[Resto] Stat Weights

[tinef]

Hey everyone,

Just wondering what you guys think the optimal stat weights are. From research I know its Int>>Haste>Mastery>>Multistrike>Crit>>Vers, but I'm unsure of the numbers.

Thanks for any help.

[skiri]

Intellect	2.0
Spell Power	1.8
Spirit	1.2
Haste	1.1
Mastery	1.0
Multistrike	0.9
Critical Strike	0.8
Versatility	0.7

[Hildrande]

Static stat weights aren't very meaningfull anymore, since all stats are fairly equally powerful, and the exact number would depend on how much you have of the other stats, and arguably also on talent choice, encounter, playstyle and overhealing profile.

Take item level over anything else and gem/enchant/eat Mastery if you like phat tranqs or Haste if you use DoC or don't have mana problems.

[Ctee]

These threads are kind of pointless. We all know the stat priorities by now. If a piece of gear dropped that was an upgrade for you, take it. Unless it was something horrible like crit and vers.

[Sprucelee]

These threads are kind of pointless. We all know the stat priorities by now. If a piece of gear dropped that was an upgrade for you, take it. Unless it was something horrible like crit and vers.

But neither are "horrible". For instance, the weights in the post above would imply that haste is ~50% better then crit/vers, which is just silly.

Yes, you want to focus on haste/mastery gear, followed by multi, but the rest are much closer than people make out.

Essentially, if there is any sort of intellect upgrade, it is worth it, barring other people in your group needing it more.

[nightfalls]

Static stat weights aren't very meaningfull anymore, since all stats are fairly equally powerful, and the exact number would depend on how much you have of the other stats, and arguably also on talent choice, encounter, playstyle and overhealing profile.

Exactly this. When I weight items, I usually use the following as an approximation but exact weight AND stat choice do vary based on all of the above:

Int: 2.0
Spirit: 1.8 (estimate: I am trying to come up with a more concrete figure - it is much harder than the other stats though!)
Haste: 0.95
Mastery: 0.9
Multistrike: 0.85
Crit: 0.8
Versatility: 0.7 (+throughput component only, there is also a DR component)

[Alroxas]

Exactly this. When I weight items, I usually use the following as an approximation but exact weight AND stat choice do vary based on all of the above:

So wouldn't it be better to just sim your character for the right stat weights since so much is dependent on how much you x/y stat already have?

[nightfalls]

So wouldn't it be better to just sim your character for the right stat weights since so much is dependent on how much you x/y stat already have?

I have my own formulation calculator for my own character, and the weights I get are generally along the lines of what my approximation is. So when making BiS lists or comparing trinkets, the approximation works pretty well for characters in realistically obtained gear.

[tinef]

Thanks for the input everyone. I realize that item level>all, but I'm just trying to find the cross over point. ie. Is a 4 ilvl upgrade worth it if it replaces my haste/mastery piece with crit/vers?

[Hildrande]

So wouldn't it be better to just sim your character for the right stat weights since so much is dependent on how much you x/y stat already have?

Can you sim healing?

The ultimate way to calculate stat weights would rely on % healing done by tranq, non-HoTs and DoC and be fight specific. I'm hoping AMR or WCL will pick this up.

- - - Updated - - -

ie. Is a 4 ilvl upgrade worth it if it replaces my haste/mastery piece with crit/vers?

Remember that stamina, armor and less damage taken all have value greater than zero.

[Deleted]

Can you sim healing?

You'd have to make an unreasonable (imo) amount of assumptions. I think Geodew has done something like this for Mistweavers. I never checked his math, but he has come up with stat weights that are regarded as "agreed upon" in the MW community from what I know.

The process could look something like this:
Use a spreadsheet like Hamlet's Healercalcs to find the strengths of each healer.
Decide on a target boss fight. Guesstimate reasonable kill time.
Decide on a healer comp.
Decide on a cooldown strategy, fill GCDs with reasonable filler spells.
Define a (number of) parameter(s) to maximize (like safety, kill speed,... )
Make assumptions about player/raid skill and define an error margin.
Sim based on those requirements.

Doesn't seem worth the hassle when "Spirit > 2-set > max ilvl, Tree or SotF & NV & Germination" probably performs above 90% of the max performance possible for most fights and has a reasonably low skill floor.

Just my 2 cents.

The ultimate way to calculate stat weights would rely on % healing done by tranq, non-HoTs and DoC and be fight specific. I'm hoping AMR or WCL will pick this up.

That's another way to go about it. Looking at logs and saying 'if everything went exactly like this again, you could switch to this gear and perform better'. I think that it's more about keeping nerds that like to look at numbers entertained (including me ) than actually having value. The obvious ones like finishing a fight with 80% mana, you don't need a combat log analyzer for. The non-obvious stuff like one stat performing 5% higher than another one are nullified by someone standing in a void zone for too long on the next pull and more/less boss mechanics being targeted at you.

Again, my 2 cents.


I'm not trying to pick an argument with you, nor trash-talk AMR or WCL. I just don't think that there will be much value coming from that. At least for rdruids.

[Hildrande]

I think that it's more about keeping nerds that like to look at numbers entertained (including me ) than actually having value.

I'm not planning on applying to Method this week, probably not next week either, and my attitude is that the effort that can be put into theorizing gear optimalization beyond taking ilvl and gemming/enchanting/eating whatever, really, could most likely make me a better raider spent elsewhere.

I tried make this huge spreadsheet with all the %SP's where I could enter the combat log data, saw enough to realise secondary stats matters very, very little over ilvl and dropped the effort. It's not all about HPS anyway, quantifying all aspects of healing to turn it into numbers is not very feasible.

[nightfalls]

I tried make this huge spreadsheet with all the %SP's where I could enter the combat log data, saw enough to realise secondary stats matters very, very little over ilvl and dropped the effort. It's not all about HPS anyway, quantifying all aspects of healing to turn it into numbers is not very feasible.

Exactly. Also since ability usage varies from fight to fight and even pull to pull, it's not like it's ever going to be "exact" anyway. You're getting, at best, an approximation from the massive amount of math/TC work.

What I have been doing, I run a formulation to get stat weights for:

1) General abilities.
2) HoTs versus direct.
3) Regrowth counting LS + either glyphed or not (changes crit value).

Then, you just estimate the actual stat weights based on how much you believe each thing will matter, based on your general ability usage. i.e. if you are heavily relying on HoTs (which every druid should be doing) with Rejuv/WG, then you need to be putting the most weight into the "HoT" stat weight which puts haste really high.

Since Tranq puts haste a lot lower and is also a high %, balancing mastery vs. haste comes out to how much you value the total healing done by Tranq, while keeping in mind the faster channel can be a blessing also (or a curse, but realistically for progress it's almost always better).

[Deleted]

I'm not planning on applying to Method this week, probably not next week either, and my attitude is that the effort that can be put into theorizing gear optimalization beyond taking ilvl and gemming/enchanting/eating whatever, really, could most likely make me a better raider spent elsewhere.

I tried make this huge spreadsheet with all the %SP's where I could enter the combat log data, saw enough to realise secondary stats matters very, very little over ilvl and dropped the effort. It's not all about HPS anyway, quantifying all aspects of healing to turn it into numbers is not very feasible.

Yeah, same here. I made my own spreadsheet, then looked at Hamlets sheet, and another one that I can't remember who made it. It's fun, and it's important to understand how spells interact with stats.
There's this saying "DPS is a science, healing is an art" which I very much agree with. Basic HPM&HPS math is like understanding how a brush works, or the difference between oil-based and water-based colors, thus crucial. But once you start trying to calculate the perfect picture it goes downhill xD

I guess we are pretty spoiled as rdruids when it comes to stat balancing this expansion. I hear MW value mastery just slightly above stamina, and half of the healers stopped being mana-capped halfway through BRF haha

[Pory]

I dropped a lot of my heroic HM items for the set bonus normal BRF gear and noticed a huge increase in my numbers but other than that, I tend to focus on intellect upgrade above whether the item has Haste or Mastery on it.

[Deleted]

Yeah, same here. I made my own spreadsheet, then looked at Hamlets sheet, and another one that I can't remember who made it. It's fun, and it's important to understand how spells interact with stats.
There's this saying "DPS is a science, healing is an art" which I very much agree with. Basic HPM&HPS math is like understanding how a brush works, or the difference between oil-based and water-based colors, thus crucial. But once you start trying to calculate the perfect picture it goes downhill xD

The trick with achieving healing stat weights is to do it retroactively. When you look at a white canvas and start to calculate what colours mixes or pencil types you need, you will naturally get largely meaningless results. But look at a finished painting. See what colours worked, what pencil gave the best strokes. Calculate backwards based on what you have already achieved, and you'll know how to optimize it in the future.

Resto Shaman Stats was a god-send when I mained my resto shaman last expansion, but it sadly doesn't seem like such a crucial tool has ever been made for the other healing classes.

[theburned]

Can you sim healing?

The ultimate way to calculate stat weights would rely on % healing done by tranq, non-HoTs and DoC and be fight specific. I'm hoping AMR or WCL will pick this up.

- - - Updated - - -


Remember that stamina, armor and less damage taken all have value greater than zero.

well stamina and reduced damage has a slight value above 0 its not very much, but a little, the extra int and extra amount of stats tends to push it over the value of stats, also those are generally "unwanted" stats thus you should less of them, and they are less hindered by diminishing returns. Personally I have a tad bit too much crit, which goes very well for my boomie off specc, also I have 2x wrists M and 2x chest M+ tier Hc giving stats that are great for some talents like DoC/hotw (dps) while others give more healing oriented stats.

but ilvl is mostly key for resto druid, and mostly trinkets and role oriented pieces, jewelry and trinkets are the most common exception.