Restoration Druid 7.3

[Valks 2,375]

On 3/2/2017 at 5:43 PM, evanbloom said:

Testing how? If not simulations, then what do you test? In your bullet list example of Mastery prio, what data supports your logical argumentation framework? It looks sound from a theoretical standpoint and all those people using Haste (myself included) must be doing something right, but I've just got this itch that something is not completely rigorous from a statistical point of view.

I'm not the actual writer/reviewer, so I can't provide proof of the actual data/numbers (I just don't have them, basically). But generally, they will take a large scaling of numbers where they do the maths that a sim would. They attribute a stat weight to each stat in increasing numbers of HoTs on the target. As you increase the number of HoTs, the weight on Mastery will increase while the weight on Haste increases. The reason why Mastery becomes more useful in an M+, for example, is because you have a smaller group of people to maintain your HoTs on, so they will all be at the 2-3 HoTs maintained mark. This means Mastery is far closer to Haste. It's always why Mastery is good for tank healing.

Haste is good in raids because you can't blanket every single person with HoTs. I'll try to get the actual numbers for you. You can also check out the WA that Orthios linked, I believe that must have some of the maths in the code.

[Valks 2,375]

9 hours ago, Harry said:

your saying that no mention of druids dispelling and no mention of the spell natures cure an error when transferring a macro, its a fundamental part of being a druid healer. 

None of our healer guides mention any dispelling at all, outside of a macro or two. It's something we can certainly add.

9 hours ago, Harry said:

wrong macro

Given that your first comment was specifically about the Remove Corruption macro, I assumed you were questioning the macro page. We've never included any mention of the dispel and you're the first one to bring it up, at least in my time of watching the comments. We will fix the macro to instead have Nature's Cure (since RC is not a resto spell, hence why I said it was an error when transferring a macro from the old guide to the new one). 

As for adding something about NC, we can do that, of course, since you have asked for it. We can't add things that people want if we've never been told they want it, unfortunately. As for your comment about us not taking pride in it, we certainly do and it's why we have such in-depth sections about healing as a Restoration Druid. We have just simply never had a request (at least in my time here) to add something about dispelling before, so we didn't.

[evanbloom 1]

15 hours ago, Blainie said:

But generally, they will take a large scaling of numbers where they do the maths that a sim would.

Thanks for the explanation! I'm not interested necessarily in the results, what I'm curious about is the procedure. Orthios said earlier that simulations are useless for healers. What you describe here -- basically what's the difference from a simulation then? ”Take a large scaling of numbers where they do the maths” what does that mean in fact? How is that not a simulation? Genuinely trying to understand the differences and the process here :)

[Guest Salvardor]

anyone did some testing with  drape of shame ?

ist it still BiS ?

[Valks 2,375]

On 3/7/2017 at 7:16 AM, evanbloom said:

 Genuinely trying to understand the differences and the process here :)

Super late here, but hopefully you're still interested:

• Damage sims take a very specific encounter type and then apply that to the patchwerk model.
  • This means that the damage numbers are always slightly inflated, due to no movement, perfect conditions etc.
• For healing sims (these did used to exist), you need to create a HUGE variety of simulations in order to actually create the proper conditions.
  • For DPS, it's rather black and white. They are either moving, or not, and they are either hitting one or multiple targets, right?
  • For Healers, there are a multitude of differentials, not limited to:
    • Type of tank
      • Are they someone that takes a huge amount of damage and just outheals it?
      • Do they create their own shields?
      • Do they use AM to stop damage?
      • Do they require chained externals to survive?
    • Type of damage
      • Magical
      • Physical
      • Sustained
      • Burst
      • Multi-hit vs. one-shot
      • Level of mob hitting the tank (boss vs. trash)
    • Types of encounter
      • Raid-wide damage
      • Tank-damage
      • Group soaking
      • Proximity damage
      • Burst damage due to tactics

Essentially, you need a crazy amount of different sims to run to get an answer, and you then need to work from there. The WA that Orthios linked essentially does all of this recording while watching people play in real time, if that makes sense? Rather than going and working out a situation for a trash pull of 3, a trash pull of 6, a magical boss, a physical boss and some group-wide damage, you can simply run the WA in a single dungeon of 20-30 minutes. As more and more people do it, information becomes more accurate and feedback means the WA becomes better, basically.

The maths that people do by hand is often easier than trying to write multiple different sims, essentially, where you need the correct rotational input etc. - instead you can use raw healing numbers and calculate it yourself.

Hope this helps!

[Valks 2,375]

On 3/12/2017 at 11:47 AM, Guest Salvardor said:

ist it still BiS ?

Will be re-assessed in 7.2 :)

[Valks 2,375]

We are currently in the process of updating our guides in preparation for the release of 7.2 - please understand that we will not be answering questions in this section about "What is better for 7.2?" prior to the patch release, since all the information will be available when the patch goes live. Please be patient and thanks for waiting!

[evanbloom 1]

7 hours ago, Blainie said:

Super late here, but hopefully you're still interested:

• Damage sims take a very specific encounter type and then apply that to the patchwerk model.
  • This means that the damage numbers are always slightly inflated, due to no movement, perfect conditions etc.
• For healing sims (these did used to exist), you need to create a HUGE variety of simulations in order to actually create the proper conditions.
  • For DPS, it's rather black and white. They are either moving, or not, and they are either hitting one or multiple targets, right?
  • For Healers, there are a multitude of differentials, not limited to:
    • Type of tank
      • Are they someone that takes a huge amount of damage and just outheals it?
      • Do they create their own shields?
      • Do they use AM to stop damage?
      • Do they require chained externals to survive?
    • Type of damage
      • Magical
      • Physical
      • Sustained
      • Burst
      • Multi-hit vs. one-shot
      • Level of mob hitting the tank (boss vs. trash)
    • Types of encounter
      • Raid-wide damage
      • Tank-damage
      • Group soaking
      • Proximity damage
      • Burst damage due to tactics

Essentially, you need a crazy amount of different sims to run to get an answer, and you then need to work from there. The WA that Orthios linked essentially does all of this recording while watching people play in real time, if that makes sense? Rather than going and working out a situation for a trash pull of 3, a trash pull of 6, a magical boss, a physical boss and some group-wide damage, you can simply run the WA in a single dungeon of 20-30 minutes. As more and more people do it, information becomes more accurate and feedback means the WA becomes better, basically.

The maths that people do by hand is often easier than trying to write multiple different sims, essentially, where you need the correct rotational input etc. - instead you can use raw healing numbers and calculate it yourself.

Hope this helps!

Thanks, Blainie, it doesn't matter it's late :) So basically when you say Icy Veins carries out testing to recommend stat priorities, you mean IV uses maths done by people instead of computers, did I understand correctly?

When you say for healing you need a HUGE amount of sims,  do you mean it is something that you would need NASA grade inhibitingly expensive processing power to calculate in a fair amount of time? Or what other reason is preventing Icy Veins from writing and running sims for healers for all that variety of situations other than manual maths being easier?

And this brings me back to what Orthios said, that ”sims do not accurately reflect in-game encounters at all”. Surely sims reflect whatever you program them to reflect?

Hope you can find some time to answer these questions while also preparing for the patch :)

[Valks 2,375]

3 hours ago, evanbloom said:

Or what other reason is preventing Icy Veins from writing and running sims for healers for all that variety of situations other than manual maths being easier?

Simply because we don't need to, basically. Through in-game testing, using raw numbers and spreadsheets, etc. you can basically work out exactly the same thing with less issues. It's not the case with DPS specs because of how much easier sims are to make for them.

You'd be putting in a huge amount more work to create the same result, basically. The best example I can give is the spreadsheet that was made by Dreamguard for Holy Paladins - it's a resource that many HPalas use for finding specific stat weights when they don't want to just follow our priority (he isn't an IV writer, but his spreadsheet actually perfectly illustrates my point) - https://docs.google.com/spreadsheets/d/1WTRKEeOWmmKqbUFz8C35rSOp_A93BzgsLyQinFXagTo/edit#gid=1959096800

This is his weights page, where you can see exactly the kind of maths I'm talking about :) (It's not for Resto Druids though, sorry!)

[Orthios 271]

6 hours ago, evanbloom said:

And this brings me back to what Orthios said, that ”sims do not accurately reflect in-game encounters at all”. Surely sims reflect whatever you program them to reflect?

I think you're overestimating what can be done within simulationcraft.  How SimC works is it takes a given APL (Action Priority List) and goes through it to determine what happens at a given time.  You can add raid events to their own APL (which is by default empty) and specify various aspects regarding the event, but there's only so much you can do.  These events, however, are based on time, not boss health percentage.  It is incredibly difficult to accurately model a fight for a multitude of reasons: 1) Group composition can very well dictate how long the fight lasts and who does the damage to adds (if any), 2) Many boss mechanics are based on health %, and the time the boss hits that % can vary widely, even within a single group, 3) Some things simply aren't possible to model in SimC right now (examples being Skorpyron's shockwave and pillars, chronomatic anomaly and elisande's speed up/down mechanics (elisande's in particular since it is a bubble that you move in and out of at will), and krosus's bridge breaking, just to name a few).

[evanbloom 1]

Thanks, Blainie and Orthios, for the answers! So it seems to be one of those situations where humans are better than robots :)

[Valks 2,375]

2 hours ago, evanbloom said:

Thanks, Blainie and Orthios, for the answers! So it seems to be one of those situations where humans are better than robots :)

Glad to see that you've got your answer! :) If you need anything else, you can always ask.

[Guest Zkar]

Grace of the Cenarion Circle as 5th taken, yeh right...

[Guest Rifap]

Shouldnt the 10% heal/sta increase be the first new trait to take? it says #5 currently.

[Guest DennyCrane]

Do not take Grace of the Cenarion Circle first, it is useless.

[Valks 2,375]

On 3/29/2017 at 10:43 AM, Guest Zkar said:

Grace of the Cenarion Circle as 5th taken, yeh right...

Hey! This is being fixed right now as we speak - this was an error when transferring things across between notes. The updated diagram and notes will be up ASAP!

Sorry and thanks for reporting it to us :)

[Valks 2,375]

21 hours ago, Guest Rifap said:

Shouldnt the 10% heal/sta increase be the first new trait to take? it says #5 currently.

See above. Thanks for helping to report it!

[Valks 2,375]

52 minutes ago, Guest DennyCrane said:

Do not take Grace of the Cenarion Circle first, it is useless.

See above - we are actually going to be recommending it as the first pick shortly :)

[Zaflis 0]

I see one error in "No shapeshift form and Travel Form: you fly to an ally's location." in https://www.icy-veins.com/wow/restoration-druid-pve-healing-spec-builds-talents

When you do wild charge in travel form, you actually leap a fair distance, not fly to ally's location. It can be used as a double jump and you can carry a person with you in stag form. It can be used to nullify fall damage by leaping just before hitting ground, or fights like EN spider boss you can leap to the left side web path on the second phase. Basically the talent only works like that in outdoor fights, because normally you can't use travel form at all.

I still see displacer beast as a very good talent that shines in indoor raids but it messes you up a little bit by activating catform automatically.

[Valks 2,375]

On 4/12/2017 at 2:39 PM, Zaflis said:

I see one error in "No shapeshift form and Travel Form: you fly to an ally's location." in https://www.icy-veins.com/wow/restoration-druid-pve-healing-spec-builds-talents

Will get this updated, thanks!

[Guest Arcuriel]

Were the legendary rankings updated for 7.2? The items page still says legends have an item level of 910.Were the legendary rankings updated for 7.2? The items page still says legends have an item level of 910.

[Valks 2,375]

18 hours ago, Guest Arcuriel said:

Were the legendary rankings updated for 7.2? The items page still says legends have an item level of 910.Were the legendary rankings updated for 7.2? The items page still says legends have an item level of 910.

This is just a tooltip error - we have since updated this. Thanks for pointing it out! 

[Guest Taz]

I would assume it's still Up there as far as raiding goes, since crit imo should be higher in raiding setups. Due to the higher uptime of hots on party members in dungeons haste/mastery increase in value and as such the power of drape decreases. 

On 12/3/2017 at 0:47 PM, Guest Salvardor said:

anyone did some testing with  drape of shame ?

ist it still BiS ?

 

[Valks 2,375]

On 4/21/2017 at 10:07 AM, Guest Taz said:

I would assume it's still Up there as far as raiding goes, since crit imo should be higher in raiding setups. Due to the higher uptime of hots on party members in dungeons haste/mastery increase in value and as such the power of drape decreases. 

Pretty much, yeah. Will just depend on how you build and what you use it for.

[Guest Zounds]

Legendarys needs revisiting as some have changed since last update.