Resto (PvE) Healing Discussion - Page 37

bubblecannon · 03/02/09, 8:44 AM

Originally Posted by Nailer

And consequently make us drift away even more from HOTs? Allow me to humbly object. They are already pushing nourish down our throat which is a curious design goal considering it fades the very reason why druid healing feels unique.

The point I was trying to make is that our direct heals don't have the oomph right now so there isn't really any 'choice'. I can't eschew HoTs for Nourish/RG spam and expect to not go OOM keeping a tank alive. If they choose to push us down the direct heal path we need either buffed nourish/RG/HT or a minor class spec redesign.

Playered · 03/02/09, 9:42 AM

The Restoration Druid Mana Budget v3.1
These are just rough figures which *should* be okay as a rough estimate.

Fight Duration:	5 min	6 min	7 min	8 min
Wisdom	110	110	110	110
Mana Spring	105	105	105	105
Mana Pot	72	60	51	45
Replenishment	200	200	200	200
Mana Pool	350	292	250	219
Intensity	375	375	375	375
Innervate(1)	250	208	179	156
Mana Tide(1)	88	73	63	55
OOC	122	122	122	122
Total MP5	1671	1545	1454	1386

This table assumes:
1) Pretty much BiS.
2) 2T7.
3) Raid buffed.
4) 1 use of Innervate/Tide (just double the select value if you want to assume 2).
5) OOC with a 2PPM used on Lifebloom.

Then we can budget some "jobs" which we would do:
1) Roll LB on a tank - 407 MP5 drain.
1a) LB incurs a start up cost of 1348 mana which means sticking a fee per LB roll which diminishes over time:
5m: 22 MP5, 6m: 19 MP5, 7m: 16 MP5 and 8m: 14 MP5 - exclude this if you intend to build up LB slowly (as in not rush a 3 stack) - thanks Dukes.
2) Roll RJ on a tank - 123 MP5 drain.
3) WG on cooldown - 536 MP5 drain.
4) WG once per 10 sec - 322 MP5 drain.
5) Nourish once per 10 sec - 286 MP5 drain (Glyph assumed, no Tranq Spirit yet).
6) Swiftmend once per min - 47 MP5 drain.

So rolling LB/RJ on one tank will be a 530(544, 8m) MP5 drain (HPS average 2300 - overheal not factored).
Sustaining RJ/Nourish per 10 on another tank will be a 409 MP5 drain (HPS average 1660 - overheal not factored, Living Seed and 44% crit raid buffed on Nourish).
You can then see that on an 8 min fight we couldn't sustain adding a WG on cooldown to that but we could throw in WG once per 10 sec, Swiftmend per min and maybe the odd bonus Nourish per min too..
Alternatively we could do LB/RJ x2 with Nourish per 10 sec and Swiftmend per min too and forgo any raid healing at all which would put our HPS cap at around 5500~ on the combined tanks.

I do not know if Regrowth will be worth using on the tanks when our mana is this constricted either considering each LB is now costing 200 MP5 more to sustain and we have lost our ability to surge through spells with a 120%~ Innervate making sure we still end up with 100% mana at the end. Needless to say our "winning" the Sartharion 3d meter will sink faster than a tauren tied to an anchor now that we will be unable to do 2x tank rolling with decent raid healing too.

Need to try and find some rough estimates on RJ/LB overhealing too which shouldn't be that hard to do with some controlled tests but it needs to be on Sarth3d to be a more reliable test which complicates things :P

Honestly though I am fine with increasing LB mana cost but they need to somewhat compensate it with the increase to the healing to keep the HPM roughly equal in order to not fuck us around on it. LB on more than one tank is a choice rather than something assumed is fine, being crippled and still having that choice is not so good without some heavy implications. Having the choice of really solid MT healing + (solid OT healing/solid raid healing/mix of both at a weaker level) does not seem unfair compared to what we have now (which is solid MT/OT/Raid combined).
Having more time to test without this specific change would be greatly appreciated because a lot of Druids will be pretty grumpy without visibly being able to see that they were too strong on top of the other changes which warranted this change too.

ps: I would laugh if there was ever the idea of these changes trying to make MP5 more attractive but I would not be surprised if we started picking MP5 as the lesser of 3 evils (Haste, Crit, MP5) on gear once past the haste softcap with these changes.

chronas · 03/02/09, 10:57 AM

Originally Posted by Playered

ps: I would laugh if there was ever the idea of these changes trying to make MP5 more attractive but I would not be surprised if we started picking MP5 as the lesser of 3 evils (Haste, Crit, MP5) on gear once past the haste softcap with these changes.

It has been a week or two since I visited the resto druid loot rank, but was wondering why all of a sudden, they are spamming Lustrous Sky Sapphire, revitalizing skyflare diamond, and dazzling forest emeralds as the gems of choice.

Is our saving grace in 3.1 uber regen? It appears to be as of yet.

just a thought

Cube · 03/02/09, 11:01 AM

Playered, where are you getting 122 MP5 from with OOC?

Assuming it's used only on lifebloom, with a LB cost of of 450 mana per cast((1350/3)*2), you're saving 900 mana per minute assuming 2 PPM.

That's 15 mana per second, which translates to 75 MP5.

Playered · 03/02/09, 11:07 AM

Originally Posted by Cube

Playered, where are you getting 122 MP5 from with OOC?

Assuming it's used only on lifebloom, with a LB cost of of 450 mana per cast((1350/3)*2), you're saving 900 mana per minute assuming 2 PPM.

That's 15 mana per second, which translates to 75 MP5.

LB will be 732 mana per cast (2T7) when the changes are applied to double the mana cost I assume.

Cube · 03/02/09, 11:14 AM

Ah, that makes sense. Thanks.

I was going by your initial cost in the 1a line... shouldn't that be doubled, as well? Or am I just completely insane(it's early on a Monday, forgive me).

Maraili · 03/02/09, 5:07 PM

Playered, I agree with your math on using OoC for LB, but I'm not sure you should be counting it on your mp5 table, as it was stated in the big blue post on mana regen that they are very seriously considering changing it from a free spell into something more akin to a mana return proc (at least the way I'm reading it). Now granted, no details on this change have been forthcoming as of yet, but it is worth bearing in mind as we consider these changes. If the ppm stays roughly the same and it returns more mana then the cost of one lifebloom, our effective mp5 goes up, but it could also easily go down if it returns less.

bubblecannon · 03/02/09, 5:49 PM

Originally Posted by Playered

Honestly though I am fine with increasing LB mana cost but they need to somewhat compensate it with the increase to the healing to keep the HPM roughly equal in order to not fuck us around on it. LB on more than one tank is a choice rather than something assumed is fine, being crippled and still having that choice is not so good without some heavy implications.

Yep, this is exactly the point we need to make clear to the developers. In itself, the mana change is not the problem. Resto druids have lower raw HPS throughput than the other healers and that needs to be balanced in some manner. Our preference, is for that balance to be met by increased HoT HPS. Time will tell what the developers feel.

That being said, I still think it would have been more reasonable to just increase our HoT cost across the board by 25-40%, and buff our HoT HPS by say 25-30%. This would have increased our HPS and also increased our mana usage whilst still giving us the choice of rolling LB on multiple tanks. They could have left RG as 50% crit because it would have been 25-40% more expensive to employ making nourish far more attractive for spot heals (and maintained a real differential between them). RG now becomes a reasonable spam nuke with the glyph comparable with the other healers whilst being too expensive for us to maintain. Nourish would then become our general tank spot heal because of it is fast cast time, cheap cost and reasonable heal. This would also have retained RJ as our 'goto' raid heal for more significant raid dot dmg instead of an LBx2 stack.

The developers do need to come clean and admit that LB is not out-of-the-ballpark HPM efficient and provide their real justification for the mana increase. (eg. wanting to make druids spend more mana? perhaps just dungeon design made LB too strong? etc etc).

Rijndael · 03/02/09, 6:20 PM

Originally Posted by bubblecannon

Resto druids have lower raw HPS throughput than the other healers and that needs to be balanced in some manner

Is this really true? It's certainly not true for raid healing, where druids dominate. I haven't noticed doing worse than other tank healers at keeping up a tank (I can keep up a tank mostly solo vs Maly 10 while still having time to throw out Wild Growths and Rejuvs on the raid). What am I missing? Is there some math you have in mind that's behind your assertion?

Shelendil · 03/02/09, 6:52 PM

Originally Posted by Rijndael

Is this really true? It's certainly not true for raid healing, where druids dominate. I haven't noticed doing worse than other tank healers at keeping up a tank (I can keep up a tank mostly solo vs Maly 10 while still having time to throw out Wild Growths and Rejuvs on the raid). What am I missing? Is there some math you have in mind that's behind your assertion?

Looking at raw healing, not effective. If we ever get to a min/max situation where exactly how much hps each healer can output (sustain) is put to the test and overhealing for everyone becomes very low, then druids will fall behind the other healers. The last time we saw an encounter like that was probably M'uru. If the content never gets that razor edge again, then it may not become a problem.

Zuel · 03/02/09, 7:10 PM

On encounters where the raid is taking steady, predictable damage, druids should be and usually always are on the top of every 'meter'--whether it be raw healing or HPS. But the nerf to lifebloom was to specifically address tank healing, an area that we certainly do not dominate by any means.

Looking at something like this is sort of staggering:

WoW Meter Online - Combatlog Replay

Not a big surprise, holy paladins are fairly insane at the moment, but still fun to look at and /boggle.

Playered · 03/02/09, 7:17 PM

Originally Posted by Zuel

On encounters where the raid is taking steady, predictable damage, druids should be and usually always are on the top of every 'meter'--whether it be raw healing or HPS. But the nerf to lifebloom was to specifically address tank healing, an area that we certainly do not dominate by any means.

Looking at something like this is sort of staggering:

WoW Meter Online - Combatlog Replay

Not a big surprise, holy paladins are fairly insane at the moment, but still fun to look at and /boggle.

While that is funny it is hard to compare that to Druids because our overhealing is not shown for the most part which cuts it down significantly for us.

If you looked at the EHPS table you will see Druids rocket to the top for a fair amount and generally this is because we spread our healing out a lot and say a couple of ticks on a MT get sniped we still have hot ticks on 2 other tanks and half the raid chugging a long. Compare that to if a Paladin gets his direct heals sniped it is a signifcant loss.

Sarth is a great fight for a Druid because it has multiple tanks that is not all spiky, random raid damage quite often and a chunk of time with high periodic damage on the raid. If the current pinnacle fight of T7 was some heavy spike encounter on both tanks and raids we would jump down a fair amount while Priests/Paladins (and maybe Shamans) would go up quite a lot.

We could really use Blizzard coding it so that HoT ticks were displayed in overhealing for these kind of discussions and comparisons as I think it would surprise us how high it is.

Rijndael · 03/02/09, 7:20 PM

Originally Posted by Shelendil

Looking at raw healing, not effective. If we ever get to a min/max situation where exactly how much hps each healer can output (sustain) is put to the test and overhealing for everyone becomes very low, then druids will fall behind the other healers. The last time we saw an encounter like that was probably M'uru. If the content never gets that razor edge again, then it may not become a problem.

I don't understand what that means. Are you claiming that a 3-stack of lifebloom plus Rejuv + Regrowth + Nourish spam with the glyph is less raw HPS than, say, Holy Light spam? I think resto druids in the current state would be rather good on a fight like Muru. Their combined HPS on tanks + raid is unmatched once they get going.

Playered · 03/02/09, 7:32 PM

Originally Posted by Rijndael

I don't understand what that means. Are you claiming that a 3-stack of lifebloom plus Rejuv + Regrowth + Nourish spam with the glyph is less raw HPS than, say, Holy Light spam?

Note: none of this factors in overhealing at all, it is raw HPS.

3x LB is 1500~ HPS
1x RJ is 750~ HPS
1x RG is 400~ HPS
This means hots make a 2650 HPS on one target.

Nourish in 3.1 with Living Seed and the glyph (or set bonus) will put out around... 9000 HPS if spammed constantly but it would drain mana at a rate of 2300 MP5 which is something we -cannot- sustain for very long at all unlike Paladins which can sustain HL for a lot longer.

I believe the Paladin HPS was assumed to be around 9000 too assuming full BiS gear and for a duration of 5-6 mins~.

bubblecannon · 03/02/09, 7:35 PM

Originally Posted by Rijndael

Is this really true? It's certainly not true for raid healing, where druids dominate. I haven't noticed doing worse than other tank healers at keeping up a tank (I can keep up a tank mostly solo vs Maly 10 while still having time to throw out Wild Growths and Rejuvs on the raid). What am I missing? Is there some math you have in mind that's behind your assertion?

IMHO we dominate on raid heals more to do with the 1 sec ticks from WG and LB than their raw HPS output. We simply end up sniping heals from the other classes. As for raw single target HPS we are far lower on reactive peak healing and about 20% lower on single target nuke healing assuming we can maintain a full LB/RJ/RG stack on the target at all times.
(I'm actually playing with a google docs spreadsheet that covers all of this. I'll post it once completed)

@Playered:
From my calcs, 9K HPS of nourish in 3.1 assumes min of 2900SP, 1 sec casts and doesn't factor in the lost GCD's maintaining the HoTs. If you factor those in, you drop in real world terms about 2/9 of that output to about 7K HPS. However, that is balanced by the HoTs themselves which roughly return nourish spam + RJ/RG/LBx3 to 9K HPS but an insane mana cost.

Every 9 secs you're looking at 1xLB, 0.5xRJ, 0.33xRG and 7xNourish: (740+223+240+3563)/9*5=2647 MP5
The holy pally in comparison is consuming 1 holy light / 2 secs at roughly 700 mana /cast after illum ~= 1750 MP5.

Now, this is only single target healing. If you factor in a Beacon and Holy Light glyph into these calculations, the holy pally throughput could be significantly higher (albeit with overheal as high as our nourish spam). [I'll also leave aside the WTF P0wn of JoL in some fights as it is a beast of a different stripe]