Saying that druids lack in raw HPS isn't exactly accurate. Druids have excellent HPS when healing periodic raid damage via rejuv - around 12K-14K at the top end. Weave in WGs and you can hit 16K. This style of healing is also very sustainable, especially with the RJ idol. Likewise, paladins have excellent single-target HPS - lets say 10K HPS that is sustainable for a pretty long time. Factoring in beacon isn't trivial as it can very rarely live up to it's potential of doubling the healing output. The HL glyph is just OP - a glyph shouldn't account for 15-20% of healing done as it does now in say a naxx clear.
Keep in mind that the HPM of 3.1 LB and HL on a single target are more of less the same. I think the main gripe with paladins is that they can spam HL uninterrupted, and I'm not aware of any change in 3.1 that affects that.
Yeah you're right that it's at least something substantial, but there are problems. The first is, as I said earlier, clipping and falling off. So if you have 240 ticks of lifebloom then do you assume that you have cast lifebloom 240 ticks/8 ticks per cast= 30 casts? If you discount clipping, how much can the value be trusted? If you say that you only clip an average of .8 ticks per cast, how sure can you be that that is a good estimate, considering the fact that the average clipping of lifebloom has never been quantized, and would likely vary largely from player to player and fight to fight.
Maybe something changed in Beta that I missed, but there was testing in the old thread that proved pretty conclusively that LB isn't clipped. It works like Lacerate, in that once you have it rolling, casting it again only adds to the time remaining, but doesn't change the time between ticks (once per second).
You bring up a good point though, how far behind perfect efficiency is a good player? To me this nerf feels like someone at Blizzard asked "how well does LB work if all the ticks heal for full and you roll it on multiple tanks?", then panic ensued. I think we all know that the reality of LB output is no where near this fantasy setting.
To get a good idea of true LB efficiency for a player the following variables would need to be parsed:
How many 1 stack ticks should have happened.
How many 2 stack ticks should have happened.
How many 3 stack ticks should have happeded.
How many times was LB cast.
How many times did it bloom.
Some of these might be ignored, but this would give us all needed information.
Maybe something changed in Beta that I missed, but there was testing in the old thread that proved pretty conclusively that LB isn't clipped. It works like Lacerate, in that once you have it rolling, casting it again only adds to the time remaining, but doesn't change the time between ticks (once per second).
I would very much be interested to see the original logic for this conclusion, because it goes against the mental model I have of the spell. It makes sense, to me, that if you refresh lifebloom 3 seconds before expiration then you are using mana at a rate higher than is necessary if you only refresh it shortly before the blooms. HPS would remain the same, as would overheal, but the de facto ratio of how much healing we pull out of the spell per how much we spend on it very much depends upon clipping, and since efficiency is what is being targeted in this nerf then I think clipping is very important.
I raid with 200 to 250 ms ping time due to Comcast being evil to me.
Many times I refresh my triple-stacks with two or even three seconds left. There are far too many time where if I try to squeeze in an extra Nourish or something I will lose a LB stack. So I usually choose to refresh early and keep the stack going than risk losing it. Patchwerk with 3 tanks comes to mind as being the easiest to try to run stacks down on time without losing them, but sometimes framerates and lag cause issues, and that is a bad fight to lose a stack on. Fights with lots of movement and dodging, like Sarth3D, are very easy to lose LB stacks without refreshing them too early.
I would very much be interested to see the original logic for this conclusion, because it goes against the mental model I have of the spell. It makes sense, to me, that if you refresh lifebloom 3 seconds before expiration then you are using mana at a rate higher than is necessary if you only refresh it shortly before the blooms. HPS would remain the same, as would overheal, but the de facto ratio of how much healing we pull out of the spell per how much we spend on it very much depends upon clipping, and since efficiency is what is being targeted in this nerf then I think clipping is very important.
Ok, I miss understood your use of the word clipping to be more like what happens with Rejuv (if I clip that last tick of a RJ 1 second before it ticks, then there will be an effective gap of 5 seconds between ticks). So there is a double wammy for clipping RJ, hurting both HPS and HPM
LB will always tick every 1 second regardless of refresh time, so clipping LB does hurt our HPM, but not HPS. Sorry for the confusion.
Back to the actual Lifebloom nerf itself. I believe that I’ve found blizzards intent, or at very least a way to make it work for us. Let’s start with some basic facts and assumptions that we know about The changes to Lifebloom itself.
1. They are doubling its cost from live
2. Whenever Lifebloom blooms it gives half the BASE mana back (As confirmed by Ghostcrawler, this is intended)
3. When the bloom portion on Lifebloom is now multiplicative. (We can now expect it to hit for 3 times as hard when a 3 stack falls off).
Currently Lifebloom costs 489 base mana (Easily verified by taking off all T7 when out of ToL), the Cost to cast is now going up to 978. Tree of Life Form reduces the mana cost by 20%, and currently our T7 gives another 5% reduction. Currently in live this gives a cost to cast of 366 mana (732 with the proposed changes). 732 mana with half the cost of Lifebloom returned (489) gives an effective mana cost of 243 per Lifebloom that is allowed to bloom. Obviously it is in a druids best interest to now let a Lifebloom expire (when looking at a mana cost only standpoint).
But slow stacking for obvious reasons has its limitations. Druids (and healers in general) prize consistency. It does me little good to have a Lifebloom tick for 400 for 10 seconds, 800 the next, and 1200 for the final 10 seconds. Not only does this cut Lifeblooms effective healing by a third, it also leads to a swing of 800 HPS. This is however extremely mana efficient.
No, I believe the key will rely on fast stacking Lifeblooms to a 3 stack and then letting them expire. This works for a variety of reasons. First, there will be little to no loss in actual healing done from Lifebloom between its current state and my new 'Blooming' model. To make things easy I’ve used rounded numbers. Lets suppose a Lifebloom currently ticks for 400 per application. Using a pattern of LB, GCD, LB, GCD, LB, 10 seconds, expire (Repeat), and assuming the haste cap is reached (1s GCD for Lifebloom), we can assume that we will have 1 Tick of 400, 1 tick of 800, and 10 ticks of 1200. Averaged out over the course of 12 seconds this leads to 1100 HPS, or an 8.3% loss in healing done. However, we are also gaining a sizeable bloom every 12 seconds as well. At 1200 healing done by a Lifebloom tick, one could expect a Bloom to hit for around 3k, this will be tripled when 3.1 hits, so we can expect a non-crit bloom to be around 9000. While I expect most of this to be overhealing, I also feel that the Bloom (which doesn’t hit in a rolling style) could easily make up for the 8.3% loss, especially considering the bloom will hit right before you are required to restack.
Mana Costs:
I used the following formulas to estimate the mana costs per Lifebloom cycle.
T = Time in seconds of the encounter
1. Live: 3 x 366 + 366 x (T - 3)/10 = Total Mana Costs
2. 3.1: 3 x 732 + 732 x (T - 3)/10 = Total Mana Costs
3. Blooming 3 Stacks (Fast Stacking): 243 x 3 x T/12 = Total Mana Costs
Then for fun, I went to compare these mana costs to both what we use on live (what we currently use) and a rolling style given the 3.1 mana costs. I simply divided the results from equation 3 by the results from equation 1 (or 2).
Minute..........Cost Compare to Live..........Cost Compared to 3.1
For short encounters as you can see, the change is fairly minimal. At longer encounters you can expect a mana increase of about 60% from live, but as my 3.1 comparison clearly shows, this will end up being a much more efficient way to stack Lifeblooms.
Further Musings
The actual increase in mana costs are actually than a simple model would predict. This model has no way to account clipping, but instead assumes that you are refreshing a Lifebloom stack at exactly the right moment. Personally, I would assume an average of 8 or 9 seconds would be where most druids refresh Lifebloom. On the 'Blooming' model, there is no clipping, since it by design allows the Lifebloom to reach its full duration.
We also gain a minimum of a 9k blooming effect every cycle, I feel this is not to be ignored. While i agree that it will largely lead to overheal, it certain situations it will be found to be very powerful indeed.
And as already mentioned, while raid healing is not the best use for Lifebloom, who can honestly say that it isn’t used for that from time to time, Lifebloom is now 123 mana cheaper than it was to use as a raid heal, that will add up over the course of a fight.
The largest weakness to this style is that it requires 2 more GCD's than a current 'Rolling' one (3 GCD's over 12 seconds as opposed to 1 GCD over 10). This leads to a loss in healing due to opportunity cost obviously. Further more it also limits how many tanks we can effectively heal using Lifebloom. If 3 out of every 12 GCDs are being used per tank, a maximum of 4 tanks can be healed using LB, at 3 tanks we can keep up a Full Stack of LB and Rejuv, and so on. When you consider all these facts, this Nerf is brilliant on Blizzards part. They've managed to significantly increase the cost of keeping Lifebloom on a tank (between 35% and 60% on an average boss fight) while building in a soft cap to the amount of tanks you can have a full stack on.
I'm not saying this is a perfect model, but it is surely something to think about.
(Edit) I forgot to meantion that these calculations obviously assume a glyph of Lifebloom. I realize that all druids may not have this, but I do. A 10 second duration seems like second nature to me now. Also, My table looked horrible. I had to fix it.
Last edited by Allinone : 03/03/09 at 5:40 PM.
Reason: Made the table look pretty, Edited out an incorrect value in equation 3
I would say this is a very interesting change, especially with regards to raid healing, I expect I'll be rolling stacks still, but the large optional heal + mana return will make the decision, of let it bloom be something other than bad.
1. Live: 3 x 366 + 366 x (T - 3)/10 = Total Mana Costs
2. 3.1: 3 x 732 + 732 x (T - 3)/10 = Total Mana Costs
3. Blooming 3 Stacks (Fast Stacking): 244 x 3 x T/12 = Total Mana Costs
These formulae should be:
1. Live: 3 x 366 + 366 x (T - 3)/9 = Total Mana Costs
2. 3.1: 3 x 732 + 732 x (T - 3)/9 = Total Mana Costs
3. Blooming 3 Stacks (Fast Stacking): 1.5 x 732 x T/12 = Total Mana Costs
For formulas 1 and 2, the tenth tick of lifebloom, if I remember correctly, goes off at the same time as the bloom, therefore you would refresh before the interval reaches 10 seconds. For formula 3, it appears to be a small arithmetic error that can be more easily seen by doing this.
I reserve the right to be completely wrong, so please correct me. Though this shows an increase in cost, efficiency of the bloom method versus 3.1 rolling would be higher, though I have doubts about relying on the bloom.
I think if you are smart about watching when the tank is below full health, and letting the 3 stack drop then, you can do better than the HPS you guys calculate. Nor is this an unreasonable cognitive load on a good druid (who already generally watches a ton of numbers counting down, while dodging firewalls, whelp add waves, and fissures). If Blizzard disagrees, and thinks it's "too much" then refunding mana on a bloom is a very puzzling mechanic to introduce, because dropping stacks "cleverly" is exactly what this change is encouraging us to do.
I think it would have been a much cooler nerf if they increased the mana cost of lifebloom, but gave a mana refund based on refreshing stacks -- the longer you refresh the bigger the refund (up to a point). You could balance this to make lifebloom less efficient, while still rewarding good play (I think refreshing stacks in a timely way is good play, and Blizzard agrees that rolling lifebloom stacks is part of core druid gameplay and they don't want to change that. Current nerfs are making us try everything BUT.)
I would very much be interested to see the original logic for this conclusion, because it goes against the mental model I have of the spell. It makes sense, to me, that if you refresh lifebloom 3 seconds before expiration then you are using mana at a rate higher than is necessary if you only refresh it shortly before the blooms. HPS would remain the same, as would overheal, but the de facto ratio of how much healing we pull out of the spell per how much we spend on it very much depends upon clipping, and since efficiency is what is being targeted in this nerf then I think clipping is very important.
I think this should be fairly easy to test and I'll give it a shot if I remember when I get home. Take off most of your spell power gear and just have a warlock lifetap while you cast lifeblooms on him. If you get a tick that is more than ~1.1 seconds apart then this should be conclusive evidence that you do indeed clip ticks. Certainly if you ever observe a 1.9s gap you are clipping ticks. The 1.1 is just due to it not really getting recorded at exactly 1.0s intervals.
Let me know if there are any flaws in this reasoning that should be adjusted before I give it a shot.
If Blizzard disagrees, and thinks it's "too much" then refunding mana on a bloom is a very puzzling mechanic to introduce, because dropping stacks "cleverly" is exactly what this change is encouraging us to do.
Dropping stacks may or may not be optimal now, but it sounds like they didn't intend for it to be:
Originally Posted by Ghostcrawler
We added the bonus to the bloom to improve its use in PvP, where you can't always keep 3 stacks rolling, and for say throwing a single spell on a rogue where you don't plan on keeping it rolling.
[...] We think rolling is a fun part of the spell. It's just too efficient and makes Lifebloom the best heal per second and heal per mana. We don't want you to change the way you use the spell, at least in terms of single tank healing. The bonus on a bloom or dispel is just supposed to be a bonus.
If the numbers aren't right, we'll continue to tweak them. It may be that the cost needed to go up by 75% instead of 100% for example.
He does however go on to say that the letting it bloom will "give you a new healing style" that's more mana efficient at the cost of having the hot up 100% of the time.
Inform your dealers and whores of my credit, and pour me a goddamned drink!
I was using the 243 effective mana cost per Lifebloom if it is allowed to bloom (as stated in Paragraph 2), and going with a simple assumption that you would be refreshing it every 12 second. 732 (The Cost of Casting Lifebloom while in Tree Of Life in 3.1) less half the cost of our base casting cost (Which is factored before talents, set bonuses, or talents). This leads to
732 - 489 = 243
For all intents and purposes, if you are to assume that you will never stack more than 3 Lifeblooms then this is what could be assumed to be your effective cost of casting the Lifebloom Spell. While I understand that it will cost the 732 to start each stack, the total mana used ends up being 243 after factoring in the returns on the base mana. So a simple equation of
243 x 3 x (T/12) = Effective Mana Used
Seems justified.
1.5 x 732 x T/12 = Total Mana Costs
does not take into account the additional mana returned due to ToL or at this point our T7 2 Piece Set Bonus
You are more than likely correct about your assumption for 9 seconds as opposed to 10, although when using gridstatuslifebloom I’m fairly certain that I see it range in numbers from 9.9 to 0.0 when counting down. Again, this is a fairly simplistic model, in real life application I would expect Lifebloom to realistically have 1 or 2 seconds clipped off its timer.
In either case, using 9 instead of 10 for equations 1 and 2 leads to
Minute..........Cost Compare to Live..........Cost Compared to 3.1
To take into account clipping, let’s assume 8s is used instead of 9s. This is the closest to a real world scenario that I would expect, and only takes into account clipping 1 second off Lifeblooms while refreshing the stack. Again, there is no such thing as clipping a blooming stack, so the more prone you are the clipping a stack of Lifeblooms, the less the impact this nerf would have if this style was adopted.
Minute..........Cost Compare to Live..........Cost Compared to 3.1
My raiding experience is raid and OT healer on 10 man. I roll LB on multiple tanks. I can't recall ever rolling LB on more than two people. Moving away from double-roll seems odd to me, as I don't do this because it's mana efficient -- I do it to buffer my relatively weaker reactive heals.
I've tried to figure out -- if LB was receiving an HPM *buff* rather than (convoluted) nerf in 3.1, would it change the way I used LB? If LB cost 1 mana, would I use it more? I then tried to think of how my heroic and 10-man healing in 3.1 if this change goes through. As far as I can tell, my spellcasting would not change if LB cost one mana.
On live w/ LB=1 mana: On MT I roll LB and RJ, and use regrowth preferably over nourish. I follow the typical RJ-on-light, RG-on-heavy model used by most. On light damage, I use RJ. On large damage, it's usually a regrowth bomb. On pressing but steady damage to all group members, WG then LB if WG is on cooldown -- these are left to bloom. In between these, I don't want to throw away mana or GCDs in case they're needed. There are some exceptions to single-target rolling, when both MT and OT are taking serious damage, but these fights are not numerous. I would not roll LB on additional people, even if it cost only 1 mana (due to single-target throughput).
On 3.1 w/ LB=1 mana: As far as I could tell I would be doing the same thing, with the change that I would be using nourish much more as due to nourish, RG, and LS changes. LB use would go unchanged in most circumstances, even if it cost only 1 mana, for the same reason as above.
On 3.1 w/ LB HPM nerf: Once again, as far as I can tell, in most situations my routine is not going to change in most situations, except for increased nourish use. In fights like Patchwerk or other double-damage fights, I cannot tell whether I would continue to roll lifeblooms and suffer a reduction in healing elsewhere, or if I would roll only on the one and rely on more direct-healing spells. It seems to me that it depends strongly on whether my role as a healing-buffer is more valuable than my role as a direct healer. I'm worried that I would be asked to keep up rolling LB to smooth out damage for other healers, but fewer RG and nourish.
So at the end of the day I'm concerned that this is simply a "too clever" change that will not have its intended effect. I'd prefer a simple 10-20% increase in mana cost rather than convoluted mechanics and decision making of bloom-or-no-bloom and the resulting yoyo for a mana bar when blooms strike. The whole thing reminds me of the hoops hunters jumped through to get shot-rotations working optimally before auto-shot became decoupled from specials.
One final note: If HPM is the main concern with LB, then why would the glyph be specifically designed to both increase the HPM *and* make multiple-tank rolling simpler? I've read that they don't take glyphs into account when balancing spells ... but I don't see how they're decoupling them here.
If Lifebloom cost 1 mana, I would glyph lifebloom, talent GotEM and Celestial Focus, reach the haste cap, and roll 3 stacks of lifebloom on 10 targets. It's pretty easy to see this would lead to ridiculous results (certainly in 10 man raids). I guess even with 3.0 costs, this is sustainable if you put some thought into it. This is probably what prompted the nerf. It's like imagine if Paladin's Flash of Light hit 10 targets, but cost the same.
Further Musings
The actual increase in mana costs are actually than a simple model would predict. This model has no way to account clipping, but instead assumes that you are refreshing a Lifebloom stack at exactly the right moment. Personally, I would assume an average of 8 or 9 seconds would be where most druids refresh Lifebloom. On the 'Blooming' model, there is no clipping, since it by design allows the Lifebloom to reach its full duration.
We also gain a minimum of a 9k blooming effect every cycle, I feel this is not to be ignored. While i agree that it will largely lead to overheal, it certain situations it will be found to be very powerful indeed.
You do gain a 9k bloom every 12s or so. However you also use 3 GCD's on Lb every 12s. That compares unfavorably to live's single GCD every 9 seconds (once the stack is established). You may be better off with a slow stack (which is more mana efficient), and then using your extra mana and GCD's for Nourish (particularly during the period when you only have one or two Lb stacks).
I haven't run numbers on this, but it seems worth investigating.
If Lifebloom cost 1 mana, I would glyph lifebloom, talent GotEM and Celestial Focus, reach the haste cap, and roll 3 stacks of lifebloom on 10 targets. It's pretty easy to see this would lead to ridiculous results (certainly in 10 man raids).
I know that it seems that way at first, but it's unlikely. You'd have no time for patch up heals, and the HPS on a single-target would be insufficient. I could not serve as a raid healer if all I did was add a 400 hps buffer, and nothing else.
In addition, the pattern of setting it up, even if you assumed zero lag and perfect haste cap, would take too long and become unwieldy (I mapped it out the first 25 casts as A,A,A,B,B,B,C,C,C,D,A,E,E,B,E,F,C,F,F,D,A,G ...) ... you can see it gets very complicated very fast.
Perhaps in some cases, you could get away with rolling lifeblooms on many targets, but for the vast majority of cases, there are better options. RJ and RG can be swiftmended, LB cannot. WG heals multiple targets on AE damage and in 3.1 can proc mana/similar heals, LB does not (since you can't be sure you've got rolls going on the right people). RJ and RG do better jobs at sudden large spikes of damage.
I am not proposing LB cost 1 mana. I took it to an extreme to see if things would change. I don't see that it would, due to GCD constraints and suitability.
I know that it seems that way at first, but it's unlikely. You'd have no time for patch up heals, and the HPS on a single-target would be insufficient. I could not serve as a raid healer if all I did was add a 400 hps buffer, and nothing else.
In addition, the pattern of setting it up, even if you assumed zero lag and perfect haste cap, would take too long and become unwieldy (I mapped it out the first 25 casts as A,A,A,B,B,B,C,C,C,D,A,E,E,B,E,F,C,F,F,D,A,G ...) ... you can see it gets very complicated very fast.
Perhaps in some cases, you could get away with rolling lifeblooms on many targets, but for the vast majority of cases, there are better options. RJ and RG can be swiftmended, LB cannot. WG heals multiple targets on AE damage and in 3.1 can proc mana/similar heals, LB does not (since you can't be sure you've got rolls going on the right people). RJ and RG do better jobs at sudden large spikes of damage.
I am not proposing LB cost 1 mana. I took it to an extreme to see if things would change. I don't see that it would, due to GCD constraints and suitability.
Two things:
(1) Check how much HPS lifebloom does at 3 stacks with reasonable gear (2.3-2.4k spellpower). It is not 400 HPS. It is more like 1500 HPS. Sustainable 1500 HPS on 10 targets is broken. (Of course someone earlier in this thread pointed out that Rejuv + WG can hit 16K HPS on the raid.. so perhaps we are broken anyways even without lifebloom). I agree that Rejuv + WG + Regrowth is a better strategy due to swiftmend and better spike responsiveness, but lifebloom would be free. Free is a pretty powerful thing, so people would opt for free, and bring enough resto druids to trivialize raid damage.
(2) If you want to roll a lot of lifeblooms there is a very simple pattern for doing it, just go down the list of 10 and put up 1 stack, then go down the list and put up the second stack, etc.
1500 HPS on 10 targets is only broken if those targets are all in need of healing. However, on most fights, only the tanks take sustained damage--the rest of the raid takes damage at random in large bursts. The thought-experiment of Lifebloom costing 1 mana would certainly lead to fantastically easy Sapphiron, but for a lot of fights is not that much of a bonus.
I haven't used Lifebloom on anything but the MT for quite some time now. Even in Sarth 3D, I'm assigned to the Sarth tank, so I'm pulling out all the stops to keep him up. At most, I throw a Rejuv on the other tanks so they can be Swiftmended if I'm in range.
Rejuv has been a better raid heal for quite some time now, even after they removed the 2t7 bonus. Lifebloom is really only effective in a triple stack, and even then, it's rarely more than 30% of my healing done.
I'm actually very disappointed with the changes to Lifebloom - I was under the assumption that Blizzard liked where Druids were in terms of spell usage, much like they're pleased with Shaman mana regen. We use all of our spells fairly equally, and depending upon the fight Rejuv and Lifebloom vary between 20-30% of healing done while Regrowth, Nourish, and Wild Growth go between 10-20%. We really have a balanced toolbox, and Lifebloom certainly wasn't the crux of our mana "issues."
Unless I missed something in these threads, Lifebloom isn't generally used for anything but tank healing. Now that we have a lot of spare GCDs, I personally have been tossing out Rejuv and Regrowth to fill the ten seconds between refreshes. If we're still not running out of mana in Ulduar, perhaps the mana cost of Rejuv or Regrowth should be looked at. Without the Rejuv Idol I can spam it through Saph and Maly without mana issues. Now that Regrowth has been balanced, it will likely be worth the "fire and forget" heal that it traditionally was and nothing more, while Nourish will now fill the role of our direct heal when Swiftmend or patience can't cut it.
I think, in order to make lifebloom good again (let's be honest the only "real" use of lifebloom is the rolling part) is to allow us to have some sort of control over when the bloom happens. Otherwise it's just going to go into overhealing. I'm still not sure why they just didn'T limit the amount of targets we can have lifebloom on instead of that change. Why not "you cannot cast lifebloom on more than 2 targets" meaning if you cast it on a third target the first lifebloom blooms. That would nerf the rolling lifebloom, buff lifebloom for pvp and still give us a reason to use it in pve imo.
Sorry about that, Allinone. I realized my mistake on my way to class , though at least I was half right.
I find that this is becoming much too complicated too quickly. Just look at all the vocab that has been invented in the past 3 pages in reference to this nerf. Slowstacking, faststacking, bloom method, etc. Blizzard's reasoning for the LB nerf is really faulty, in my eyes:
1. Buff in pvp
Now lifebloom use in pvp is becoming about as bursty as the damage is, now that it's detrimental to roll a stack. So now we have the pleasure of hoping opponents dispel us at the right time or that 10 seconds before a big hit we happened to have put up a lifebloom.
2. Make a stack falling off less of a penalty
They accomplished this. Unfortunately this also makes rolling a stack more of a bane than a boon. A complete reversal.
3. Make rolling on multiple tanks less awesome
They accomplish this by making the blooming method the only way that it will work at current regen levels, which causes three times as many gcd's used in a cycle, resulting in greater difficulty and time investment in rolling on multiple people.
4. Make it more raid heal friendly
They made it more efficient, but that's not exactly going to make it rival WG or even RJ. To steal the example from someone else, no one will use it as a main raid heal even if it cost 1 mana just because it would likely be sniped and it's too slow to bloom.
It seems pretty clear that it is significantly better to bloom stacks, and that rolling is worse in hps(maybe) and hpm. They say that they want us to have to make a more difficult choice when choosing to roll on more than one tank, but it has become an unreasonable one. We gain the "choice" of using 3 gcd's every 12 second window per tank or we can "choose" to roll them on the tanks instead and spend twice as much mana for a huge loss in efficiency, much as we have the choice of using unglyphed healing touch over regrowth and nourish for tank heals. Even if they tweaked the exact numbers behind it to the point where the choice between rolling and blooming was not so cut-and-dry, do they really think this is an elegant solution? This spell stacks up to 3 applications and ticks once every second while refunding half of its base cost when it blooms, where the bloom is multiplied by the number of applications and rolling it is penalized by clipping and other human mistakes. Is there any spell that is even half this complicated? I think they need to go back to the drawing board on this one and consider putting LB in a curse category so we can't roll it over 2 stacks or whatever, because this implementation is clunky and unwieldy. I would rather they fiddle with the coefficient on the bloom, or increase the base heal of it while increasing the cost by 40%. Speaking of which, it seems apparent to me that though they say they have "...done a lot of calculations on this..." that they really haven't both because they increase it by 100%, which seems too simple a number to have a significant mathematical basis, and also because they would have found out that they're making lifebloom bloom oriented instead of hot oriented, though they say they still want us rolling it.
Further Musings
We also gain a minimum of a 9k blooming effect every cycle, I feel this is not to be ignored. While i agree that it will largely lead to overheal, it certain situations it will be found to be very powerful indeed.
While I appreciate the extra value on the bloom, I think that it would be virtually all overheal, if not all. At the moment my lifebloom heals 50% overheal. My experience is that virtually all blooms overheal by some amount. So any amount added to the bloom is simply going to add to overheal.
The exception, of course, is if you game the bloom. I'm unlikely to try to game anything more than 4k health ... which is just slightly more than my bloom now. I don't think I'd be willing to game someone who is down 8k just because my bloom heals that much. The very rare occasion where this would come in to play would be where I'm about to do a patch up heal and notice that LB has 1s left. I think this would be so rare as to negligible in the evaluation.
I am in agreement with Paininabox regarding complexity. I am very worried about my GCD situation and my spell "rotation" should this change come about.
re: Controlling the bloom -- if swiftmend triggered the bloom, would that be better? worse?
Rijndael:True, I was a bit close minded about the order it be applied, and my number was wrong (I copied a 1 stack, not a 3 stack for me). Sorry about that. I was trying to say more what Paininabox, GTthebard and Jurik have said: solely rolling lifeblooms is not a good way to heal -- maybe broken in some fights, but only a few. It's just a though experiment as to the value of lifebloom HPM, and not meant to be taken as a serious suggestion.
Another "interesting" post by GC about the Lifebloom changes in general: There is an obvious flaw in his logic on #4 but please save the ranting about it for wow boards/friends/family and not this thread.
Originally Posted by GC
1) I am still reading. I read all of these threads.
2) I would never suggest trying to turn length of a thread or attitude of the posters into any kind of meaningful statistic (as some posters suggested). Inevitably, the thing that will persuade us is the strength of your arguments.
3) Here is the standard caveat not to post insulting, mindless or uselss replies that will get you banned.
4) Remember our stated goals. In PvE the goal was to make rolling Lifeblooms on tanks more expensive. We didn't nerf the amount healed by the spell. Currently, rolling LBs + Rejuv on a tank can be as much healing as a priest or paladin who is just spamming all of their heals on the tank. That is a totally fine thing for a druid to be able to do. But currently the druid can do all of that on more than one tank AND also add in some raid healing. That is just too good. We're cool with rolling LBs on a tank and raid healing. We're cool rolling LBs on two tanks. But once you are rolling LBs on 2 (or 3!) tanks and have enough mana left over for a lot of other raid heals, then you are just too efficient a healer. I understand not all of you were able to do that. But many players were.
5) In PvP we realize that the goal may be to keep rolling LBs on as many players as possible, but realistically there are lot of situations where you are going to fail to do that (dispels, CC and LOS to name a few). In those cases, LB will now heal for more and your overall cost for Lifebloom will be about the same as it is now (and it might possibly be a buff in tree form). For those druids who really could keep 3 LBs rolling on multiple targets in Arenas, it will be a nerf, but it is probably a nerf that was needed anyway from a mana-per-heal standpoint. We think Resto druids are under-represented in Arenas at the moment. We don't think the way to improve that is by letting Lifebloom rolling heal for an amount we feel is unbalanced.
6) Once we get the PTR updated, get out there and try it out.
If Lifebloom cost 1 mana, I would glyph lifebloom, talent GotEM and Celestial Focus, reach the haste cap, and roll 3 stacks of lifebloom on 10 targets. It's pretty easy to see this would lead to ridiculous results (certainly in 10 man raids). I guess even with 3.0 costs, this is sustainable if you put some thought into it. This is probably what prompted the nerf. It's like imagine if Paladin's Flash of Light hit 10 targets, but cost the same.
Yes but it would generally be better to lifebloom the tanks, as well as put rejuv on them, because the damage is focused at the tanks. Additionally lifeblooming 10 targets isn't that feasible on fights with movement, because you have to be in range of your targets basically at all times for it to work. I don't see this working on 3D, which is what comes to mind as a fight with movement.
I don't think the mana thing will be a killer, and the multiplied bloom will really be a boon to healing tanks during healing lockouts (i.e. enraged web wrap on a hypothetical boss that hits twice as hard as maexxna), something that's already kind of a druid niche. Allowing your bloom to go off could be a great source of burst healing now that it's been made a somewhat viable tactic. This is especially true if you have a timer that tells you when a big burst will be needed. You could single handedly top a nearly dead tank off from an unbalancing strike type attack with a nourish -> swiftmend + bloom in a very short time frame.
03/03/09, 6:31 AM
