Magic Accuracy

taking out the headband and gaining 1 luk as well - which should be +1 macc, so the 1 acc is worth around 5.5-6.5 accuracy.


Blessing of the Spirit adds acc too, is this the only char?

Ohhh, totally spaced that. Level 5 Blessing on my character. I'm guessing level 1 or 0 on KMSTryst.

Just to be 100% clear you want the 6th stat down:


That is 100 in this picture.



If we assume this char has no blessing, then magic avoid is
INT + 2*LUK + EVA
Ie in this pic, 69+2*18+0
In other pics, 61+2*4+5 (from Bless)

Dunno why acc is proving harder to figure out.

I could have sworn that's what I had before... Ugh. This is getting annoying.

125 macc in this image, correct? If so, my OP was correct, just miss-labeled.

OK so if the table is correct now, the formula might be:
Mag acc = Floor( 1.2*INT + LUK + gear acc)

So luk mages will be missing more than lukless mages! Really nice...

It's still beneficial to use level 80 class equips since they get a better bonus from the potential system (and can be makered with 2 crystals) and possibly the level 163 estaff.

So luk mages will be missing more than lukless mages! Really nice...

Not necessarily. LUK could also appear in the hit-rate formula, for example hit rate could be dependant on (MagAccuracy+LUK-Level) or some such, so that high level lukless would still be in trouble.
Early testers said LUKless couldn't hit. So until different info trickles in that's what I will assume.

I got some more hit-rate data. It's really weird.

Char Lv (INT / LUK)	Magic Acc	Mob Lv	Attacks (Hit / Miss)	Ratio
51 (222 / 49)	315	52	104 (99 / 5)	0.9519
51 (222 / 49)	315	54	104 (90 / 14)	0.8654
51 (222 / 49)	315	56	104 (75 / 29)	0.7212
51 (222 / 49)	315	58	104 (70 / 34)	0.6731
51 (230 / 55)	331	52	104 (98 / 6)	0.9423
51 (230 / 55)	331	54	104 (86 / 18)	0.8269
51 (230 / 55)	331	56	208 (156 / 52)	0.7500
51 (230 / 55)	331	58	104 (70 / 34)	0.6731

Seems pretty erratic when dealing with abrupt accuracy changes, but I kind of see a trend where the difference between the two accuracies gets smaller as the player's level approaches the mob level. I'll retest these later.

Your data seems to support the "Always hit stuff under or equal to your level" theory. I like that system. xP

It's still beneficial to use level 80 class equips since they get a better bonus from the potential system (and can be makered with 2 crystals) and possibly the level 163 estaff.

You'd need to get roughly 85 or so INT bonus from Potential gear to make up for level 80 gear, right? That's what, +20%? I have no idea how practical that is.

Your data seems to support the "Always hit stuff under or equal to your level" theory. I like that system. xP



You'd need to get roughly 85 or so INT bonus from Potential gear to make up for level 80 gear, right? That's what, +20%? I have no idea how practical that is.

considering a lvl 70+ gear can give you minimum +9% and a maximum +27% in stats it is

OK so if the table is correct now, the formula might be:
Mag acc = Floor( 1.2*INT + LUK + gear acc)

This also fits the accuracy for my mage.

EDIT: Hit ratios seem to fit the curve: [20 - (Mob Lv - Char Lv)] / 20

This also fits the accuracy for my mage.

EDIT: Hit ratios seem to fit the curve: [20 - (Mob Lv - Char Lv)] / 20

Linear graph? Sounds fair enough.

EDIT: Hit ratios seem to fit the curve: [20 - (Mob Lv - Char Lv)] / 20

It doesn't even factor in magic accuracy. o-o

Seems to fit really well, though. What's the point of having "magic accuracy" if it's entirely level dependent?

It doesn't even factor in magic accuracy. o-o

Seems to fit really well, though. What's the point of having "magic accuracy" if it's entirely level dependent?

To kill off Newt sniping? I don't know, I can't think of any better reason. Pretty funny how they list accuracy values in the stat window and have it do nothing.

Not even Nexon is that irrational.... I might make a mage and run some accuracy test before and after adding something like 60 INT to see if there's any significant difference in hit ratio.

Does anyone here remember / know how to run a Binomial Hypothesis Test? It's been like two and a half years for me.

Maybe someone in kMS who can get ahold of the new aufheben circlet and/or ludibrium cape can do some testing on monsters a certain number of levels above with and without the item since those items would seemingly make the character's accuracy differ enough for significant results.

Char Lv	Mob Lv	Macc	Hits	Misses	Ratio	Estimate
10	26	28	31	80	0.279	0.2
10	26	82	20	64	0.238	0.2

Derp.

Char Lv	Mob Lv	Macc	Hits	Misses	Ratio	Estimate
10	26	28	31	80	0.279	0.2
10	26	82	20	64	0.238	0.2

Derp.

Doubt that Magic Accuracy will actually decrease hitrate. Perhaps the hitrate varies and the higher the target's level compared to yours, the greater the variation of the hitrate? Maybe Accuracy/Magic Accuracy do not do anything unless your level is equal or greater than the target's, and the hitrate is not equal to 100% in this scenario but just very close to it. The Accuracy/Magic Accuracy value could affect the actual hitrate, and each point will contribute very little to the hitrate (which explains why there's a need for the 400 Acc hat and 420 Acc cape).

Could you do it again or something? You'll probably get rather different results...

It doesn't even factor in magic accuracy. o-o

Seems to fit really well, though. What's the point of having "magic accuracy" if it's entirely level dependent?

Maybe accuracy matters for low-luk? Not much data here with luk-less yet, except at very low level.
The formula could be much more convoluted, but simplify to what you have when luk is high enough.
Or they could use different formulas depending on whether the character level is above or below the mob level.

If that new test is Lukless, perhaps the hit rate went down due to greater spread between int and luk.

It was shown that physical accuracy did not seem to matter for mobs with lvl below char lvl.
http://ellinforest.com/forum/showthread.php?t=7541

You need a lot more than ~100 trials to accurately determine a hit rate. Go for 400-500.

That test was done on a lv 10 Cygnus before and after Auto-assigning stats. 28 macc was with 20 INT.

I'll try it again, but with more hits. The test was done in the Golem MD, so sniping is easy.

Char Lv	Mob Lv	Macc	Hits	Misses	Ratio	Estimate
10	26	29	90	310	0.225	0.2
10	26	83	80	320	0.200	0.2

What the pineapple . . . .

I don't claim to have flawless counting skills, so I assume I miscounted here and there. I know there was a 400 maximum hit because I attacked until the anti-bot system kicked in. (100 attacks * 4 hits per Magic Claw = 400)

Throwing on +/- 5 to each hit rate:

Macc	Hits	Ratio
29	85	0.213
29	95	0.238
83	75	0.188
83	85	0.213

I have a hard time believing that Nexon would create a new stat window, include new stats, and then not use them....