Unfortunately, there is no right answer to this, no magic crit number. I'd say as high as you can comfortably go without sacrificing other stats, if you gave me some current numbers for you (healing/crit etc) I could probably be a little more helpful. I highly recommend playing around with the two spreadsheets in the main post and seeing how your HPS and efficiency compare with a few gem/gear options, and see what numbers you like most.

I've always thought that Consecrate healing friendly targets would be a good way to make that work; an AOE HoT, but with it's own twist, a bit like Holy Nova. It would probably have to be a Holy Talent, though or Prot Paladins would soon become very popular tanks with melee heavy raids!

*gasp* A useful role that has synergy for a Prot Paladin? No way.

For 2.3 I would expect people speccing 42/11/8 - Talent Calculator - World of Warcraft for keeping the 3% crit up on the boss?

Honestly, I think rank 8 will still probably be useful. I use it because I want to heal my non-BoL'd raid for ~3k, where FoL doesn't do sufficient HPS, and higher is mostly unnecessary. Overall I end up with pretty close to 50/50 split of HL(7,8,11 depending task), and FoL.

For gemming I've been using the gleaming(spell crit) cut for yellow, and royal(Healing/mp5) for blue (as well as red if we actually got red sockets). My HL crit rating is ~42.5% buffed. I'm looking into if luminous(healing/int) or brilliant(int) cuts would be a better use of these sockets.

From my WWSs, I'm finding on healing intensive fights, I end up casting and completing a heal every ~5.2 seconds. These fights being naj'entus raid healing, mother tank healing, and bloodboil tank/felrage healing. I'm getting that the mana saving/regen and healing (my HLs tend to have lower over-heal than crits, so they're contributing some to healing) really don't seem to be worth it. I'm getting 10 crit rating being equal to having a 1.5 mp5/+1 healing gem for rank 8 HL (improves up to 2.3 mp5 with rank 11).

In doing my calculations I've made simplification assumptions which should favor crit. The fights I took samples from tend to be roughly 6 minutes. I divided by the number of heals I landed in the fights (FoL/HL), to get the number of seconds/heal. I then divide the seconds/fight by seconds/heal to get heals/fight. Treating the FoL heals as HLs inflates the number of heals done in a fight. I'm getting 69.3 heals in a fight.

4 mp5 over 6 minutes gets you 288 mana. I then found out how many crits you need with FoL, and HL ranks 7, 8, 11 (assuming all casts were with only one of those spells). And then converted the percent crits into rating, and then rating into rating/mp5, and finally 10 crit rating into mp5.

For healing I took my HL overheal and divided by the crit rate, finding about 17% of the extra healing done by critting was effective healing. My average HL heal size worked out to 3800 (over the course of a whole raid). I crunched this down to 10 crit rating contributing approximately the contribution of +1 healing.

Yes my calculations include a fair number of assumptions and roughly taken samples, but is there anything in there that should make these numbers significantly lower?

You assumed all your overheal was due to crits. If you took only the crit overheal and divided that by the number of crits you would probably get significantly higher added healing due to critting. However I doubt your 17% would become anywhere near 25-30% and even if it was doubled to 34% of your crit healing being effective healing it only buffs 10 crit rating from +1 healing to +2 healing, so the end result is still quite the same - crit being not so good. However I would make sure that the crits really are a significant part of your overhealing and not the minority of it.

I suppose imp might purely depends on your raid comp, if someone else has it for every raid there isn't much point and might as well save some mana, but I guess for general use might as well grab might instead.

I personally spec into Imp BoM for my pve spec, but my guild raids melee/hunter heavy, so the extra 44 ap for 8+ people is always nice. Also the other paladin is always PvP specced, so I volunteer to take the bitch spec. I'll probably go 42/11/8 when 2.3 hits.

I never said you should ever leave your raid without imp BoM... Your raid should always have it! It's just that you don't nescessarily have to be the person who buffs it since you're keeping judgement up. It's really a matter of wether you have someone with it in your raid on a consistent basis or not.

edit:
i removed what had before since i was misunderstanding you.

at first, i was reading your post as if you viewed crit/illumination as a form of 'mana regen', as if mindlessly spamming HL results in mana regen. I didn't understand how crit/illumination was best when in a spam environment.

after thinking about it more, i see how crit/illumination are 'at their best' when in a spamming environment relative to mp5 or +healing. When one starts to cancel heals, mp5 becomes better for regen. When one starts to use FoL, +healing is all that matters, since FoL is already efficient (no massive need for regen). I guess, i'm just saying 2 things. 1. crit/illumination is still very good as a 'scaling reduction in mana cost', regardless of spam casting. 2. spam casting might be overrated in that it tends to result in more overhealing.

i guess i've never thought the mindless continuous spamming of HL was that great since those that did it tend to have significantly higher overhealing percentages. They went out of their way to cast, thinking that they 'regen mana' when casting, but in reality they still loose mana (even with a 100% crit rate), and end up waisting more mana than anyone else. I'm just not convinced that HL spamming was a effective as people thought considering the significantly higher overhealing.

Remember though, you don't get extra loot for finishing the fight with more mana or for using less mana pots. Overheal is not a bad thing unless you have people dieing elsewhere. But more overheal in a very general sense means the targets were topped off (hence overheal) and that is job 1 for every healer. Overheal means, for tank healers especially, they're casting proactively and not reactively-they aren't overcancelling. These are all *good* things. Now if your raid healers have 40% OH and your rogues are still dieing, that's different-but I think the cases where OH can really be viewed as a negative are far fewer than when it's a positive.

That's not to say the highest overheal is the best or the prime metric for healers, but this sentiment of OH being the debil is something that I think needs to be shaken out of people

Good points... I should probably add these into the main post...

I'll do it in a bit.

Damnit Thosin!

On a general note.. in patch 2.3 should guilds be ensuring that 1 pally go imp BoM, one pally go 42/11/8?

good point. OH doesn't matter if not going OOM.

i guess when i said "I'm just not convinced that HL spamming was a effective as people thought considering the significantly higher overhealing." i should change, "as effective" to "as efficient". the _effectiveness_ of the healers I've seen do the constant spamming of low rank HL has been good and reliable, even though the _efficiency_ doesn't look so hot when factoring overhealing. Spam casting low rank is probably a more reliable than strat of big heals + canceling a lot. A couple times i've canceled when i should have let the heal go through >_<

OH doesn't matter if you're not going oom and if nobody dies. Doesn't matter how much mana you had, if you overhealed when someone died it means you *maybe* could've healed that person instead of overhealing another.

Back to the Libram issue:

(It seems that swapping Librams invokes the GCD, so my dreams of using a macro to combine the new Libram of Mending with my every-10-seconds cast of a low rank HL will never fly.)

Lets assume that I'm stuck with using the same Libram all the time, and that my usual job is small heal spam on the tank, while keeping Light's Grace up and being ready to dump a big heal. At the moment I can use my Libram of Absolute Truth, a low rank of HL, and everything is fine and dandy.

After 2.3, that (from what I've read - does anyone have hard numbers?) wont be practical. So I can go back to my Libram of Souls Redeemed and FoL spam, and just use low-rank HL to keep Light's Grace up. Or I could do that but use the new Libram of Mending, giving me 22 MP5. I suppose this really boils down to: 105 +healing, or 22 MP5? I suspect 105 +healing is the way to go

Also beware - if you click that macro right before your holy light becomes "available" (as in right before you finish casting another spell or right before your global cooldown comes up) it'll still swap librams and start a new global cooldown. So use with care (just like /stopcasting)

I asked on SA, but no one there seemed to know yet - has anyone found out where Girdle of Stromgarde's Hope is going to drop when it comes out in 2.3? Neither World of Raids nor MMOChampion have mentioned where it drops, even when listing ZA drops by boss/timed event rewards.

I'm rather anxious to get my hands on it, it being a major upgrade for me and all.

I've been trying a new gear/gem setup and I really like it. The idea is to have as much crit as possible without sacrificing healing gear for spell damage or making stupid choices, like 10 crit on gloves instead of 35 healing. I replaced all of my healing gems for 10 crit gems and only kept 2 healing mp5 gems to keep my metagem working. I also replaced one of my healing trinkets with Xi'ri's gift. Finally I got the spell crit and dmg scryer shoulder enchant since the dmg adds to healing. Altogether I traded 234 healing for 120 crit rating. I did some calculations for my play style and I lose about 87 healing per heal and gain 27 mana per heal, but also gain slightly more burst healing (1 more crit in 20). My goal is to have over 50% raid buffed holy light crit once the t6 chest decides to drop.

It's not the highest HPS setup and not meant for every type of healing, but I like being able to crit more reliably since most of my healing comes in short bursts followed by lots of canceling. I tried it out in a couple raids and was satisfied with the noticeable increase in crits (had about 48% over a 20 minute fight as opposed to my normal 40%). It's slightly more efficient as well, about a 1.7% gain

Just wanted to let anyone that is interested in a crit setup know that it works pretty well if you use a lot of big heals, have a shadow priest, and do mostly single target prehealing.

According to your experiences, you need more HP/S, which is better provided by +healing (or haste) than by spell crit, as the theorycraft says. 10 crit is less than 0.5% chance to heal for 1.5X or 0.25% more effective healing. 22 healing just adds that much more. Just read up if you want exact numbers. On top of that, the regen, which is actually the primary benefit from crit, is also not that high that even the benefit from regen+extra healing doesn't make crit gems worth it.

If you want to max your HP/s you should socket everything with pure +healing gems, not pure +crit gems. If you're more limited by mana than HP/s socket 9heal/2mp5 gems and in yellow sockets 4int/2mp5 (afaik those are the best for that purpose). Anyway the point is that stacking crit is good for neither.

Another thing to note is that +healing will let you land a bigger heal every time, while with crit sometimes you crit when you don't need to, and sometimes you need to crit and you don't, in which case the tank dies. So +healing (and/or haste) is also the safer route.

One thing to add to your gear valhallen for that set up is either the gladiator gauntlets or the merc gloves. Both of those gloves have a 2% increased crit to flash heal as well as passive crit.

Theory crafting is fine and works well for predicting healing over long periods of time, but when you are only looking at short periods of time, such as one or two heals after the tank takes damage, I think crit does more healing than more +healing, once you reach a certain level of healing of course. Small amounts of crit are also worthless, and I would never trade a 22 healing gem for a 10 crit gem by itself. But in large amounts, I think crit is more useful for prehealing with high ranks of holy light than extra +healing, as long as you trade them in a total 2 to 1 ratio. I'd rather heal the tank for 5000 half the time and 7500 the other half, than 5400 2/3 and 8100 1/3, but that's just my preference, because in the event that I need a crit at that moment for whatever reason, I'd like the odds to be in my favor.