Magic Accuracy - Why so serious?

[Nevex]

A similar argument could be done with m.acc-CHR-Charm-Light/Apollo staff.
Has anyone ever tested if m.acc does indeed affect Charm (since staff does)?

I also remember reading somewhere that magic accuracy affects skillchains resist ratio, and being a hidden stat would explain how low level targets are more likely to recieve full damage from skillchains.

[Wizerd]

My preliminary tests a while ago on RDM suggested that 1 Magic Accuracy = 1% hit rate increase (for landing a full hit, lower tiers were generally split between 1/2 and 1/4, with a small portion being the full 1/8 resist).

It also suggested that monsters have magic evasion numbers comparable to melee evasion, assuming the melee accuracy formula is used. NQ staff gave roughly a +10% magic accuracy increase (not to be mistaken with +10% hit rate increase), and HQ gave roughly +15% (it was closer to 11% and 16% but the sample size was restricted to 200-300 each). I never tested INT or MND's affect on Magic Accuracy, nor skill's affects. Just "Magic Accuracy +#" and staves against a baseline of 500 casts.

Eventually I do plan on getting around to testing more.

A warning to anyone who plans on testing enfeebles: you have to wait until the enfeeble wears off and record the time on each monster. Any enfeeble that builds resistance won't qualify unless it's on an hmpede (which will lose aggro and reset any resistance). Simply landing the spell vs not landing it means you're comparing 1/1, 1/2, 1/4 resists to 1/8 resists, and that is useless. Also, I do believe enfeebles have a 1/16 resist factor (1/16 being the full resist, 1/8 being the enfeeble that seemingly lasted 3 seconds). Nukes, however, do not, despite what anyone will ever tell you.

Also, my tests all ran between 52% hit rate and 90% hit rate, meaning if the inflection point (that's what they call it right?) was 50%, then it might require 2 magic accuracy to reach 1% hit rate below that point, which was a point someone brought up when I first posted my results.

[pchan]

A warning to anyone who plans on testing enfeebles: you have to wait until the enfeeble wears off and record the time on each monster. Any enfeeble that builds resistance won't qualify unless it's on an hmpede (which will lose aggro and reset any resistance). Simply landing the spell vs not landing it means you're comparing 1/1, 1/2, 1/4 resists to 1/8 resists, and that is useless. Also, I do believe enfeebles have a 1/16 resist factor (1/16 being the full resist, 1/8 being the enfeeble that seemingly lasted 3 seconds). Nukes, however, do not, despite what anyone will ever tell you.

I agree on partial resists for nukes, as I've never seen 1/16 out of the many tests I did. Also, length duration for slow and paralyze seem purely random as ES slow II or paralyze gave me 1 min and 30 sec a few times while the typical duration of slowII is 3 minutes (I think it's 4 minutes for blindII), w/e accuracy you have. For sleep I and II, I have never seen anything else than 1/2 duration. For gravity and silence it seems totally random too as you can verify with Elemental Seal.

[Ambir]

That flan test is the first I've seen that actually gives an idea of the relationship between EnfSkill, base stat(here INT), MagAcc and the actual chance of landing a spell (over a decent number of tests). Anyone know off hand roughly what level bat a 75 RDM can sleep about 50% of the time? Be interesting to see how the stats on different mobs at different levels...

I do have a significant amount of MagAcc gear. Would be interested in doing some more tests to confirm...