Math calculations for FFXi

[Masamune]

Rdshift: have a look over there @ Allakhazam.com: Final Fantasy XI

it's far from "done" but yu'll understand what i meant by "missing features". i sent a new version to author, should be a release in next weeks maybe, up to him.

[Masamune]

On topic of math calculations, i got a weird result while parsing solo Lesser Colibris lvl63/64/65 just outside Whitegate @ Wajaom. When checking the crit rates i got for each of these Lcolis levels (images attached), one would say i'm about 16-17% crit rate, with samples size ~1k+ hits.

BUT

When i tried to calculate to retrieve these values, i got :

Elvaan THF75/nin, 4 merits crit, dex 70+8, equipped with Heart Snatcher main + Palhuwan Khazagand, should get :

- on Lcoli lvl63, AGI62 : dDEX=78-62=16 -> 6% from ref table + 4% merits = 10% + 6% HeartSnatcher + 1~5% from Body would get me minimum = 17% crit rate.

- on Lcoli lvl64, AGI63 : dDEX=78-63=15 -> 6% from ref table + same as above, would get me minimum = 17% crit rate.

- on Lcoli lvl65, AGI65 : dDEX=78-65=13 -> 6% from ref table + same as above, would get me minimum = 17% crit rate.

Conclusion: is Palhuwan body really that low of 1% crit rate (or even 0.5% lol) ? or did i miss something into my calculations/parse analyzis ?

P.S.: Reference Table i use is from one of the testings done @ wiki :

-120	5%
11	6%
21	7%
26	8%
31	9%
36	10%
38	11%
40	12%
42	13%
44	14%
45	15%
46	16%
47	17%
48	18%
49	19%
50	20%

[Callisto]

Does the level adjustment to cRatio happen after or before the ratio is capped? That is, with high enough attack, can you still hit the cap even fighting something several levels above you?

[ringthree]

During my research, i found 2 versions for handling new dmg results while equipped with a 2hands weapon (compared to old formulas) :
a) Redshift@KillingIfritWARforum is multiplying pDIFs by 1.05 (if i'm not mistaken)
b) Seedo@KillingIfritMNKforum raised cRatio cap @ 2.064 instead of 2, Crit pDIF mod @ 1.06 instead of 1, Max Crit pDIF cap @ 3.15 instead of 3. (and that for ANY weapon, not only 2handers)

The rest of the formulas are same as the old ones.

Personnally, i use version b), matches pretty well my parsed results for TPing (i didnot verified thoroughly for WSs yet).

I tested the spread on the pDIF and got lower and higher damage hits that could not be accounted for by either of those statements (at least I am pretty sure).

[Callisto]

When I tried I got something around the same as the 1.05(I found the multiplier to be 1.26, it would math out the same to just multiply the end result by 1.05), though I also had seen lower than the previous low cap iirc, and it made me believe that the -.8 amount for pDIFMin was also increased, I want to say it was like .88, I'll try to dig up wherever the hell I posted that stuff later.

[Masamune]

Ringthree and Callisto, please post yur parsed results, i'm very interested into these lower and higher dmg numbers than calculated ones with either new pDIFs a) or b).

Post also yur merits and gears.

Personnally i got my high/low numbers from the parses posted above + as war/sam perdu voulge + war/nin maneater/joyeuse, both @ Gcolis 81/82. First, i found higher and lower dmg numbers than the ones calculated with old formulas, until i met Seedo @ KI mnk forum who gladly submitted me these new caps explained @ post above. My results then totally matched, thks Seedo.
Also, i have to say i fucked up 3 meripo parses before getting one eligible for analyzis, (ie no Dia on mobs, no cor only brdsX2 and large enough sample size for EACH mob levels ~ 1k+ hits). For example, one dia is enough to get yu chances to get higher dmg numbers than calculated ones; or also eating debuffs like str/dex down to get lower numbers than calculated... (/sigh i can't post my parses, even zipped they higher than forum reqs...)

[Callisto]

Ringthree and Callisto, please post yur parsed results, i'm very interested into these lower and higher dmg numbers than calculated ones with either new pDIFs a) or b).

Post also yur merits and gears.

Personnally i got my high/low numbers from the parses posted above + as war/sam perdu voulge + war/nin maneater/joyeuse, both @ Gcolis 81/82. First, i found higher and lower dmg numbers than the ones calculated with old formulas, until i met Seedo @ KI mnk forum who gladly submitted me these new caps explained @ post above. My results then totally matched, thks Seedo.
Also, i have to say i fucked up 3 meripo parses before getting one eligible for analyzis, (ie no Dia on mobs, no cor only brdsX2 and large enough sample size for EACH mob levels ~ 1k+ hits). For example, one dia is enough to get yu chances to get higher dmg numbers than calculated ones; or also eating debuffs like str/dex down to get lower numbers than calculated... (/sigh i can't post my parses, even zipped they higher than forum reqs...)

I was unfortunately on 360 at the time and did it using pen & paper(lol), so no parses to back up my stuff

[Milkster]

I know this will probably sound noobish but I tried the search function and wound up in 2006 trying to find information on sTP. This is my first meele job to 75 and I was wondering how do I calculate how much STP I need to reach a 5-6 Hit. The delay off weapons and how that figures in etc.

[Ultimagi]

Store TP = CEILING( CEILING( (100-x) / y, 0.1) / z, 0.01)*100-100

Where x is the number of extra TP gained from multi-hit weapon skills (0 for pure hit build),
y is the number of hits you want to reach 100% TP in,
and z is the base TP gained from the weapon (these values are listed for every weapon on wiki).

If you don't know how to use CEILING(x, y), it means round x up to the nearest decimal place indicated by y.

[urat]

I know this will probably sound noobish but I tried the search function and wound up in 2006 trying to find information on sTP. This is my first meele job to 75 and I was wondering how do I calculate how much STP I need to reach a 5-6 Hit. The delay off weapons and how that figures in etc.

http://www.bluegartr.com/forum/76242...just-sams.html

If you scroll about halfway down the OP you'll find a decent equation for finding out how much tp you need for your 5-6 hit, but its pretty simple.

You need 100 tp to ws, so 100/5 = need at least 20 tp per hit.

So simply, 20 divided by your tp/swing, then times that number by 10 and roof it (round up to nearest whole since 1 tp store is actually 0.1% boost) and that'll be how much sTP you need.

You may want to take a look at some of that topic and acc vs haste topics around here, since in some circumstances, mainly when your acc < 90%, acc and haste gear is better than tp gear, since haste increases dot AND tpm, as does acc, whereas stp only increases tpm.

[pahnphoenix]

i don't know how 2h pDIF works, but 1h pDIF seems to work exactly the same as it did before the 2h update. the only difference in the 1h damage equation is that now all physical damage is multiplied by a random number between 1.0 and 1.05 (that's why you're seeing 3.15 as new crit cap).

or at least, that's what some JP guy found, and it makes the most sense considering the loss of a cluster of hits at 1.0 pDIF in subsequent parses (but the retention of a cluster at 1.0-1.05 if i recall). however, i believe people have reported NQ hits that dip below expected minimums and not seen the expected maximum if it were simply 1.0-1.05*old pDIF (i believe i reported a low below old minimum, but it's been too long). so, the theory doesn't look perfect.

as for 2h pDIF, i'm at a loss, but admittedly i haven't followed much of the testing (if it's even being done). i just use cRatio for that, and assume it caps higher than 1h cRatio. i could be catastrophically wrong i suppose. it would be a big deal if 2h cRatio capped at 2.0 and just happened to randomly jump up to 2.6~ when converted to pDIF. it also wouldn't make much sense, and i'd hope people would notice "hay guise, i just added 60 ATT and nothing happened what the fuck?"

[Milkster]

pDIf and all the pXXX/fXXX modifications etc. Or if anyone could send me a PM about it.