And yes, when things go wrong, mp5 is likely to be more useful than int (since things go wrong towards the end of a fight usually). Hence my proposition that if a fight lasts on average 6 minutes, then assume it will last 8 minutes and find the corresponding int to mp5 value in her graph. Then, so long as things don’t really REALLY go wrong, your int vs. mp5 ratio will if anything overrate mp5 (in the event that the fight doesn’t last 8 minutes). How much you offset this length estimate should depend on the variability in the length of the encounter (eg. if you avereage 9min 30s and enrage is 10m, you don’t go off by much. If it’s say, Lurker though, where you can get by for a long time without a full raid, maybe you want to slide up the graph by even 3min).

The reality of the situation is, you will rarely tweak your mp5 to int ratio on a fight to fight basis. More often, you have maybe two or at most 3 potential choices to use as part of your primary healing set. As noted before, 2.5 int costs the same as 1mp5 (again, with the exception of Royal Nightseye gems). Using Phae’s graph, we can see then that for any fight over 3min 30sec (estimate, and remember 2.5int = 2.75 with kings), mp5 makes better utilization of your itemization points than int.

Most, if not all challenging (i.e. if you’re T5 and in kara this doesn’t apply) boss fights in this game last longer than 3min 30sec. Because of this, the most important piece you can get from her analysis is: always gear for mp5 over int (in cases of equal itemization level). The only exception to this rule should be when your int is so low that you are losing out on the effects of innervate, etc. IMO, the absolute ideal balance would be to have enough int that one innervate will exactly fill your mana bar. Everything else I would want in mp5 and spirit (from a strict regen perspective, obviously you need healing & stamina too).

And yes I enjoy this as much as you do Kalaghan

Now to our subject again.

I get your point about the need to burst heal. Let’s say that the accident happens on the 1st min. With intellect you have 45mp more to spare, while with mp5 you have 12mp. Will the less mp would be enough? Maybe yes/no? Depends on the dps and your total mps. Probably yes cause your mana pool is rather nice anyway and the fight is still young. Now consider having to burst heal again after two minutes when the boss enrages and your mana pool is running low. You have already spent your 45 mp from intellect and now have nothing more to spare. On the other hand, the 1 mp5 would have given you 12 mana more by that time which might be adequate to allow you to throw an extra rejuvenation. And of-course the mp5 value in this matter increases as we move to more lengthier fights that might include killing frenzy phases or more cave-ins or shatters.

I agree on your max desirable mp5 point. But I have to say that such levels are out of reach for most of us, at least until we get some T6. I also agree that after a certain amount mp5 loses desirability because you can maintain enough mana for the whole duration of the fight. BUT I must point out that this effect comes through mp5 and not through intellect. Get the Dreamstate build for an example. You get a huge mana pool, but you also get a huge mp5 regen to compensate for your need to spam heal.

To sum-up an intellect increase
Has no time-related positive effect.
It’s utility is negatively affected by boss DPS and can be actually 0 after the fight’s duration exceeds a certain point (dependent on boss dps again provided a given heal output), since you will have spend all of it by then.

In your example, the utility of intellect past the 1st minute is actually 0.
So introducing boss dps as a mean value or something would make the model much more accurate.

http://www.wowhead.com/?talent=0tcMubIZZxxcbeq0xoh
http://www.wowhead.com/?talent=0tcMubZZxxcLeqMxox

They’re still not my favourite builds, and I would question how far your current style will heal you into deep content. It sounds to me like you are making a concious decision to sacrifice a bit of healing power and effeciency so that you can play the style you enjoy most. There’s nothing wrong with that. As long as you (and arguably your guild) are aware of the trade-offs you’re making, then the more power to ya. =)

@Dimitris

I would have to strongly disagree with you. I think Phae’s analysis is fine. What about the reverse situation. Rather than consider a 3 minute fight, consider a 3 minute fight where you have to heal for X total health in the first 3 minutes. Suppose it takes 36 mana to heal for X.
1mp5 = 36MP
3int = 45MP
But oh no, someone didn’t get out of that cave-in/AoE/cleave in time and suddenly now you have to heal for more than what your 36mana allows. Suddenly your int up front is more valuable. There’s room for variability in either direction, thus improving the value of either stat.

Additionally, consider enrage timers. Illhoof, Netherspite, Hydross, FLK (Fathom Lord Karathress), Leo, and VR (Void Reaver) all have enrage timers. There is no way for the fight to go beyond the 10 minute mark (at least not by more than a few seconds), so you can put an absolute cap value on your mp5.

To say that it’s comparing apples to oranges just doesn’t seem reasonable. There is a clear baseline value that is the only thing that matters from either, the mana they provide (well, for druids at least…int does improve spell crit which is more relevant for other classes).

Granted, if a fight on average lasts 6 minutes, you might not want to use her graph to determine the value of int to mp5. This is because the distribution of fight lengths (given that it’s a variable) probably does not follow a normal distribution. eg. It’s more likely the fight will go longer than that it will be shorter. In order to make the analysis any more accurate you would have to delve into high level statistics and model the probability of various fight lengths, as well as a work on a model that would show the distribution of volumes of healing needed early in the fight (i.e. front-loaded mana use, providing additional value to int). To expect such statistical work would be utterly absurd though.

Your criticism doesn’t invalidate her analysis, it only serves to point out a fact she has already addressed (within all reasonable expectations). The relative values of the two stats are dependent on fight length. That’s why the X factor in her graph is the length of the fight. If you’re concerned about the fight going longer than it’s average of 6 minutes, then shift down the graph to the X = 8min marker.

P.S. The $10,000 per year for an unlimited number of years is called an annuity and the forumla for determining it’s value in present dollar terms is:
NPV (Net Present Value) = $10,000/r
Where r is the current rate of interest (in this case it would be annualized since the $10,000 is per year, but effectively you only need to match the time units used).

@Phaelia
Sorry if I’m taking over your comment board! Hehe…I don’t have the the motivation to get my own blog going but I love discussing the finer points of being a resto druid.

OK to use your graph.
Let’s assume a 3min fight (36 ticks).
For 1mp5 you get 36 MP.
for ~3 intellect you get 45 MP.
The 2nd actually sound better does it not?
But what happens if the dpsers actually slack cause they are eating pizza and the fight lengthens to 6mins? (Provided that you already have made your choice?)
You will still get 45 MP from intellect, but mp5 would now tick 72 times giving you double the mana.
This ex-post gives mp5 more desirability cause you cannot predict for sure how much the fight will last or how much mana you have to use.
My point is that there is no error in your logic, but your not taking into account how each stat individually gets affected by the game or affects the game itself, thus making it rather unsuitable for decision making, unlike your other calculations that compare stats with effects that are more close to each other eg mp5 and spirit. Comparing mp5 and intellect only to each other is a bit like comparing apples to oranges.

@Dimitris: MP5 does not last forever; it stops when the fight stops. Getting mana back after we’re /dancing over Gruul’s dead body doesn’t help me. Which is why the value of Intellect vs. MP5 is directly related to the length of the fight. To correct your example, $10k per year “forever” would really be $10k until you die. You can take your average life expectancy given your current age and determine how many $10k payments you can expect over the course of your lifetime. So if you were 25 years old and your life expectancy is 90, then you should get $650,000 over the course of your life. Would you rather have that paid out right now or divided up into $10k increments over the next 65 years?

@Ermengol: Don’t worry about it! I had to doublecheck myself. I wasn’t aware that the tooltip had changed. =)

So the correct example would be something like the following (in terms that a woman would appreciate : p )
Let’s say that you like DK shoes so much that you want to buy as much of the models they produce. Your usually spend 10k a year to them and you are offered either 1 million $ now or 10k $ per year “FOR EVER” (like mp5 is forever, or at least for as long as the fight is). Of course there is a big chance (DK being a succesful company and all) that will keep making more shoes for much longer than 100 years… Which one sounds better then? (Provided of course that you manage to evercome that minor age problem :p )

There are actually some financial models that can be used to calculate that with indefinite payments and all abut I am sooooo bored to remember/search for them…

At some point of the history of WoW MP5 stats were introduced but the tooltip remained the same for a while, hence my mistake

/cry