Int/2.

New 10/2/08: This thread is now a placeholder for my scavenged data.
Note: About half of this was taken from a Japanese site Nekonecat cited a looong time ago on SW: http://ms-jinrou.seesaa.net/article/19757346.html. If you're interested about this stuff, I recommend running it through a translator (...learn to put up with the crappy syntax, it gets the essence across) and perusing it if you haven't already.

INT	Magic	Spell Base	Mastery	MDEF	Min	Max
20	20	20	0.15	0	4	15
20	43	20	0.15	0	7	31
20	43	40	0.6	0	35	62
20	65	20	0.15	0	10	47
20	65	40	0.6	0	56	95
51	65	20	0.15	0	13	50
51	65	40	0.6	0	62	101
65	65	20	0.15	0	15	52
65	65	40	0.6	0	65	104
349	358	120	0.6	0	1492	2144
334	358	120	0.6	0	1483	2135
284	374	60	0.15	110	???	1031
284	374	90	0.15	180	???	1539
115	150	40	0.6	0	160	251
???	430	100	0.6	85	???	2149
???	435	100	0.6	85	???	2215
???	430	60	0.6	85	???	1270
???	435	60	0.6	85	???	1321
???	633	336.15	0.6	580	8817	12111
553	693	224.1	0.6	0	7014	9361
618	754	201.15	0.6	0	7176	9465
902	1747	567	0.6	900ish	73197	92481

On it Captain!
I'll send some data your way as soon as I start training....
With Heal?
I'll make it SR!

If you're comparing spell damage to heal damage, keep in mind that they're totally different, and their formulas for damage are different. We've already known for a long time that int independently had a small effect on non-heal magic damage. Also, the number of targets hit is pointless because non-heal mob spells deal the same damage to all monsters hit by the spell, whether there are one or fifteen monsters.

This sounds pretty easy to keep track of. I'll try to remember next time I'm farming gold teeth; I usually do that for about an hour. It should be simple because I'm usually only hitting one at a time using Cold Beam.

Which current formula do you consider to be the most accurate? I want to know if I should change the one I currently use on my calculator.

I do not know, I honestly don't.

Just at a glance, the fraction version is the most sensible-looking (as I've said) but it doesn't have INT/2 in it, which is why I made this thread in the first place.

If you're comparing spell damage to heal damage, keep in mind that they're totally different, and their formulas for damage are different. We've already known for a long time that int independently had a small effect on non-heal magic damage. Also, the number of targets hit is pointless because non-heal mob spells deal the same damage to all monsters hit by the spell, whether there are one or fifteen monsters.

I realize. I know that it's been done before, and the INT/2 is known (it's part of Neko's version of the formula) but the F part of his is a bit shaky. And the other two I've seen don't include INT/2.
I didn't mean number of targets hit, I meant the specific type of monster. Mainly for MDEF and element attribute, but you get the level too just in case.

I've found the formula that Nekonecat linked from that Korean site to be very effective in calculating magic damage thus far.

FormulaMax=(~mag * 3.3 + (~mag ^ 2) * 0.003365 + ~int * 0.5) / 100 * ~smatk
FormulaMin=(~mag * 3.3 * ~mastery * 0.9 + (~mag ^ 2) * 0.003365 + ~int * 0.5) / 100 * ~smatk

It seems almost perfect to magicians even in the 150-ish range, so with stats that high, I'd guess it's pretty close to the real deal. I'd also like to point out that it has your INT / 2 (INT * 0.5 is the same) in there.

K if you want some weird data, here's some.

902 int
1747 magic attack (with gigajuice)

Level 10 meteor

I've found the formula that Nekonecat linked from that Korean site to be very effective in calculating magic damage thus far.

FormulaMax=(~mag * 3.3 + (~mag ^ 2) * 0.003365 + ~int * 0.5) / 100 * ~smatk
FormulaMin=(~mag * 3.3 * ~mastery * 0.9 + (~mag ^ 2) * 0.003365 + ~int * 0.5) / 100 * ~smatk

It seems almost perfect to magicians even in the 150-ish range, so with stats that high, I'd guess it's pretty close to the real deal. I'd also like to point out that it has your INT / 2 (INT * 0.5 is the same) in there.

I'll keep that in mind.

There are always mixed opinions about which formula is the most accurate. My beef with this one is the 0.003365, which seems like it was pulled out of a hat (or maybe Excel) and my attempt to rationalize it ended up a few damage points off even at level 13.

Cyanne, nice one. Is it just me or is there no juice in the upper right hand corner?

Cyanne, nice one. Is it just me or is there no juice in the upper right hand corner?

It last what... 5 seconds? Could have used the juice, used Meteor, used another one, then the juice died before Meteor(2) hit? :x

Yeah, the juice died before the meteors fell.

I just calculated the max and min of that meteor without mdef, and the damage range came out to be 77872-93476. The max neutral hit on the screenshot was 92481, and the lowest was 87659 (possible 88970 or 78970 on the bottom left). The formula does seem quite accurate, but the minimum seems a bit low.

Edit: Here's another picture of me hitting 73197, taken by a friend. Considering zakum's natural mdef and the fact that it probably has mdef up casted on it, that formula is REALLY accurate.

On an interesting side note, you can't hit 99,999 damage on a monster with an elemental weakness to the element you're using. I'm not sure why. Or, at least, it kind of seems like that in most cases.

Err. So what happens when you would've hit 66666x1.5?

Actually, the 99999 damage cap works so that if you would hit that much damage or more, it's capped off at 99999, THEN magic defense is applied.

On an interesting side note, you can't hit 99,999 damage on a monster with an elemental weakness to the element you're using. I'm not sure why. Or, at least, it kind of seems like that in most cases.

You can with critical for some reason.

 Spoiler

That's a lot of criticals for a 15% chance... but uh. That would signify that critical is applied after defense, which we already know. Doesn't explain why defense would come after range-correction, though. Assuming 1 and 99999 are done at the same time, that would mean a damage number annihilated by defense would be a MISS instead of a 1. And we know that's not true.