Aye, it's difficult to say here. From parsing the log, your gross Rejuv tick in your raid-buffed state is 2,926 and your gross Rejuv crit is 4,389. That said, at this moment in your run (live logs so this will change), your average RJ effective crit value is 634 and your average RJ effect tick is 487 This is 1.3 x on effective crit value. If I used my prior formula, I'd get a ~3.4% effective healing provided by the T9 4 pc RJ crit bonus:


((avg effective crit - avg effective tick)*number of crits)/total effective healing

((634-487)*923)/3,954,120

But, it's very tough to say how valid that is. Basically, unless your periodic RJ is able to heal for more than 2,926 on any given interval, the crit didn't buy you anything. Without parsing every individual application of RJ, I'm not going to be able to determine this. I have to go with the average crit wouldn't be higher than the average tick unless there were cases where you realized some value on crits above and beyond your normal ticks. That said, it's entirely possible that just the randomness of your crits hitting more often when there was something to heal versus your normal ticks hitting at full health. Any of a variety of things could impact that. It's definitely going to be hard to model accurately even with live data. I have my perfect-world formula fairly locked down but the live data is fairly challenging to say the least.

I'll have to give it more thought but we are about to start ToC as well (west coast time).

Don't really understand how that changes anything at all. It just means our overhealing numbers will be SLIGHTLY smaller?

The fact that Rejuv will not crit when the target is at full health means normal ticks on full health targets register as 100% overheal and there is no representation of crit ticks whatsoever on full health targets. This makes it far more difficult to valuate the bonus. This effectively inflates the average effective crit tick in comparison to the average effective normal tick. If crits would still occur and register as overheal on full health targets, I could say with more certainty that the predominant reason the average effective crit tick would ever be higher than the average effective normal tick is because the crit has hit for more than a standard full effective value / no overheal rejuv tick on a certain percentage of intervals. As it is, any delta between average effective crit tick and average effective normal tick could be chalked up to one registering on full health targets and the other one not doing so.

Both registering still wouldn't be perfect as it is possible due to variation that crit ticks healed for more than normal ticks on targets that were very near full health (e.g. normal rejuv is 2500 but due to a target being at say -1750 when a crit RJ hit and a target being at -1000 when a normal tick hit, it would appear the crit had more value when in actuality either one could have healed for the same amount in both situations. Either way, both registering on full health targets would definitely eliminate a good portion of the skew if not all of it. At the end of the day, while this methodology has no impact on our output, it does impact our ability to measure the value of the bonus.

EDIT: Also, as I expected it might from having seen prior parses of Twins with Val'anyr, the remainder of the data was hard to draw any conclusions from as Blessing of the Ancient Kings was making up almost 70% of effective healing and affecting the output quite drastically.

A good night's sleep and some time to think while walking my daughter to school brought some clarity (it seems like clarity anyway). I was using two criteria: 1) Crit Rate and 2) Overall Effective Rejuv healing as a percentage of total effective healing. My thought was in that these were effective numbers and I had already dealt with overheal. But, as I have looked at live data, I can see that we have not put overheal to bed at all. I think I have a workable formula now that should be a good estimate of T9 value and it's based on the 2 input above as well as 3) Percentage Overheal of Rejuv.

Looking at it this way, T9 with a 30% crit rate where Rejuv is making up 40% of your total effective healing and has 0% overheal, 4 pc T9 would net you approximately 6% net effective throughput. Now, if all I do is slide my overheal percentage from 0% to say 60%, 4 pc T9 would net approximately 2.4% VALUABLE total throughput. Slide overheal to 80% as we've seen in some parses and the value drops to 1.2% VALUABLE total throughput.

Again, there is variation and RNG to deal with but the law of average should bring us to a point where those numbers begin to make sense. This definitely captures a more complete picture I believe. So, if this is a valid way of looking at things, the next thing we can do is agree upon what a general average of crit%, % Rejuv of effective total healing, and % overheal looks like given the damage profile of ToC and our gear going into it. My initial thoughts would be 30% crit as a good round number, Rejuv 40% of effective healing, and Rejuv overheal value of 70%. Given those three input, I would place a value of approximately 1.8% effective VALUABLE throughput to T9 4 pc.

So, now some folks might be thinking 1.8% versus 10-12% of T8. But I don't think we can do that. Going through this process has definitely made some of the things Playered has noted earlier about T8 and his pointed questions about how much of T8 was really increasing VALUABLE throughput instead of meter inflation. I really think to accurately account for that we should also take T8 effective throughput and modify that with an overheal value. So, if T8 shows 10% total effective healing on a meter but there was 70% Rejuv overheal on that encounter, perhaps we would look at this as 3% real, VALUABLE healing.

What the above really is trying to capture as it regards appropriate T8 / T9 valuation is the robbing of Peter to pay Paul effect that Playered spoke of. I am capping VALUABLE because it's somewhat of a new term that I'm trying to place a meaning to versus EFFECTIVE. Based on the data, we have likely gotten too caught up in EFFECTIVE and not paid enough attention to VALUABLE. For instance, how much VALUE did a quick instant heal of Rejuv buy us if we used it on a target that was near full health and the remainder of that Rejuv was eaten up in Overheal? It's effective heal may be high, but it's value may not.

Anyway, very preliminary but lets say 1.8% VALUABLE total healing throughput from T9 given the profile I used above of 30% crit, 40% RJ total effective healing, and 70% overheal (baseline ToC numbers I'm swagging at). Let's also say that, using that same ToC profile and now accounting fore overheal devaluation, that 3.0% is a reasonable VALUABLE total healing throughput number from pre-nerf T8 and 1.5% a reasonable VALUABLE total healing throughput number from post-nerf T8. At this point we are bringing the bonuses back down to earth where their impact is not as influential as we have made them out to be in prior conversation. If this is the case, then we can get back to the basics of focusing on the core stat itemization (spell power, spirit, int, haste, etc.) that we understand very well and, to some extent, be less concerned about the impact of the 4 pc bonuses.

I'm typing this out in kind of rush before I head out for meetings and design reviews but thoughts on the general conclusions I'm drawing and the premises I''m basing them on would be interesting. I think I'm on the right track to nailing this down (at least in my own head) but a level-set from the community would be welcome.

Arythorn,

If you want to make a transition from "EFFECTIVE" to "VALUABLE" you also need to consider the insurance value. My auto insurance is effective when I'm in an accident (or have to show it to some official for some reason). However it is valuable whenever I want to drive, in that it reduces my financial risk associated with driving.

If Bob the mage is down 4k in health, and you are certain that he won't take any damage for the next 30s, do you want Rj+4t8, or Rj+4t9? The answer is you want Lb (or perhaps 1/6 of a WG). It provides all the benefit you need, and it costs less.

In practice, you don't see Bob's situation very often. You don't know for sure if he'll be damage-free for the next 30s get hit for 20k in 1s, or take steady damage for a while. You'll probably use Rj before Lb because it provides a significantly better insurance policy against that uncertainty, even if it costs a little bit more.

Will Bob benefit more from your t8, or from your t9? Some incoming damage/heal patterns will kill him in either case. Some will leave him at full health in either case.

I think it is pretty clear that t8 is a better bonus if both
1) Bob has damage right now.
2) Bob is in more danger between now and now+9s than he is likely to be between now+9s and now+18s.
I think (2) slightly favors t8. (1) significantly favors t9, unless you only cast Rj on people who already have damage.

If you have a log of someone with 4t8, it is easy enough to see how effective the bonus actually was.

I think you can make a reasonable prediction of the potential effectiveness of t9 with any log tool that will show you a graph of health-deficit vs. time for individuals.

Suppose my Rj ticks for 2k and would crit for 3k. I can look at Bob's health graph and quickly see that his health deficit on some fight was over 2k about 25% of the time, and over 3k about 15% of the time. I could then make a reasonable estimate that for Bob, 15% of my Rj ticks would have been fully effective, and another 10% would average about 50% effectiveness. If Bob is representative of my other ranged DPS (stands in the fire no more than usual), I can make extrapolations to a larger group.

I'm not sure which online log parsers (if any) will give you graphs of health deficit vs. time. Wowlogparser is a desktop combat-log parser with that capability (it calls it a received healing plot). It can also show when each incoming heal landed, so rather than see how much he was at low health, I could count how often my Rj ticked when he was at low health. Unfortunately, I can't filter that graph to show only health + my rejuv ticks.

I am still unsure if I want to break my 4pcT8 bonus after reading the last few pages of replies.
A lot of the utility of 4pcT4 lies in the fact that the initial heal is instant, even if it is relatively small.
For example, when I see someone's health get low in a raid, I like to rejuv them (if there is not already a rejuv applied, and sometimes even if there is), and then Swiftmend immidiately after. I like to think the instant tick is enough to handle damage they may take during the GCD, and I place a high value on that.
If I see a group of people take big hits at once, I will WG the group, and then begin rejuving each of them. This will do about 3-4k healing on each of them relatively quickly, and provide enough of a cushion for a pally or priest to nuke them back up.
Overall, I worry that I might see more raid deaths without access to that instant heal.

As far as healing meters are concerned, we can be certain that the instant heal of 4pcT8 will not be overhealing, we cannot be certain about any subsequent ticks. (except for constant damage fights like Twins, etc.)
4pcT9 will probably provide a net gain on healing meters during 'constant raid dmg' fights, but with the possibility of more raid deaths (IMO). I suppose it wouldnt provide much gain on "normal" fights with mostly burst dmg.

For the record, I did 6700hps on this week's Heroic 10 ToGC Twins fight. There was a holy pally doing around 5000hps, and a h.priest not far behind him.
45-50% of my healing came from rejuv (i dont remember the exact numbers). I will be sure to get the software for a parse of next week's run.

edit: also, i am glyphed for nourish right now, because i was tank healing during NB last week, however normally i would use the WG glyph

My latest version of Resto Druids stat weights are out for late AC geared Druids. As always, Full buffs, Living Spirit, Improved mark of the Wild, 8 minute fight, 80% Replenishment Uptime, Improved Tree of Life were all assumed. I also estimated the benefit that crit would get from Rejuv (as it is obvious that we are still hammering that value out). For those of you who wish to follow along in Paininabox's Spreadsheet, my spell distrubution looked like

LB (Hot) - 5%
LB (Bloom) - 1%
RG (HoT) - 5%
RG (Dir) - 1%
Rejuv - 60%
Nourish - 7%
HT - 1%
WG - 20%

Near BiS pre hard mode gear was used to create this list. Due to the better mana regen on the list, I've upped my Mp5 -> SP converstion from .8 to a .9. To estimate the value of Rejuvenations crit, I simply assigned it the same value as Rejuv's Spell Power, and figured out the total benefit by hand (All other critable spells aside from LB made this seem like a reasonable asumption).

Stat Values:
Mp5: 1
Int: .554
Spirit: .591
Spell Power: .9
Haste: .503
Crit: .517

Resto Druid Loot Rankings, Pre AC Hard Modes

Things that surprised me, but still make sense.
Haste and Crit are almost as valueable now point for point to Int and Spirit. With the new BiS gearing used, a resto druid could be pushing 3700 SP raid buffed, the more Spell Power you have the more valueable crits and haste will become. Also, with this latest batch of gearing, i'm flirting with the haste cap (where i used to have 100-200 extra haste). Crit has turned into a relevant stat again due to its effect on Rejuv, but as is apparent...the total value of that is still getting hammered out. A resto druid in this gear will have ~700 mp5, plus nearly 26k Int, which is largely the reason i raised the value of Spell Power compared to mp5.

Any thoughts or suggests would be appreciated. Keep up the good work

The way I am attacking estimating how often those that had my Rejuv on them were at or near full health is by using % Overheal of Rejuv. It may not be perfect but it should give us a good ballpark value on a given encounter or set of encounters and it is much easier to glean from a parse than health deficit vs time correlated to whether Rejuv was ticking per raid member. My supposition is that, if it is the case raid-wide that my Rejuv targets were most often not near full health, we will see Overheal drop and vice versa rise if they are near full health a preponderance of the time.

I would say assuming 60% RJ in late T9 with no 4T8 on Coliseum encounters is not realistic especially considering that Coliseum logs tend to peg RJ lower than 60% even now still with 4T8.
Sticking around 40% RJ, 10% LB (HoT), 25% WG, SM/RG/Nourish at a rough total of 20% and a left over 5% for the additional extras seems more logical.

Maybe it is just me, but the link does not seem to work.

That's because there are two "http://". The proper link is this.