New Heal Formula

Wow thanks!

Data get.

LVL: 50
INT: 262
LUK: 6
MATK: 350

Fighting Coolies (level 57, 180 mdef) for around an hour.

Projected: 355-1347
Observed: 254-1250

Is that without defense?
With defense, it'd be 247~1257.
That's without adjusting for the level difference (which I don't know how to do). Maybe magic defense doesn't have a level component at all.

Thanks for the data. Keep it coming! -needs that unmaxed Heal-
I'm thinking of remaking my cleric though, so I could get that.

That was projected without defense, I figured that that would be most helpful to figuring out exactly how level is a factor (or if the defense reduction is correct, for that matter).

Well.
Using the weapon defense formula on it gives 229~1191, which is quite a bit off.

lol. Maybe heal does ignore.

I'll do ya a favor Sap. My cleric should be lvling tomorrow after I get 40 coins, I'll go to something like skeles (maybe not that high), but you get the point, and see if level makes a diff.

Also, I'd think there would be a full multiplier, like every monster after the first takes the same amount of reduced damage. Its hard to test that with heal, you'd need a HELL of a lot of screenshots, then you have to tell which monsters are in what order. That doesn't account of the lag when you pressed the button to when the damage appared to how much they moved.

...You're going to miss, Techno. I hope you realize that. Go to V/H or something.

Also, with area mobbers, all the monsters get hit at the same time. There is no first monster. Every one takes the same amount of reduced damage depending on how many were hit in total.

Okay, so I've been doing a little bit of messing around (and I do stress little - Scania is impossible with KSing), and I bring you 2 Coolie hits!

I originally also had a question as to whether or not misses count for the damage decrease, but it appears that it's applied equally to all monsters as opposed to something like Iron Arrow. Hopping into a mob of 5 yields much more stable damage than attacking a single target. It doesn't decrease per target, or so it would seem. This does need some more testing, but perhaps my results as I increase the number of targets I hit will tell us.

Anyway:

Level: 50
INT: 262
LUK: 6
MATK: 350

Enemy level: 57
Enemy MDEF: 180

2 targets, observed damage: 167-967

Those should be pretty close. I thought I saw a couple on the low end that could've been lower, but I wasn't certain due to two numbers overlapping or whatever, so I didn't count them. Should keep that in mind as far as calculations go.

Is that without defense?
With defense, it'd be 247~1257.
That's without adjusting for the level difference (which I don't know how to do). Maybe magic defense doesn't have a level component at all.

Thanks for the data. Keep it coming! -needs that unmaxed Heal-
I'm thinking of remaking my cleric though, so I could get that.

Magic defense doesn't have a level component. I basically do the same damage on a Viking as I do on a Snail. >_>

Level component = you're below the monster's level and the game nerfs your damage because of it.
If you're level 146, your damage to a Viking won't get a level reduction regardless.

And Bui, are you sure you hit a 967? I'm getting strange results here.

Positive. I made absolutely certain that I witnessed the numbers I submitted.

However, the 96x might've been with a miss on another target, I still miss sometimes. Could be that misses affect things oddly.

EDIT: I think I'll do some more testing since I can't sleep anyway. I'll record min/max results for 2 hits and 1 hit/1 miss. That might help more.

EDIT2: Pretty sure misses are a nonfactor. I'm getting 94x fairly often with both single miss and no miss between 2 targets.

 1 hit 2 hit red hit blue hit

 Just the 1

Didn't feel like waiting around for max damage on both when it seems as though they're the same. :[
Oh yeah, and I saw a 970 in there. My earlier data showed 967 as the max observed, so it's only three points, but ya.

EDIT3: ...971? o_O It's almost like it keeps rising even though I spent around an hour on 2 targets earlier.

... 972. I should almost stop looking for LOW damage because it's hard to process that and take a screenshot of high damage if I see it.

ok so. It seems much harder to hit minimum than maximum. Either that or I'm just having the opposite luck of earlier (?).

 185/Miss

 176/other hit

I think it's pretty safe to say that missing doesn't matter, which is good for my purposes. I'll check out 3 mobs in a bit. I can produce screenshots of any of the data I put forth, so just let me know if something else seems wonky.

Hopefully a 967 works for you - I have observed up to 972 though.

 FINALLY GOT YOU, peach

Surprisingly tired, I might be able to sleep now. If so, I'll do 3 targets later. If not, you'll know.

Level component = you're below the monster's level and the game nerfs your damage because of it.
If you're level 146, your damage to a Viking won't get a level reduction regardless.

And Bui, are you sure you hit a 967? I'm getting strange results here.

Lemme rewind time for a moment. xD

 Spoiler

If level played a role like in physical damage, I wouldn't be able to hit that much at that level.

Ah, I see.
So magic accuracy is really the limiting factor, not level defense.

@ Bui - That's odd, then. According to your data, the 2 monster multiplier must be >0.788403, but it also must be <0.773176. Contradiction.

And yes, I'm sure I didn't mix the two up.

I'll go to DTs or something to bring in the final results, but that pic up there is pretty difinitive.

BTW Sap, is that magic formula (the quadratic) the real one, or an assumed?

It's copied off SW.

Got some more data:

INT: 331
LUK: 77
M. Atk: 429

Predicted Damage: 909 ~ 2444
Predicted Damage on 4 targets: 576 ~ 1549
Observed Damage on 4 targets: 455 ~ 1408

This was done on Voodoos, which have 210 magic defense.

So with defense applied in full, that's -126~-105
Or, on your 4 target range, 450~1444
A pretty nice min damage fit there, though I can't say the same for the max.

Any specific things you'd like me to test before I level? I'm starting to get up there on % and I'd think if this is anomalous, it's good for data. Planning on messing around with 3 targets after an APQ.

While this formula is much cleaner, I still think it's somewhat inaccurate because of the /999. That just feels like it's wrong. A number should probably be adjusted here or there. Nice work though.

Divide by 1000 instead?

I had 1000 at first, but one of my data samples only fits if I make it 999.
That particular sample is only off by 1 damage though, so it could easily be a rounding error.

Bui: Yes, more targets would certainly be helpful.