New healer optimization - Healer Playstyle

Hey !

Wanted to get a bit of help understand the “New Healer Customization Option: “Healer Playstyle””.

  1. First of a “bug” ? or actually a feature idk:
    When updating the “healer playstyle” slider in BiB it impact BiS result as well (Bis and BiB have their own slider so i’m not sure if moving the one in BiB should impact BiS when BiS is set to default)
    image
    (Also if you change the slider; already displayed result in upgrade finder and BiS doesn’t update)

  2. Another bug : 2c3df5e616234aae88e244e6305410b3
    image
    This ilvl lariat has no value (it has different stats compare to the two others) It’s a neck i modify to see how it was doing in “Add to my bag” and in BiS it has no value. If i set the stats back to “default” it has value :
    image
    (maybe a cache issue?)

  3. The main topic
    To be honest; i’m a bit lost and unable to choose properly what is the setting i could use for the playstyle slider. I’ve looked at result in both extremes and some middle point. Yes haste is more valued in one case and not in the other. I’ve looked at one item i have in my vault “Eye of the vengeul huricane” it has quite some haste on it compare to my lariat crit vers.


    Weirdly there is a small bump in the curve. And almost 7% upgrade if i play all reactive ? meaning compare to my current gear; having this particular neck in a reactive playstyle i will heal 7% more ? (7% more hps ? Seems weird)

Could you share some more info on how it’s works ? Is it like; “proactive 100% 0%Reactive mean you will do ramps perfectly ?” Or Is it like, you try to snipe as mutch low health target as possible in reactive mode ?
Shoudln’t it be tied to the raid difficulty ? Like “Heroic ~= more snipe effect”; “Mythic ~= more proactive healing” ? (Currently heroic or mythic difficulties has no impact on result right ?)

Why do i ask; i feel like the default “middle ground” feel not where i want to be.
To be able to find the sweet spot, i would use the simulator to see what’s the best “HPS output” and see what the gap between a more reactive gameplay or “what do i lose by playing more in the reactive side”. All this is based on the assumption that it change the “rotation”. Does it ?

Idk but could we have a button on the side with “find the sweet spot” where AMR internally wil compare all the level in between reactive and proactive gameplay; calculate the “distance” to the max HPS and max DPS use a square root normalization to get one value :
image
(with a beeing the slider from full HPS to full HPS)

With that the result should be either full proactive healing to be the best output or have an optimal sweet spot or with a certain treshold be within 5% of the max ?

I can’t seem to replicate the issue with changing the setting on BiB affecting BiS. The settings should be completely separate for the those sections.

It is on my list to revisit getting the upgrade finder to automatically update or reset when you optimize in BiB… actually tried to change that a couple weeks ago but ran into an issue.

I’ll look at that issue with BiS list… I think what happens is when you modify an item it ranks it, but then if you run BiS again, it doesn’t re-rank the custom list entry after the new optimization.

I’ll have Swol answer your questions about the details of how this setting is used for each spec, but in general: this is a pretty quick first implementation of this option. Right now it is more meant as something you can set “to taste”. We probably need to tweak its range for some specs as well to give it a smooth and predictable value across the range of the slider.

Sometime soon we will create multiple preset strategies with this slider set at values that we think are good for each strategy that you can use as a reference. Then you would push it one way or the other incrementally based on personal preference rather than some kind of calculation.

The extremes of the slider are meant to be pretty “extreme” – so that you can force the optimizer heavily into one mode or the other. Since it is affected by haste, it will be prone to the occasional dip or peak – just the nature of it.

Right now the heroic vs. mythic difficulty setting is mainly for tanks, as it adjusts the armor constant. We may end up using it for other specs eventually though.

1 Like

I know you probably want this setting to be some sort of fantastic calculation that figures out how good haste is… But that’s not what this is for.

I’ve already done fantastic calculations that show haste isn’t that great for most healers when you want to have max output. (Resto druid is only one I’ve found to scale well with haste)

Players have told me they want haste anyway. I can’t magically make the kits scale better with haste. So, instead we are making this setting that captures the “feel” of haste being better by making healing that can “snipe” scale more with haste. I set it up so all the way to the left you use my calculations that I think are correct for max output. All the way to the right haste becomes your best stat.

Mistweaver, in particular, seems to scale very poorly with haste. Which I know flies in the face of what is popular right now. I’m using this setting for now to default to something more mainstream that gets a chunk of haste. I will revisit the model in the coming weeks as well. I need to make this new setting interact in some smarter way with fight length too.

Thank you both of you for comming back to this post !

Hopping back after work, seems i can’t either… so probably a miss read on my part

No prb !

I hope it does for healer too at some point i think it can be interesting. I think @Swol mentionned dmg pattern in the current calculation method so maybe it could influence it ? idk

Got you ! Thanks for clarifying. I will probably stick to the “zero haste” for now. Changing all gems and enchant and recrafting stuff is too expensif to play around with it. Till this point i’m pretty happy with my perfs/gear suggestion; i only need to play better and not die :slight_smile:

I think i understand something now. When i change the slider; if goes from “the optimal proportion of stats AMR found” to “100% haste lover” and in between it value more and more haste by “removing” value from other stats.

It would explain why in the base value “Horn of Valor” is value ok (+0.64%) and when asking haste for less (-0.43%)

It’s been discussed in another topic, but just like swol said, there is a significant difference in effective healing vs potential healing. Most guides want you to be top of the charts, which in theory you would be with amr calculations IF enough damage was occuring. the problem is most raids run with either super geared healers, or extra healers, which makes effective healing more “lucrative” because otherwise your heals are simply not getting through as effective. Them adding the slider gives you the option to play a less efficient but “better” looking healer. Unfortunately this is just the situation with healers and 90% of the time raid leads don’t care to analyze healing as long as you “look” good. Often times you will see in progression fights the real story of healers, and that’s usually your high haste builds being out of mana fairly quickly and not really making a difference vs random damage. however if your team is really good about minimizing damage the high haste builds will make you “look” better as you will get quicker heals in before the other healers. So consider what you are doing, and who your other healers are to decide which build might be optimal for you. The other guides don’t explain this well and just blanket pick haste for heals.

1 Like

Yeah I think i get the idea now.
That’s really interesting :thinking:.

So technically, looking for potential healing can be interesting in progression where we may not be overgeared yet (or a healer may die in the fight)
And then swap more and more into effective healing during farm session. Am I getting it correctly?

I generally have always favored balanced stats. I tend to never pool stats for healing as it just doesnt make sense to me. id rather be the middle ground between the two so i can lean one way or the other depending on whats needed. a lot of times this means I might not be the highest log, but it also keeps people alive better. for example if someone is at 1% health, no matter how much haste you have if your other stats are weak and your heals are low but quick, you might be able to put an “effective” heal on them, but is it going to keep them alive when that aoe smashes? Vice versa, you may have a heal with tremdendous power and crit 100% of the time, but what if you need to heal multiple people who are low? you gotta pick someone to not heal because your cast time is too long to get much heals in.
So i prefer to be right inbetween. A decent heal, and enough haste to get even one or two more people in.
Druids right now are almost an exception to this, because of the hot mechanics scaling well with haste, but even then I still like to keep a good amount of mastery or crit to beef up those ticks. so I’d consider that. You may not win any top logs this way, but it will help you minimize deaths in your team. Now as I said, once people are a bit overgeared for the content and damage becomes minimal, haste will set you apart from the other healers logwise. But in general by that time people arent closely monitoring logs either.
So consider this. if you are trying out for a better raid team or fighting for a position, or streaming and want top logs, dump everything into haste. Otherwise i’d favor a more beneficial stat blend.

1 Like

Can you link me to the post with your calculations and analysis on how healers scale with haste? Super interested, thanks.

I don’t have a post on the forums with a calculation/analysis written up, but I’d like to do that. The optimizer itself shows the results of my calculations and analysis that I’ve done so far. Once I get everything working how I want it to work, I’ll endeavor to write up some analysis where I’m seeing significant differences in the recommended value of haste between what I’m calculating and what other guides/discords are suggesting. It should be an interesting analysis.

I still need a couple weeks to finish implementing all the talents for our talent optimizer, and I want to spend some more time with the healing models in particular to polish them up. I think there are some pretty decent “sanity” checks one could do that show how haste doesn’t scale as well as other stats for a spec like, say, holy paladin. And then taking into account mana pool can at best not affect the value for haste, but usually will tend to make it a less desirable stat if you are in any sort of mana-limited situation.

The mana-limited portion of it is what I see the most disconnect with the rest of the community on. I see people say that they never run out of mana and that mana isn’t a problem… which tells me they aren’t doing as much healing as they could potentially do. That means they don’t need to (in which case “optimization” because extremely fuzzy/feely because they are in a position where topping meters becomes more of the goal than max output), or, it means they are leaving a lot of healing on the table. It could also just mean that people like haste, which is fine, but if you’re going to optimize based on what you like instead of math - you don’t need a mathematical model.

1 Like

they are in a position where topping meters becomes more of the goal than max output

I don’t understand how you think “max output” can result from doing lower HPS.

HPS on meters is measuring effective healing done, it’s not measuring how much healing you could possibly do with your mana if you always had someone to heal. Max output is the latter. This is one of the never-ending design issues with healing in a game like WoW.

The only healing that benefits anyone is healing actually done.

Why optimize around a theoretical concept which is completely divorced from how the game is actually played?

Assuming the boss dies, the length of a fight is largely independent of what the healers do (i.e. the “S” in HPS is approximately constant) and the highest HPS achievable for the fight is, ipso facto, the maximum output achievable for the fight.

Maybe your playstyle slider would be more intuitive if the two ends were labeled “HPS” and “HPM”.

I think the question here is how effective do you want to be. high haste will definitely make you “look” nicer in low damage fights. But in a situation where you are progging or lots of damage is going out, your effective hps will be higher with a focus on maximum output and mana efficiency. Like i stated before it really boils down to your group comp and your teams ability to mitigate incoming damage. if your team isn’t taking any damage and you are running 4 heals for a 20 man group, haste will let you get more effective healing in, but when you reach a boss where everyone is taking tons of damage and things are messy, then max output is going to give you “higher” hps as there is a need for it. amr assumes there is a need for healing and maximizes your mana efficiency and output. if there is no need for healing it doesn’t matter what stats you have unless you are just trying to parse or look good.
there used to be an addon that tracked the most useful healing in the raid.
for example it gave points based on health % restored and other factors such as what their health% was before the heal. unfortunately most people stopped using it because of hps trackers, but the assumption that the highest hps person is the most effective healer is where a lot of people are wrong.
for example as a druid i can literally just keep hotting everyone in the raid as much as possible, but those 2k-3k ticks arent more effective than say a 20k heal on someone with 50% health, but because its always healing my hps will be high as long as people aren’t at 100%.

Still looking forward to some kind of evidence to support the black box.