I've been doing pretty much the same as you. Entering the fight with 3 HP ready to be used and hopefully at least 1 proc of conviction be nice also

I assume you are averaging the procs as Value = Buff * Duration / (ICD + Time-to-Proc), so that Time-to-Proc = Buff * Duration / Value - ICD. That gives:

Name	Buff	Duration	Value	ICD	Time-to-Proc
[Tear of Blood]	1710	15	400.78	60	4.000
[Fall of Mortality]	1926	15	269.64	75	32.143
[Fall of Mortality]	1926	15	269.64	90	17.143
[Witching Hourglass]	1710	15	324.68	75	4.001

I see a couple of issues:

• I'm not sure what the chance to proc from crits is for Tear of Blood, but I know our crit chance is fairly low. Are you sure that 4 seconds is a reasonable expectation for average time to proc? Is that number pulled from logs?
• It would appear to be an error in your original table that normal and heroic Fall of Mortality are valued the same despite the difference in ICDs.
• Both versions of Fall of Mortality and the Witching Hourglass have procs that read similarly ("Your spells", "Your healing spells") and show as a 10% proc chance in Wowhead, so one would expect them to have the same time-to-proc. It would appear that you are not using the same method to evaluate these three trinkets.

While I believe there's no set answer to this, I do know that progressing into Hard Modes has had me pushing out far more Divine Lights than Holy Lights, particularly on tank healing and in conjunction with ToR. Considering how poor our spell coefficients are right now, until the change to mana tide and Divine Plea, I'm running with Jar for the additional regen.

Unfortunately, I despise archaeology, so I'll never get Tyrande's. Seems like my next best choices are Fall of Mortality and DMC: Tsunami.

Without detailed math, it's not obvious how converting more HL->DL with the additional mana from JAR compares to healing a little more with all spells due to the additional Intellect from Core, just to pick an example. My hunch is the opposite of yours - my Jar sits in the bank while I use Core/DMC:T - but I won't attempt to claim that it's more than simply a hunch.

I think playstyle is also a significant factor in the comparison. Relative to the other paladins in my raid, I spend a lot more time in melee range getting mana from SoI and HP from Crusader Strike. I have a larger proportion of DL and HP heals while they use more HL. We end up with similar healing output overall, but with my playstyle I see mana as being a little more easily available and HP heals (mana-free) contributing a larger proportion of my total healing. Consequently, I value Intellect slightly higher vs. Spirit than they would.

Offtopic: I wonder if it's by design that the acronym for Jar of Ancient Remedies is JAR.

Your argument is:

1. A little Intellect provides a small amount of healing if you spam HL
2. Heroic mode encounters require a lot of healing.
3. Therefore Spirit is somehow unquantifiably better than Intellect.

It's certainly not a compelling comparison of how well our overall healing performance scales with Spellpower vs. Regen or Intellect vs. Spirit.

I think your earlier post had the right answer to this :

All this complex math won't take into account the other healers, the encounter itself, etc.. So far none of the 6 heroic encounters I healed had me going oom (at least after I learned when and where to use what cooldowns) but I find it impossible to get into any sort of a spell "rotation". As long as I don't overreact to incoming damage and start panic healing my mana is fairly stable and I never had to swap in a spirit trinket.

First off, I estimated the time to proc on Witching Hourglass to be 4 seconds, I personally use this trinket so I will go and retrieve some logs and verify that.

The proc on Tear of Blood is (supposedly) 30% on crit, and I originally estimated buffed crit to be 30%, which sounds a bit high. I will revise that to 22%, which gives any given spell a 6.6% chance to proc the trinket. Witching Hourglass has a 10% of any spell to proc, so I can scale Tear of Bloods time-to-proc based on what I find with the Hourglass.

Those numbers with Fall of Mortality look pretty messed up, so I'll redo those calculations, hopefully with some logs.

Edit: Time to proc based on 9 procs from Cho'gall and Val & Ther for Witching was 7.222 seconds, which puts Tear of Blood at 10.942 seconds.
Time to proc for Fall of Mortality based on 8 procs from Maloriak, Magmaw, and Omnotron was 7.159 seconds, with the ICD of normal and heroic both being 75 seconds. The proc value, duration, ICD, and chance for heroic vs. normal are identical.

Updated table is below:

LoD/HR/PotI is a moot point due to Witching Hourglass having the same feature (and all other trinkets). Although 2pc T11, Infusion of Light, and Divine Favor could potentially decrease time-to-proc, so that is something to keep in mind.

Doesn't quite even out though as BoL cannot crit (assuming it can even proc them?!).

Working with the Tear proc & 22% crit you end up with approx 11.4% chance to proc when healing a target other then yourself.

((1-(0.78^2))/2)*0.3 = x
(1-((1-x)^2))*100 = 11.40%

(If anyone can simplify/show any errors in that please do, my brain is a mess and tried to model each crit/chance to proc separately per cast)

With the 10% on the others you reach approx 27.1% chance. (Assuming a non-beacon-target, non-self heal).

(1-(0.9^3))x100

(tried to think of any reason that wouldn't be accurate, unless I missed something it seems fair)


Also, the no. of targets seems to make much less of a difference than I thought with reference to the tear, if you work out the chance to get a proc on each crit inside each cast.

Working with a 6 target LoD and a PotI proc I got a result of 22.2% chance to get a proc... whereas, with beacon transfer raising that to 14 heals, the others would reach 77%.


EDIT: After a few hours sleep and a re-read I just wanted to try and sort out what I was trying to do with the tear part.

Started out just by working out the chance to get a crit from both the heal and PotI and *0.3. Then I realised that getting a proc FROM the crit was similar to getting a crit itself (can't just use the sum of the whole, 50% chance to crit over 2 casts has a 75% chance to get a crit, not a 100% chance).

X is just the chance to get a proc - off the avg chance to get a crit from a 2 cast situation /2 to seperate them to allow the second part to be worked out, which is the sum of the whole which is (hopefully :p ) the chance to proc from the chance to crit in the 2-cast situation.

I apologise for both the complete mess of my maths and the perhaps over-simplified explanation, was never good at showing my method and I know some people who come here don't instantly grasp some of the concepts behind the maths.

You guys needed Tear of Blood times to Proc, right?
I have some logs if that will help your math at all. I've been using Tear in raids for like 4 weeks now (in fact, so has the other Holy Paladin too, if you care to dig through her data).

World of Logs - Real Time Raid Analysis
The buff is called Cleansing Tears.

My gear is pretty bad, mostly blues, yes, but honestly? Who else would be using Tear of Blood anyway? I'd say it is a fair comparison given that people like me are going to be the people using it.

Halfus 10N Kill
Edit: Proc prior to fight
First proc: [19:42:37.052] (TtP: N/A)
Second proc: [19:45:26.537] (TtP: 109.485 sec)

Valiona and Theralion 10N Kill
Edit: No procs within minutes of fight start
First cast: [20:12:27.677]
First proc: [20:12:28.365] (TtP: 0.688 sec)
Second proc: [20:13:54.162] (TtP: 25.797 sec)
Third proc: [20:15:22.865] (TtP: 28.703 sec)

Not counting Argaloth because it is a joke fight that I probably wasn't casting half the fight on and Magmaw because it has a huge regen phase where I didn't cast either.

Omnitron Defense System 10N Kill
Edit: No procs within minutes of fight start
First cast: [21:11:08.615]
First proc: [21:11:22.443] (TtP: 13.828 sec)
Second proc: [21:13:12.349] (TtP: 49.906 sec)
Third proc: [21:14:56.709] (TtP: 44.360 sec)
Fourth proc: [21:16:14.537] (TtP: 17.828 sec)

Maloriak 10N Kill
Edit: No procs within a minute of fight start
First cast: [21:31:54.256]
First proc: [21:31:54.724] (TtP: 0.468 sec)
Second proc: [21:33:50.943] (TtP: 56.219 sec)
Third proc: [21:35:06.974] (TtP: 16.031 sec)
Fourth proc: [21:37:46.287] (TtP: 99.313 sec)

Chimaeron 10N Kill
Edit: No procs within a minute of fight start
First cast: [22:36:39.521]
First proc: [22:37:40.193] (TtP: 60.672 sec)
Second proc: [22:39:10.912] (TtP: 30.719 sec)
Third proc: [22:40:28.365] (TtP: 17.453 sec)


Average time to proc over listed encounters: ~38.098 seconds

Oh, derp. Let me fix it.

Edit: Done. Only Halfus had a proc before the fight.

Asside from using what I consider to be 'incorrect' stat weights, your table is missleading at best. You have chosen an arbitrary Int value, to determine the effective value of spirit for mana based trinkets, and then applied a spirit weight to that, when all these values varry with Int.

The Doll for example is 321 Int plus 4200 mana per 1 min. While the Core is 321 Int + 1926 Spirit for 20s/120.

So lets consider a 2 min segment, so 8400 mana from the Doll. Core gives 20s of 1926 extra spirit. The core will become a better trinket at 17,000 int. Obviously this is not a realistic situation, but it still shows how the trinkets will vary in stats when some are based on non-scaling returns.