Final Form Magian Weapons

[Cream Soda]

Back in the testing spasm the boards were under a few months back, multi-hits were parsed by Mont and Masa and Soboro was found to have a 30/50/20 distribution without DA. I'm not sure of the exact page (or even thread) but it's probably on somewhere on page 4-5 of Advanced.

Or that

[Thorny]

6922.08 + 13768 = 20,690 to kuro's 21,800 if that's correct, 5.3% in kurodachi's favor.. very underwhelming

the real question is if people weren't willing to use soboro over hagun before because of tp feed, why would they want to use kurodachi now?

[Byrthnoth]

Ok, I messed up my original calculations because I'm apparently windows-calculator retarded this morning. Plugged it into excel and verified my equations are correct and got:

6-hit Soboro: 2.93 rounds to 100TP on average
6-hit Soboro with 15% DA: 2.908 rounds to 100TP on average

For anyone wondering, Soboro's Attack rounds to 100TP are minimized at 41% Double Attack - 2.8924 rounds to 100TP.

[Cream Soda]

I like to feed tp. One thing to consider when doing the comparison though is the possibility that zanshin cannot proc 40% of the time (on any round you attack twice).

As I've explained to a few ppl via PM to get their opinions. I've done some testing w/ zanshin.

Mnk/sam h2h = no procs
Sam/blm h2h(one punch) = procs
Sam/nin club = procs
Sam/nin club x2 = no procs

So I've come to the conclusion that either h2h/DW are specifically excluded, or any time you attack more than once per round, zanshin isn't possible.

[mdkuser]

6hit soboro
2hit possibilities: 1/3(27/81)
Triple x2 = 1/3 ^ 1/3 = 1/9 = .1111
Triple + Double = 1/9 = .1111
Double + Triple = 1/9 = .1111

3hit possibilities: 14/27(42/81)
double + double + anything = 1/9 = .1111
triple + single + anything = 1/9 = .1111
single + triple + anything = 1/9 = .1111
double + single + double/triple = 2/27 = .0740
single + double + double/triple = 2/27 = .0740
single + single + triple = 1/27 = .0370

4hit possibilities: 11/81
double + single + single + anything = 1/27 = .0370
single + double + single + anything = 1/27 = .0370
single + single + double + anything = 1/27 = .0370
single + single + single + double/triple = 2/81 = .0247

5hit possibilities: 1/81
single + single + single + single + anything = 1/81 = .0123

2 * .3333 = .666
3 * .5185 = 1.5555
4 * .1358 = .5432
5 * .0123 = .0615
Average Rounds: 2.8262

2.8262 rounds for 6hit soboro, 450 delay
3.7 rounds for 6hit kuro going by mdkuser's math, 450 delay
@140 STR(easy enough to change, i'm not on xi nor do i know normal STR+ numbers atm so just taking a wild guess)
15 fstr cap on kuro, 12 on soboro, both would cap on lolibri but not on gekko things really, for the sake of argument though

kuro becomes D:83 and soboro becomes d:52
140 * .75 = 105 WSC + 83 = 188 for kuro and 157 for soboro

19.74% ws damage for kurodachi
30.92% ws frequency for soboro
= 9.34% ws damage for soboro

assuming 4 fstr during tp phase, soboro becomes D:44 and kuro becomes d:72.. 72 * 1.4 = 100.8*pdif per attack round on kuro, 44 * 2 = 88*pdif per attack round on soboro

100.8 / 88 = 14.5% tp damage for kurodachi

with a 70/30 ws/tp split( i don't feel this is the right way to calculate damage, since gekko situations vary pretty heavily for ws/tp split and even changing GKs will change split), that'd put soboro about 6% ahead overall while 60/40 would put kurodachi 0.42% ahead

now, looking at it my way

1.5625 FTP
188 157
2.3 avg pdif on gekko

2.3 * 1.5625 * 188 = 676 damage gekko from kuro
2.3 * 1.5625 * 157 = 564 damage gekko from soboro

1.2 avg pdif during tp
120 damage per attack round from kuro
105.6 damage per attack round from soboro

3 minute analysis(meditate relevant)

march x2, hasso, haste, 20% gear, both 450 delay: .35 * 450 = 157.5 / 60 = 2.625 seconds per attack round
180 / 2.625 = 69 attack rounds
69 * 120 = 8280 damage from tp from kurodachi
69 * 105.6 = 7286.4 damage from tp from soboro

69 / 3.7 = 18.56 ws + 1.5 from meditate = 20 * 676 = 13520 ws damage from kurodachi
69 / 2.8262 = 24.41 ws + 1.5 from meditate = 26 * 564 = 14664 ws damage from soboro

In 3 minutes, soboro should deal 21,950 damage with a 33.2:66.8 TP:WS split.
In the same 3 minutes, kurodachi should deal 21,800 damage with a 38.0:62.0 TP:WS split.

If you change the STR and pdif around, kurodachi can pull ahead. Realistically, though, they're almost identical in overall damage output.

actually, the 3.77 rounds till 100 TP was for 95% acc; so for 100% acc you should take

3.58 for kurodachi
2.8x for soboro. ( with 0.3/0.5/0.2 distribution)

in the end soboro=hagun, less than 0.25% difference.. while kurodachi = +3.9% ( seems about rigth with your math. you basically compared 95% acc kurodachi to 100% acc soboro)

[Thorny]

used 3.7 not 3.77, just pulled it out of your post.. acc assumption wasn't specified

[Soho]

The proc rate on the h2h don't matter. Anything short of 100% and they suck

You misread my post.

I'm asking about 2H weapons (ie Polearm, GAxe, GKt etc)... Not H2H.

Has anyone verified that the proc rate on OAT 2H weapons is 40% yet?

[ZhiGaruda]

So are we all in agreement now that OAT GK is bazing awesome?

[Thorny]

So are we all in agreement now that OAT GK is bazing awesome?

if by 'bazing awesome?', you mean one of the weakest OAT weapons and a waste of krab shells vs bow, sure

[Shinryuu]

Pink words.
\(´A´)/
I really want to know if the bow is an ammo-eater.

[Byrthnoth]

Well, I get these numbers under the stated conditions, units are attack rounds per WS

0% DA, 100% ACC:
Kuro 6-hit: 3.7776
Kuro 5-hit: 3.056
Soboro 6-hit: 2.93
Soboro 5-hit (Yoichi onry): 2.417

15% DA, 100% Acc:
Kuro 6-hit: 3.58
Kuro 5-hit: 2.89
Soboro 6-hit: 2.909
Soboro 5-hit: 2.375


I'm still not sure how to correct for 5% misses. It would shift the distribution towards higher numbers and really hurt the Kuro 5-hit (where 2x2 is the only way to get 100TP). Well, actually I think I just figured out how to do it, but I don't know if I want to put the time in and modify my spreadsheets, especially because it messes up my main simplifying assumption. >:/ isn't like I even have SAM75

[Shinryuu]

I'm still not sure how to correct for 5% misses.

Divide by .95; actually, take Zanshin into account too.

[pchan]

soboro 6-hits, 95% accuracy 17% double attack ignoring zanshin :3.025275865 rounds
soboro 5-hits, 95% accuracy 17% double attack ignoring zanshin :2.486170610 rounds

[pchan]

Divide by .95; actually, take Zanshin into account too.

No.

[ZhiGaruda]

if by 'bazing awesome?', you mean one of the weakest OAT weapons and a waste of krab shells vs bow, sure

But still one of the best GK's for sure. bazing

[Demosthenes11]

if by 'bazing awesome?', you mean one of the weakest OAT weapons and a waste of krab shells vs bow, sure

how would you come to this conclusion after it was just math'd that it's the best non relic gkt? lol

we are talking about 10 krab shells, not 20k currency. stop throwing a fit

[pchan]

Now with 15% DA/95% acc

soboro 5-hit : 2.457205001
soboro 6-hit : 2.979435372

[Byrthnoth]

Now with 15% DA/95% acc: soboro 5-hit : 2.457205001

So, based off what I've been seeing and doing, Soboro SAMs should never sub WAR or probably even use Brutal Earring... unless they're doing a 5-hit with Yoichi. I don't know about you, but I get Brutal with /DRG being about a 0.3% increase in WS frequency. Sheeet, whatever Bushi does for DoT is probably better than that. /WAR and Brutal together give about a 0.7% increase in WS frequency.

Can you set DA% to 0 with 95% Acc and tell me what the Soboro 6-hit gets? Also, how are you calculating this?

Right now I figure I could break it up like this:
Set of probabilities - Separate into 2/3/4/5 hits after 2 attack rounds. Multiply by the probability distribution of hits and misses, re-separate into a new distribution of 1/2/3/4/5 hits. 5-hits (for a 6-hit) qualifies as 100TP, the other options move on to the next stage and repeat the process, but only the added hits are accuracy corrected. Continue until 5 attack rounds and call the difference between 100% and the sum of your probabilities 6+ attack rounds.
Right now I use the difference between 100% and the sum of my probabilities to avoid calculating one of the more difficult probability sets, so I don't really want to do this myself.


Edit: Ah-hah! Another option would be to just accuracy correct the Soboro distribution itself and use the new probabilities, which I could do in my spreadsheet right now if it wasn't for 0-hit cases.

[Thorny]

how would you come to this conclusion after it was just math'd that it's the best non relic gkt? lol

we are talking about 10 krab shells, not 20k currency. stop throwing a fit

not sure how i'm 'throwing a fit', i'm pointing out that it's not anywhere near as great as people are making it out to be.. much like the rest of these

depending on how the ranged ones work, they're both best nonrelic(and competitive with relic if the enm- isn't needed, but not sure if there are any places that applies really)

pole/scythe are both large improvements for nonrelic owners, pole could beat relic/mythic

dagger is a decent improvement for a nonrelic thf, and great for dnc

GK is marginal, barely beats soboro given 30-50-20 spread, and would lose to soboro if it was 25-50-25(which is well within margin of error for the posted test) or 33-33-33

not to mention GKs in general are for shitty pdif, not sure where you'd use it AND not care about tp given.. jol only weapon? or piercing resistant mobs in ein, about it

if you're a completionist sam that can never get amano(lol?) then get it and use it on jol or the rare situation you're fighting piercing resistant things.. better off using hagun for tanking, polearm or soboro /rng for zerging, polearm for weak things

[HarleyM]

if by 'bazing awesome?', you mean one of the weakest OAT weapons and a waste of krab shells vs bow, sure

has the math been done on the OAT bow vs the OAT gun?