Hey Guys, I was hoping I could get some feedback on my healing as a druid so I can make any improvements in the way I approach healing.
Just some background info, I currently sit around 1863 +healing unbuffed and get around 2005+ healing buffed as well as 10.4k health and 10.3k mana. Generally Im used as either healing the mt or group patching which depending on the situation I'd throw a LB or two on a player depending on how much health they are losing.
Anyway, heres my wws report from Hyjal, excuse the RoS attempts, first time we attempted RoS so it was bound in wipes!
Well as most of the last 5pages or so of posts say, it's really dependant on the rest of your raid, the group makeups etc etc. As far as I could see from a quick glance, you seem to be alright anyway. You're not lagging behind other healers, and you have a decent balance of spells, at least in my opinion. Depending on how much mana you end up with, you can start using more regrowth which might or might not result in more healing, but as far as I understand, there's no secret magic for druids, you just have to chain cast from start to end, and use up as much mana as possible without ever running oom. How much you heal, what spells you use and how long you can stay vary depending on your group your role your assign or your gear(and the encounter). As long as the tank is not dieing, and you're not last/second to last on effective heals, you're doing your job fine imo.
184, but yeah.
I believe its true that GCD may be triggered client side, which means that latency shouldn't influence the casting as much (if any at all). So it makes sense that you can reach 5 GCDs with rather low spell haste even at high latency. I think Giansm calculated it to be 113 spell haste if there was no latency at all. So latency does affect it, just not as much as we thought, I guess.
But Im not very good at network and latency when it comes to these details, so I'll leave that to someone who knows more
He did this with only 182 spell haste and a ping of 360ms ?
Yes. As has been noted, it's not so much the latency that matters as the volatility of your latency. If you assume a completely static latency, then all your casts are delayed by the exact same amount so this doesn't affect your ability to roll LB stacks (see post 446 for a clearer explanation). Hence, all that mattes in the spell haste.
7/5 = 1.4
So as long as you have spell haste greater than (1.5/1.4 - 1)*1570 = 112 haste it's plausible that you could sustain 5 LB stacks. It's actually possible to do it on a cycle or two with even less haste than that, if your latency happens to be improving at the same time (see 449-451). However, the fact is you will rarely (if ever) play in an environment where latency variance is insignificant. Because of this, you need more than 112 haste to accommodate for those times when your latency is increasing.
Consider this example with no haste (assuming no human error - i.e. you cast as the ms your GCD is complete):
Cast 1 with 1000ms latency at t=0. Server side, it is cast at t=1s
Cast 2, 3, and 4 with 50ms latency occur at t=2.50s, t=4.00s, and t=5.50s respectively
Cast 5 with 1500ms latency (an unfortunate lag spike) then occurs at t=7.50s
So cast 1 to cast 5 has a 6.5second delay. i.e. It's the 500ms change in latency that slows down your cycle, and you're able to keep your LB stack up despite playing with significant latency. When you start aiming for 5 cast cycles, you often don't have this full second window for variance (at 112 haste, you have exactly 0 room for a latency spike) and so it becomes much more difficult to keep up your LB stacks. In order to have the exact same room for error/latency in a 5 cast cycle, you would need to get sufficient haste to fit those 5 casts into a 6 second cycle (1.5sx4= 6 seconds). This requires a 1.2second GCD, which equates to 392.5 haste. It's still possible with (much) less, but the room for latency variation isn't as high as it is in your standard 4 cast cycle.
Over large periods of time this latency is largely irrelevant. For druids though, who are operating within 7 second windows, any variation in latency can have a significant impact on our ability to heal.
Well, it looks like there that it should be possible to do a 5 cast cycle (5 lifeblooms, or 3 lifeblooms and 3 rejuvs, or 4 lifeblooms and 2 rejuvs or whatever. i know those add to 6, really its only 5 casts per cycle though, go through it in your head) with only a small amount of haste if your latency is perfectly stable at any value and you can do perfect button pressing. over the course of an actual fight, and with small variations in latency, and not having a robot playing your character, I still think closer to 200ish haste is going to be needed to really maintain it without it dropping pretty often. Anyone do some practical testing on what haste counts they need to actually maintain a 5 cast cycle for several minutes? hmm.... g15 macro spamming sounding tempting.
Um, anyone test if spell haste changes the GCD associated with swapping weapons. Like, i for one, and i'm sure a lot of people use casterweaponswap to change between s3 mace with +healing, s2 mace with spell surge, and solarian's staff with 20 spirit on the fly. Normally i don't lose any time to that because the mod times the weapon swap gcd inside a spell gcd. Is that gcd also reduced? I have a total of 0 haste gear currently, so not much of a way to test it myself. I suppose i could borrow my roommate's paly.
On another note, i didn't do my testing for how many times i cast hots compared to how much they actually tick because my range check in grid is busted and i didn't want to see how bad i did without that in hyjal.
When you start aiming for 5 cast cycles, you often don't have this full second window for variance (at 112 haste, you have exactly 0 room for a latency spike) and so it becomes much more difficult to keep up your LB stacks. In order to have the exact same room for error/latency in a 5 cast cycle, you would need to get sufficient haste to fit those 5 casts into a 6 second cycle (1.5sx4= 6 seconds). This requires a 1.2second GCD, which equates to 392.5 haste. It's still possible with (much) less, but the room for latency variation isn't as high as it is in your standard 4 cast cycle.
actually, that's not even enough, because that 1 second cushion that we have now on 4 casts is basically .25 seconds of latency variation + human error per cast. To get that with a 5 cast cycle you'd have to have 1.25 total seconds free, so fitting 5 casts into 5.75 seconds. That pushes you over 400 haste needed to get 5 as easily as 4 now. Well, i guess its probably somewhere inbetween, as the latency spike variation would get the full second to work with anywhere it was, but the human lag would apply on each cast. But we can safely say its over 400 haste to make a 5 cast cycle as easy next patch as a 4 cast one is now.
That's really the issue. Maybe other people do, but i'm not perfect at keeping lb up, I'm estimating in times where i'm trying to stack it on 4 targets, i let one slip probably once a fight or so, more than that if i haven't rebooted my pos computer in a while. People with more flaky latency or computers or slower reactions would be even more often. It takes hundreds of haste to get the same cushion you have now, so unless you're already almost never letting lb drop, it would take like, every peice of haste healing gear in the game.
Really, its going to come down to how perfect you are at button timing, how stable your latency is and probably most important, how terrible it is for lifebloom to drop once in a while, because its almost certainly going to unless you stack obscene amounts of haste.
actually, that's not even enough, because that 1 second cushion that we have now on 4 casts is basically .25 seconds of latency variation + human error per cast. To get that with a 5 cast cycle you'd have to have 1.25 total seconds free, so fitting 5 casts into 5.75 seconds. That pushes you over 400 haste needed to get 5 as easily as 4 now. Well, i guess its probably somewhere inbetween, as the latency spike variation would get the full second to work with anywhere it was, but the human lag would apply on each cast. But we can safely say its over 400 haste to make a 5 cast cycle as easy next patch as a 4 cast one is now.
My example noted that I was assuming no human error, but I suppose I should've made it a bit clearer that everything else was on that premise as well. Going back to the non-human environment though, I don't think you can look at it as 1.25 total seconds free (i.e. derive it based on 0.25 seconds per cast). It's not the latency variation per cast that matters, but rather the latency variation per cycle. Looking at it as an average per cast doesn't work because you will often have negative (latency improvement) bits. If I have a 5 spell cycle, all I care about is the latency difference between cast 1 and 6, cast 2 and 7, cast 3 and 8, etc. If cast 1 is at 50ms latency, and cast 6 is at 100ms latency, then my latency variation is 50ms. It doesn't matter if casts 2-5 were at 50ms latency or 2000ms latency, I'm still going to keep that LB stack up with cast #6. Hence, the number of casts in between doesn't matter, it's just the change in latency over the course of the cycle. Since the cycle length hasn't changed, 1second is still all that's required to maintain the same room for variation. However yes, the human error component would be multiplied by 1.25 due to there being one more cast in the cycle.
Either way, to reliably maintain a 5 cast cycle without letting lb drop, with a decently varying latency and anything short of absolutely perfect button pressing requires an absurd amount of haste. If you're just barely ok with a 4 cast cycle now, it would take around 400 haste to get that same 'just barely ok' factor with a 5 cast cycle. Looking at what gear is actually attainable, you'd be taking a pretty big hit to your other stats to pull that off.
Anyone know of an addon that tracks latency variation?
It's not just latency, you also need a strong PC post-2.2. If you experience frame rate issues in encounters like Teron or even worse some AOE-heavy part of MH trash, you won't be able to take full advantage of haste effects.
Yeah, I noticed it was easier for me to roll 4 stacks after I cleared out and toned down my addons folder, dropped some of the video settings, and stared at walls & floors for busy encounters. On Teron I run straight through, stare at the wall behind the boss, and zoom my camera into it. There's nothing in my area I need to watch for not standing in there.
On council, I have a much harder time because I'm zoomed out, watching all the spell effects and getting ~15 fps.
Haste doesn't excite me much, but at least I will be able to roll 4 lifeblooms without much work with the limited haste I plan to have on occasions where I normally would've only run 3.
Ditching any profession you have on the first news of a new patch item is just foolish. Do yourself a favor and wait until you see the whole patch.
Anyone still think they might actually change the t6 4piece bonus to a lifebloom buff?
Anyone still think they might actually change the t6 4piece bonus to a lifebloom buff?
Originally Posted by EU CM
It easy to see why people would view this change as a nerf, but the truth is that the coefficient was already artificially inflated.
Resto druids are still very strong at what they do. I notice when I'm raiding with mine that I don't do very well on the healing meters, but that's just because of the way that druid healing works.
Edit: *meh* this thread is older than I thought it was, but it's still as good a place as any to post this
That completly messed up statement makes me doubt they will do anything to further enhance LB beyond where the current nerf has it.
The underlined sentance is baffling, but this is the only scrap of information about LB we have been given since its change - and it sounds like it wasn't aimed at a direct PvP change but an overall "Lifebloom is too good" change.
Changing the 4T6 bonus to LB will atleast make everyone shut up untill WoTLK however as it will cover it up mostly, but it doesn't seem they want that to be the case from what I can make out of it.
The problem is just, we have very good healing spells in theory, and the majority are ignored because of how good LB is.
Nerfing LB is fine but it really should include an additional buff to make those other spells actually live up to their potential and be able to compete with LB.
That completly messed up statement makes me doubt they will do anything to further enhance LB beyond where the current nerf has it.
The underlined sentance is baffling, but this is the only scrap of information about LB we have been given since its change - and it sounds like it wasn't aimed at a direct PvP change but an overall "Lifebloom is too good" change.
Changing the 4T6 bonus to LB will atleast make everyone shut up untill WoTLK however as it will cover it up mostly, but it doesn't seem they want that to be the case from what I can make out of it.
The problem is just, we have very good healing spells in theory, and the majority are ignored because of how good LB is.
Nerfing LB is fine but it really should include an additional buff to make those other spells actually live up to their potential and be able to compete with LB.
Well understandable anyway, while the effective heal isn't that high on single tank bosses, the difference between 0-1-2 resto druids(not sure about higher since we only have 2 in guild) is pretty visible for anyone. 2lifeblooms smoothes spikes so much. I guess I could see why they'd want to nerf lifebloom since it's so efficient, I also wish they would have buffed something else, but well doubt it, it'll go like that until wotlk most probably. Definitely not broken, and probably still better than holy paladins in most situations.
As for the ditching profression thing Ribeye, I don't think anyone said you should ditch your tradeskills right now. But you should probably look into making preparations(like farming daillies for gold or leveling the gathering related skill on an alt) since it doesn't hurt to be prepared before everyone gets the same idea as you and the prices go through the roof. I'm still going to drop tailoring whenever I can, since it's mostly useless anyway but for my PMC. Once I replace my PMC with equivalent gear I'll get LW, and if for some reason they nerf the patterns before it goes live, I can just drop LW and take something else. It's not like tradeskills are such a big investment, both money and timewise. Obviously everything is relative, but I don't care about spending 1k or 2k on leveling tradeskills, I have enough ways to make money(that I don't actually use since I don't need money currently).
It's hardly the first news anymore either, if anything the current build, at least in my opinion, will be pretty close to what goes live(barring the bugs and I guess a few adjustments). I'd say, in 3-4weeks top. After all, there's pretty much only muru and twin eredars left to tune, and twin eredars are pretty close to their final version I believe? The classes changes seem to be mostly done, and they're now adding "bonus" stuff, like the cleave not hitting CC change, totally unneeded, so probably not very high on their priority list, if they had time to fit that in, that means they're not thinking too much about other changes. I'm just going with my experience of predicting patches going live anyway.
Yeah, I'm not exactly dropping what i have now, but I've been farming copper and light leather outside IF while I'm waiting in 2v2 queues on my hunter. I'll definitely want to pick up lw on one of my chars, if not both, and maybe JCing for the neck and a stash of gems. Even if i don't pick them up, i figure others will, so, good.
I'm torn because i was all set to finally drop herbing and alchemy for enchanting and lw, then they finally put something worth having in for alchemy. So, need to decide between lw and enchanting. Or get enchanting and enchant all the rings i have, then drop it as soon as the vashj ring 2.0 drops off sunwell trash and level lw. The major thing is that blizzard has been getting better and better with making trade skills actually useful.
On the subject of lifebloom, i would have been so much happier if they would have just nerfed the efficiency instead of the hps. Make it cost 9% more mana instead of 9% less hps. That would fix a lot of itemization issues as well and give us a reason not to stack pure red gems.
That completly messed up statement makes me doubt they will do anything to further enhance LB beyond where the current nerf has it.
Wouldn't read too much into it, said GM has shown quite significant lack of class knowledge in the past.
He's stated, for example, that druids can tank any boss in the game. While not exactly untrue, the amount of effort that needs to go into a druid tanking Illidan (hello shear), RoS P2 (No reflected deaden -> raid dps issues), archimonde (especially for horde before fear mechanics were changed) or Kael (pyro without a shield is ouch) is just too high for the approach to be reasonable for pretty much any first kill.
How does the blue dragon card scale with the changes on PTR? I'm considering getting it for future potentional use, my guild has cleared ssc/tk and most of us are rerollers from guilds that have cleared BT so we should progress fairly fast. But I'm thinking with the changes on PTR to spirit + that card + potentionally healing touch spamming outside of hotting it might work pretty well.
I also kinda doubt spirit will be quite as insane as it is on ptr now, so it will be possible to run out of mana as a druid :P
Spi/Int is only 50-100 MP5 boost... can people please get rid of the misconception that we have unlimited mana in 2.4?
Innervate should of given you not far off full mana if utilized well currently so the change on the PTR only means the same effect happens, or its slightly stronger while you can if you wish sit and spam HT(Max) on someone for no reason.
Blue Dragon? if your contemplating using that trinket and using HT (responsibly as in cancle/casting) it's somewhat counter productive, you want to be hitting every GCD - and why are you asking about a regeneration trinket if you think we have no mana problems in 2.4 ??
Well it is a little annoying that the blue posted what he did. If his druid is not performing well on the healing meters than maybe that should say right there that druids are not in any way in need of a PvE nerf.
You're absolutely right Playered. We won't have unlimited mana. However, we will be getting a huge increase in our total per fight mana pool. 100/mp5 for a 5 minute fight is 6,000 mana, if you never leave the 5 second rule. With expertly timed LBs just before the stack expires should allow you get get 1 second of OO5R every other cycle, if you are only hotting the MT. If you must innervate your self, according to the spreadsheet, I'll gain an extra 7,000 mana over 2.3 innervate(with my stats, 500 spirit, 460 int unbuffed). So, under fairly ideal circumstances, I'll be getting an extra 13,000 mana during a 5 minute fight.
Over the course of a 5 minute fight, if I'm only healing the MT, then I'll use:
11040 mana on lifebloom(184 mana * 60(number of reapplies during a 5 minutes fight)
8625 mana on rejuv(345 mana * 25(1 application per 12 seconds for 5 minutes)
10560 on regrowth(704 mana * 15(1 application every 21 seconds for 5 minutes)
______
30225 total mana
I start with 11,000 mana. I regen 22920 from the 382 mana/5 I should have in 2.4. 33,920 total mana in 5 minutes. I have an extra 3,000 mana, without potions, innervate or a shadow priest. Granted, this is a ideal situation, where I'm not raid healing or using swiftmend and only healing 1 tank. I'm not sure what the mana cost of swiftmend is(16% of what?) but you can easily add that total(used every 15 seconds) to the list above and see where it comes out.
If I remember correctly, a super mana potion is 100mp5, so thats another 6,000 mana, for a total near 40,000. What does a decent shadow priest return? Last Mag fight I looked at for me, the SP in my group gave me back near 20k over the 11 minute fight. Does about 150/mp5 sound about right for a 2/6 2/4 geared SP? So were looking at 49,000 mana in an ideal regen situation in a 5 minute fight. Counting innervate, thats 66,000(if by some chance you're able to get back all 17,000 of it). You could also add in mana spring(3,750 mana) and mana tide(24% of 11,000(right?) for 2640) if you've got a shaman in your group.
While it is certainly not unlimited, it is a huge mana pool to work from.
Remember, all these numbers are for an ideal situation. One tank to heal, all the mana you regen you're able to use and you don't lose out on some mana when you are full.
Last edited by grutak : 03/07/08 at 12:48 PM.
Reason: shadow priest numbers changed cuz i found the wws
While it is certainly not unlimited, it is a huge mana pool to work from.
Remember, all these numbers are for an ideal situation. One tank to heal, all the mana you regen you're able to use and you don't lose out on some mana when you are full.
If you're going to make an analysis regarding our mana potential, I would think it'd make far more sense to consider either the typical/average scenario or a worst-case scenario (maybe 3 rolling LB stacks and a rejuv/swiftmend option in the 4th cycle spot) than assuming best case scenario numbers. Granted, it's interesting to see the extent of our maximum potential. But it's when the going gets rough that mana is an issue, hence that's where the 'unlimited mana' statement should be challenged. If I get some spare time this weekend I'll try to adapt your example to such a situation (if no one beats me to it).
Spi/Int is only 50-100 MP5 boost... can people please get rid of the misconception that we have unlimited mana in 2.4?
Innervate should of given you not far off full mana if utilized well currently so the change on the PTR only means the same effect happens, or its slightly stronger while you can if you wish sit and spam HT(Max) on someone for no reason.
Blue Dragon? if your contemplating using that trinket and using HT (responsibly as in cancle/casting) it's somewhat counter productive, you want to be hitting every GCD - and why are you asking about a regeneration trinket if you think we have no mana problems in 2.4 ??
What I'm wondering is am I going to have any use for that trinket in 2.4, in a fairly typical scenario of hotting 3 tanks and using rejuvs/swiftmends in between, am I gonna get oom in 2.4 spamming pots every cooldown. I'm also planning on making the alchemy trinket (once I get the LW chest).
You're right Kalaghan. I should have, so I will. But I think I'll keep it independent of fight length and just use a cast cycle compared to a mp5 cycle.
I'll start by assuming a 6 second, full stack lifebloom on 3 targets with the 4th GCD being rejuv or sm, with every third extra GCD(every 18 seconds) being used for sm.
3*184 = 552 for lifebloom plus 357 mana(average of 2 rejuv's and 1 sm) for 909 mana per 6 second cycle. This is 757.6 mp/5. According to the new spirit formula, my new IS5SR regen will be 382(after buffs) with 500 spirit and 480 int unbuffed. With mana pots(100mp/5) and a shadow priest(150mp/5), my mp5 will be in the neighborhood of 632mp/5. I'll need to make up 125.6 mp5. With about 11,000 mana, it will take near 440 seconds to go OOM.
With the new changes, innervate will give back about 17,000 mana. If it is used the second you go oom, you'll have atleast another 11,000 mana to spend, not including the about 6 seconds you'll be sitting at full mana with the innervate buff still on. 11,000/17,000 * 20 seconds is about the mark you hit full mana.
So, thats 886 seconds untill OOM.
If you include improved mana spring totem(62mp5) and drums(25mp5), I'm down to needing to make up 38mp/5. I'm not sure how to calculate mana tide totem as mp5, but that is another 2600 mana.
I'm not the most math savvy guy around, so please correct me if I make a mistake.
Two things jump out at me. First, I think it's easier to just say that innervate will give you a full mana bar after 20 seconds. So you are at full mana at 440+20 = 460 seconds. Second, since 440 seconds > 6 minutes, you would have infinite mana under these ideal circumstances (since innervate will be back up before OOM if you start at a full mana bar and given that regen). Personally, I want to see if I can make do without a shadow priest and without my own innervate (which I suspect will be going to high int dps classes post 2.4) for most fights. That's where things will be getting difficult (potentially).
Edit: Definitely a much more useful and relevant analysis though. Don't mean to come down on you for it, it's certainly a help.
[quote=Darion;662556]6. Has anyone done studies of the percentage of a HoT that is 'used' during raid situations? ie: Druid throws a rejuv out. It ticks 4 times. How much of those 4 ticks is actually healing vs wasted due to the player getting healed to full by someone else?[quote]
I realize I'm a little late in responding to this, but I'm fairly sure there is a quasi efficient way of doing this if you are using a single resto druid in the raid.
In WWS parses it records "buffs gained", add up what you want (rejuv for instance) and that should be the number of rejuv's you cast in a fight. Now I say "should" because I'm unaware what happens if you overwrite your hots mid duration, pretty sure it doesn't count it. In anycase, its better than nothing.
Doing some quick calculations on last night's Hyjal run, came out with 43% overheal on rejuv with 84% of those rejuv's being on the tanks.
Depending on if this is a plausible way to gather overheal information I'll have more concrete numbers sooner than later.
In WWS parses it records "buffs gained", add up what you want (rejuv for instance) and that should be the number of rejuv's you cast in a fight. Now I say "should" because I'm unaware what happens if you overwrite your hots mid duration, pretty sure it doesn't count it. In anycase, its better than nothing.
Doing some quick calculations on last night's Hyjal run, came out with 43% overheal on rejuv with 84% of those rejuv's being on the tanks.
Depending on if this is a plausible way to gather overheal information I'll have more concrete numbers sooner than later.
Am I correct in thinking you're calculating it like this:
(effective healing by rejuv) / [(# of Rejuv buffs received) * (max healing per rejuv)]
?
If so, I think that gives a reasonable approximation for rejuv (it will still be off due to swiftmends, meaning the denominator is a little inflated). Unfortunately, lifebloom healing (which I think is the most important focus) is so different from rejuv in terms of both frequency and duration of HOT, so I don't think the result is transferable. And I suspect (confirmation?) that lifebloom would not get picked up as a new buff when a stack is refreshed?
03/05/08, 11:57 AM
Similar Threads
