Saturday, March 7, 2009

3.1 Lifebloom wrap-up.

.

EDIT: LB numbers towards the end of this post changed to more accurately reflect what my own LB numbers are self-buffed. =)

I got on the PTR one last time yesterday to test out rolling the new Lifeblooms under more raid-ready conditions (note: they fixed the Lifebloom bug! Now when your LB blooms, it gives back 50% to *you*, not your target. Woohoo!). My partner in crime, Vortexx (who graciously and without complaint respecced survival so I could rock out LBs with Replenishment up) and I met in Darnassus to test it out. After spending ten minutes trying to find a paladin in Darnassus, I gave up, hearthed to Dalaran, *finally* got someone to give me a Blessing of Wisdom, right before he got on a flight path no less, realized that it was only the 10 minute buff, frantically ported back to Darnassus, ran to the training dummies, and got Texx to start hitting things.

He kept Replenishment procced constantly, I had the 91mp5 BoW, MotW, and was in tree form. I was rolling two stacks of LBs, and ran out of mana in 3:21.

Under the same conditions, but with rolling two stacks of LBs, AND refreshing Rejuvs on both targets, I ran out of mana in 2:50.

After all is said and done, we're not going to be able to heal like we used to. Like Ghostcrawler said, we should be able to roll one stack of LBs on a player and still be able to raid heal. And I think we'll be able to do that.

I still do have a few more thoughts though.

1. What about new druids? Trees who are just starting doing 5 man heroics, say. They've got a pretty small mana pool and less spirit for an innervate and mana regen. I don't think that they'll be able to roll LBs on a tank for very long at all. Is rolling LBs on one tank going to be reserved for those who have over 1000 spirit and int? I would hope not. I'm not asking for it to go back to the way it was before - I understand why they're making this change, and I've resigned myself to it (yes, I know it's not officially live yet, but I'd rather get used to the idea now, rather than after the fact). I just hope that when Ghostcrawler says "You should be able to roll LBs on one tank and be able to raid heal", that this type of healing isn't reserved solely for the raiding elite.

2. If we put a 3 stack of LBs on someone, let them bloom, and then restack them immediately, I think that our overall raw healing numbers will be higher than if we just rolled them. I'm not putting mana efficiency into the equation right now, just sheer healing numbers. Let's compare the current way of healing (Putting 3 LBs on and then rolling them) to the new way (putting a 3 stack of LBs and then letting them bloom, followed up by an immediate placement of 3 more stacks).

Our numbers: One LB ticks for 400, two LBs tick for 800, three LBs tick for 1200, on both the current server and the PTR. The bloom hits for 9300 on the PTR. Let's also assume that you have the glyph of LB, so each stack rolls for 10 seconds.

Healing numbers on the current server (rolling LBs):

Secs: 1 2 3 4 5 6 7 8 9 10
Heal:400 800 1200 1200 1200 1200 1200 1200 1200 1200
LB1 LB2 LB3 tick tick tick tick tick tick refresh


(And yeah, I know that in the first 3 seconds there, the global cooldown is typically more than one second, but we'll just say it's one for ease of math - the numbers will be a little bit off, but close enough I think)

After that round, it just stays at ticking for 1200 each second. Let's say we do this for 2 more rounds, so 30 seconds total. So we have (400 + 800) + (1200 * 8) = 10800 for round 1, and then (1200 * 20) = 24000 for rounds two and three, for a total of (10800 + 24000) = 34800 healing done.

Healing numbers on the PTR (letting the 3 stack bloom):


Secs: 1 2 3 4 5 6 7 8 9 10
Heal: 400 800 1200 1200 1200 1200 1200 1200 1200 9300
LB1 LB2 LB3 tick tick tick tick tick tick bloom


And then this is done three more times for a total of 30 seconds.

(400 + 800) + (1200 * 7) + (9300) = 18900

18900 * 3 = 56700.

Old way of healing, after 30 seconds: 34800 total healing done.
New way of healing, after 30 seconds: 56700 total healing done.

Looks like that bloom really does make a big difference.

Obviously there are issues of overhealing (is the bigger bloom really worth it?), and issues of mana efficiency (can I keep up restacking those LBs *and* throw other heals around for an extended period of time?), but overall the total healing done is far and away better than rolling LBs on the current server.

I think it's going to take some getting used to, however. Vortexx and I went to two-man Onyxia, partly to test out my new healing style (letting LBs bloom), and partly for fun.

My mana seemed to be ok once I let the LBs bloom and got half of it back. I was throwing out some raid healing as well, some rejuvs, regrowths, and WGs (even though Texx and his pet didn't really need them) and I ran out of mana quicker than I'm used to. But overall it wasn't like /castcastcast OMG OUT OF MANA. It is REALLY hard to let those LBs bloom, though. All my instincts tell me to refresh them. I even caught myself doing it a couple times. The urge to refresh them is a freaking force of nature, let me tell you. Guess we'll just have to get used to it.

.

7 remarks:

Syll said...

wow great analysis! This gives me hope. I think that it is going to be a fun challenge to remember to evaluate bloom or roll on a tank when the lifebloom nears the end. I'm sure we'll all incorporate this as an option and press onward. I second your concern for the littler trees coming up, though. I hope that it's not too hard for them to remain effective before their mana is snuffed out.

Thanks again for visiting my blog today! I'm adding your site to my blogroll. It's wonderful!
Sylly

Keeva said...

I was just talking to someone last night about how it is instinct to refresh LB just in the last moment; it's almost an autopilot function now.

That is going to be a very hard habit to break.

Peli said...

Only problem I see is that, especially if raiding is anything like it is today in T8, our LB Bloom will be 100% overheal and it's almost impossible to time the bloom so that it isn't overheal since hit happens 10 seconds in the future.

The LB change is going to be a HUGE hit to tree raid performance in an effective HPS format.

Bell said...

IMHO, the new style will be a combination of rolling and blooming; roll when your target does not need the bloom, let it bloom when they do.

Anonymous said...

It would be interesting to know how your mana lasted when letting the stack bloom compared to keeping it rolling. It would be nice to think we have the option of rolling or blooming, but the mana return of the bloom may oblige us to always go for the bloom.

Personally, I'd prefer Blizzard left this alone and removed Replenishment from the game instead.

Unknown said...

Actuallly I do see one problem with your analysis. after the frist 3 stack, it whould be cheaper to refresh it and keep rolling it then to let it bloom and the immediatly restack it. Even if alll of the bloom is non overheal. Math as follows.
Once you have three stacks ticking you get 1200 MpS if you refresh it it cost you X mana and you get another 9s (10 ticks minus one as you refresh it) of 1200 Mps
9 x 1200 = 10800 hp for once cast, or X mana.

if you let it blooom you get a 9800 hp heal and you get 3xX/2 = 1,5X mana back. You then speand 3X mana to restack it, netting you a 3X-1,5X = 1,5X mana cost.
Helth healed thgouh is alot greater
since you get 9800 +400+800+(9x1200) = 21800hp healed for 1,5X mana.

Conclusion: Mana per time wise it is still cheaper to keep rolling then to immideatly restack. But if the Blooom can be used then it's much more mana per health wise effective to let it bloom.
the only way to come close to todays manacost is to not fast stack it, but rather to build your three stack up over 18 sec, to geet the msot out of each cast helthwhise before refreshing it.

Anonymous said...

To bad those "blooms" will be wasted and 8 out of 10 times be overhealing.

The only reason we are going to do it is to get some of the mana back or hell is freezing over and the tank just happens to need a bigger heal.