My original testing after 04.09.2009 update:Test#1, Test#2
http://i685.photobucket.com/albums/v...kintest001.gif
SHORT VERSION:
• Curepot/MND/skill do increase Stoneskin amount
• Stoneskin is 25% of cure amount
• Calculated cure amount determines skin amount
• Amount restored to target does not affect skin amount (cure for zero works)
• Divine Seal doubles the resulting skin amount
• Day/Weather also increase the skin amount appropriately
• Maximum skin amount appears to be 300
• Strong indication that this was changed! (ninja fix? grrr )
EDIT: added maximum, and Divine Seal
Recently, I dug up my tests to answer a question, but I realized I never got around to retesting.
I felt compelled to point out a mistake I made in testing -- the idea was to match HP motion, using Sublimation (-2) with Orochi Nodowa (+1) to cancel out Auto-Regen. However, subli chews outta stoneskin regardless of regen (order of operations) ... so really it would have been WAY more accurate to just skip Subli & Nodowa.
But then, I was so bothered by my mistake that I went back and re-tested it:
----
Rya's Afflatus Solace
Cureskin Re-Test
09.29.2009
Rules:
•1000hp -250hp (from devo/martyr) = 750 without skin
•watch auto-refresh to time JA mid-tick (ensures auto-regen doesn't skew things)
•test each cure 3 times
•reset JA timers @ Nomad Moogle for rapid testing
Results:
Spoiler: show
----
So basically, cure potency does affect skin amount, and I don't just mean "+potency" gear, I mean the actual calculated cure amount!
I made this other distinction by fine-tuning MND a bit while maintaining 27% and 1000hp. I used Furen's calc to find my cure value, then took a measurement of my actual cureskin:
@ 81mnd/0%pot/276skill
Cure V should be 644hp
results: 911hp = 161skin (161/644 = 25.00%)
@ 83mnd/27%pot/276skill
Cure V should be 821hp
results: 955hp = 205skin (205/821 = 24.96%)
@ 106mnd/27%pot/276skill
Cure V should be 866hp
results: 966hp = 216skin (216/866 = 24.94%)
Conclusions:
• the cure amount (even if it lands for zero) does affect cureskin amount
• cureskin appears to be 25% of the cure amount
What's interesting, though, is that in spite of my inferior method of testing when it first came out, I definitely did not see this huge of a fluctuation in skin amount. Yes, they would vary by a few HP depending on how long Subli had to nibble on the skin, but certainly not to the extent that I saw this time (lowest tested 5skin was 161, highest was 216):
Prev test: c5 @ 640 = 156 skin
This test: c5 @ 644 = 161 skin
Prev test: c5 @ 812 = 156 skin
This test: c5 @ 821 = 205 skin
Yeah, I'm 99% confident that they changed this since I tested it last...
In any case, it's good to know~