Someone wanna double check the mathFollow

Cause if so we might just have a solid conversion of MP5 to Spellcrit for both HL and FoL


http://bloodpaladin.blogspot.com/2007/11/holy-paladins-healing-vs-spell-crit-vs.html

from the same page

The thing about Paladin healing is you never stop healing, unless you are bad, poorly geared, or don't use mana pots proactively.

I've written a VB (yes its the only programming language I know) program to generate a random sequence of heals and work out how long it takes to go oom, unfortunately changing the crit chance doesn't seem to affect the time to oom very much, which is probably a programming error, if I get it working it should give a reasonably accurate relation between mp5 and crit.

Well the flaw in your argument is that you assume continuous fol spamming, which won't get you much further than kara, to get a true comparison you'd need to know the proportion of time spent not casting and the ratio of holy lights to flash of lights.

My program got horrendously complex (I didn't use any subroutines at first, and it got out of hand) so I started afresh with some more elegant code, though i'm not sure a simple answer is possible, it depends on your current stats, the amount of time you want to heal for and the rate of healing you need to maintain.

Well, it is easy to adjust my math to particular situations if you like. Even separately for every boss fight out there it is possible (the ratio differs according to the type of a fight!). Do the following for preferably a few fights against the same boss:

- track the total time the fight lasts, say x seconds
- track the total numbers of FoL cast, say a FoLs cast
- track the total numbers of HL cast, say b HLs cast, base mana cost HL =
- average all three numbers if multiple fights are available, transforming x in an average X, a in an average A and b in an average B

General information:
- casting time FoL = 1,5 seconds, base mana cost 180
- casting time HL = 2,5 seconds, base mana cost 840

- over the cause of the entire battle the total mana M returned through the talent in the holy tree for 1% spell crit (not rating!) then equals:

M = a*1,0%*60,0%*180 + b*1,0%*60,0%*840

- the mana returned per 5 seconds M5 through the holy talent then becomes

M5 = 5*M/x

This is mana-return wise for that particular battle directly comparable to mp5.

As an example, 1 fight recorded:
X=600 seconds (10 minutes)
a=200, meaning 300 seconds of FoL casting
b=100, meaning 250 seconds of HL casting
idle time nonhealing is 600-300-250=50 seconds

Total mana returned for 1,0% spell crit (not rating!)

M = 200*1,0%*60,0%*180 + 100*1,0%*60,0%*840 = 720

Mana returned per 5 seconds:

M5 = 5*720/600 = 6.

For this example, ONLY OBSERVING MANA RETURNS, 1,0% spell crit (not rating!) equals 6 mp5. BUT: 1,0% spell crit also means 1,0% more healing done over the full course of the battle... making the 1,0% fully superior to 6 mp5. To what ratio? Well I would say 1,0% more healing done can be converted to a +healing number (quite easy math) and thus the 1,0% spell crit (not rating!) equals 6 mp5 and an amount of +healing.

Pfffffffff I call it a day.


Edit: some minor text adjustments.

Edited, May 29th 2008 5:27am by Immunios

You are correct, Leodis. To be fair: I am not a holy paladin and to be even fairer: I have not played my prot paladin for quite some time now so to be fairest: I totally forgot about Light's Grace. /facepalm

However, all the above calculations, with a correction factor for the Light's Grace talent, should be ok to have at least a good estimation of when a spell crit item should be valued over an mp5 item or not.

When the comparing values are far off from one another, it is an easy choice. When they are close to each other, it will be a plus in some fights and a minus in others, depending mostly on the random number generator on both the damage the tank takes and the healing done by the paladin. In the latter case, there is no real "best" item of the two, so then choose what seems appropriate for you/ for the fight.