's sim is using "nighthold's" hotfixes then it explains everything. But are You sure, that sims aren't on current, pure 7.1.5 build?