How to apply DA/TA to multi-hit weapons mathematically?

[Cloudius]

I've been working on a Drk zerging spreadsheet to compare stuff like weapon vs weapon,
haste vs HP vs DA/TA vs Acc.

I'm almost done, but I can't figure out how to apply DA/TA to multi-hit weapons.
The best info I could find was this:
www32.atwiki.jp/studiogobli/pages/17.html

* As for attack frequency occurrence ratio of [rideiru] 30: 50: With 20 average 1.90 attacks
o With DA probability 10% 1.91 times, with DA probability 20% 1.92 times


* As for attack frequency occurrence ratio of [jiyuwayusu] 50: With 50 average 1.50 attacks
o With DA probability 10% 1.55 times, with DA probability 20% 1.60 times


* As for attack frequency occurrence ratio of [kurakenkurabu] 5: 15: 25: 25: 15: 10: 3: Whether even average 3.82 attacks of 2 the [do] [tsu] you saw
o Therefore DA probability it rises with getting near to 2.00, frequency of attack it decreases

The 1st part was good info, stuff that do 1-3 has have a 30:50:20 ratio, and an avrg of 1.9 that needs 10% DA to add a 0.01 to it. But what about TA?

The 2nd part about joytoy was too easy and I dont rly care about. But the third part about KC is kinda add, an avrg of 3.82? and I didn't get what they're talking about DA and 2.00, also what about TA?

Any info, or a better source, or a better translation would be greatly appreciated!

[VZX]

I'll do the 3rd part:

The attack frequency distribution of kraken club is as follow :
5: 15: 25: 25: 15: 10: 3: 2 with the average round/swing 3.82
Hence, as DA increases, the average round/swing is getting closer to 2.00

3.82 = .05*1 + .15*2 + .25*3 + .25*4 + +.15*5 + .10*6 + .03*7 + .02*8
, if you are confused on that part

EDIT:
As for double attack and triple attack, it overwrites your multi-hit swing to 2 swing if it proc.
i.e.:
Assume 10% DA rate on KC
There is 90%*5% chance it'll do 1 attack/round and there is 10%*5% chance it'll do 2 attack/round
There is 90%*15% chance it'll do 2 attack/round and there is 10%*5% chance it'll do 2 attack/round = 15% chance to do 2 attack/round
There is 90%*25% chance it'll do 3 attack/round and there is 10%*25% chance it'll do 2 attack/round
There is 90%*25% chance it'll do 4 attack/round and there is 10%*25% chance it'll do 2 attack/round
...
There is 90%*2% chance it'll do 8 attack/round and there is 10%*2% chance it'll do 2 attack/round
And then average it

[Cloudius]

Thank you VZX for your reply.

I'll do the 3rd part:

3.82 = .05*1 + .15*2 + .25*3 + .25*4 + +.15*5 + .10*6 + .03*7 + .02*8
, if you are confused on that part

I understood that, but what I meant was that it's different from what we know and use.
=((0.05*1)+(0.1*2)+(0.15*3)+(0.2*4)+(0.2*5)+(0.15* 6)+(0.1*7)+(0.05*8 ))
This is what is usually being used (to my knowledge), and it goes well with the 4.5 avg we all know [((1+2+3+4+5+6+7+8 )/8 ) or ((1+8 )/2)]. So what's up with that avrg and ratio on studio global?

EDIT:
As for double attack and triple attack, it overwrites your multi-hit swing to 2 swing if it proc.
i.e.:
Assume 10% DA rate on KC
There is 90%*5% chance it'll do 1 attack/round and there is 10%*5% chance it'll do 2 attack/round
There is 90%*15% chance it'll do 2 attack/round and there is 10%*5% chance it'll do 2 attack/round = 15% chance to do 2 attack/round
There is 90%*25% chance it'll do 3 attack/round and there is 10%*25% chance it'll do 2 attack/round
There is 90%*25% chance it'll do 4 attack/round and there is 10%*25% chance it'll do 2 attack/round
...
There is 90%*2% chance it'll do 8 attack/round and there is 10%*2% chance it'll do 2 attack/round
And then average it

This I also understand, but I guess I worded my OP wrong, so I apologize.
I meant how do we make this a function I can use?
i.e KC's avrg = 4.5, DRK has 5% DA and 2.5% TA, so his new avrg = X swings/round

I got everything else for the spreadsheet worked out, but Im blocked by this DA/TA thing.

[Suiram]

The point they are making is that you don't want to use DA, since DA proc prevents a multihit kraken swing, so look at it like a gauge. On the left side of the gauge (0% double attack) you are swinging 3.82 times per attack round. If you're on the right side of the gauge (100% DA) you're swinging 2.0 times per attack round. You could calculate what your average number of swings are in the middle, but why bother since you know it'll be less than the 3.82 you get with no DA.

Likewise for TA, Except this time 3.82 is on the left and 3.0 is on the right.


If you really want to make the spreadsheet anyway, you'll have to figure out which of DA or TA is checked first when both are present. If you only have DA rate you want:
(DA rate)*2 + (1 - DA rate)*(3.82/1.9/1.5 depending on weapon)

For triple attack alone you'd want:
(TA rate)*3 + (1 - TA rate)*(3.82/1.9/1.5)

If DA is checked first you'd want:
(DA rate)*2 + (1 - DA rate)*(TA rate)*3 + (1 - DA rate)*(1 - TA rate)*(3.82/1.9/1.5)

And if TA is checked first you'd want:
(TA rate)*3 + (1 - TA rate)*(DA rate)*2 + (1 - DA rate)*(1 - TA rate)*(3.82/1.9/1.5)

[Omnipotent]

That's what she said.

[Suiram]

On a side note, this Studio Gobli information about the multi-hit distribution finally puts to rest the age-old question, "What if I could dual wield kraken clubs lol"

Dual-Wielding is the winner hands down, since 3.82 swings on a Del264 weapon is 1.45 swings per 100 delay. Meanwhile, the distribution of swings with two krakens and the 8-swing attack-round cap results in an average of 6.935 swings per attack round. With 15% DW that's 448 delay, and with 20% it's 422 delay, working out to 1.55 and 1.64 swings per 100 delay, respectively.

[Yarko]

I've recently done some Kraken Club attack rates test, and although I realize 642 round of attacks aren't a huge sample, the # of attacks rates resulted to be for me:

2,6 : 10,44 : 22,59 : 31,46 : 20,56 : 9,5 : 2,02 : 0,62

which are somewhat different from gobli's 5: 15: 25: 25: 15: 10: 3: 2

I'll try to expand my sample once i have more time, but was wondering if anyone else has done their own tests and has their own data?