What's Better?

[Howard Roark]

a bit late but: for weapons with same TP gain as destroyers (delay 41-59) there are several STP tiers that are nice to achieve, where your numbers of hits to reach 100 get a larger boost on average than other numbers. 21 STP is a tier for example ( full usu + rajas+brutal+ WT), going from 19 to 21 stp by using WT is about 5-6% WS frequency boost, while you expect it to be only 2% normally. This is easilly seen by counting teh hits needing to reach 100: it drops by one for almost all 8, 7 and 6 - hit asuran fists. The previous tier is 16 IIRC, and the next is above 30.

On the other hands, the numbers are completly different for H2H with spharai's delay for example ( never computed those).

thanks, was beginning to wonder if anyone would reply

didn't know about the tiers either, thx

[Scritch]

Aurore Doublet or Acc+10 Wardecors for PUP? (Other augment is used for Pet: stats).

Relevant stats:

Aurore: STR+4 DEX+4 Accuracy+8 Double Attack+2%

Wardecors: STR+5 DEX+5 Accuracy+10

For WS and situations where I can't get away with Goliard Saio.


EDIT:

Also, Wagh/Hades or a DMG+22 Delay+61 Taipan Fangs?

I finished the Pet: MAB trial for my fangs, and I want to know if it'd be better to use them over Wagh or Hades until I can finish another magian

[Angelcaido]

V.fork/pole grip with turban/aurum/3xhomam

or

Stingray(STP+8 )/pole with 5xperle

or

V.Fork/rose with 5xperle

Finally got the v.fork last night and wondering if i have a use for it.

[Nightfyre]

Finally got the v.fork last night and wondering if i have a use for it.

town gear

[Nameless]

Yea, Stingray easily beat v fork.

[Vodou]

a bit late but: for weapons with same TP gain as destroyers (delay 41-59) there are several STP tiers that are nice to achieve, where your numbers of hits to reach 100 get a larger boost on average than other numbers. 21 STP is a tier for example ( full usu + rajas+brutal+ WT), going from 19 to 21 stp by using WT is about 5-6% WS frequency boost, while you expect it to be only 2% normally. This is easilly seen by counting teh hits needing to reach 100: it drops by one for almost all 8, 7 and 6 - hit asuran fists. The previous tier is 16 IIRC, and the next is above 30.

On the other hands, the numbers are completly different for H2H with spharai's delay for example ( never computed those).

I'd like to see the math on this. While it's obvious that it's better to WS in fewer hits, all things being equal, I tried to recreate what led you to these particular tiers and I'm missing something.

[drwaffles]

Just checking something regarding SAM GKTs, looking specifically at two Magian weapons:

DMG:92 Delay:450 Additional effect:Weakens defence +15

vs.

DMG:91 Delay:464 Store TP +5

I have Rajas and Usukane Haramki, but not Sune-ate. From what I can see, I can get a 6-hit build with the first weapon if I WS in Haramaki, and with the second I can use Sword Strap (for 451 delay) and still 6-hit without having to WS in Haramaki. So basically it boils down to an extra 2 STR on WS from Osode vs. 1 more base damage, 1 less delay, 2% DA from Pole Grip for both TP and WS, and the fairly negligible additional effect. Thus, the first weapon seems superior. I thought I'd check since people were drooling over the STP+5 GKT when it came out, so please correct me if I'm wrong on anything.

[Deader2818]

Why do you need to WS in Usu body for the first one?

I assume you TP in Brutal/Rajas/Usukane body which gives you 16.9 per hit.

16.9x5 = 84.5

If you WS in only Rajas and Brutal thats 16.2.

So 84.5 + 16.2 = 100.7

[Medelus]

How does X's knife compare to the STR+6 augmented Kila for Mandau offhand?

[Ned]

Given the completely horrendous look of full Perle and the "every allatard in the world has it" stigma, can someone please point out some ridiculous flaw in my math and tell me that Turban/Haub+1/Homamx3 is still better for PLD Atonement TP? Please?

Homamx3, +1Haub, Turban = 14% haste, 25 acc & 6 dex (=28 acc) , 6 STR, 15 atk (inc str), 110mp/hp
Full Perle =14% haste, 18 acc & 14 dex (=25 acc), 22 str, 36 Atk (inc str)

So perle has same haste, very nearly same acc 1 handed (same if 2 handed), loads more attack, more def and 2% crit. The only thing u get from the mix set is 100 ish HP&MP.

I can't see any time you would use homam mix set unless that HP/MP is important??

[Vodou]

Just checking something regarding SAM GKTs, looking specifically at two Magian weapons:

DMG:92 Delay:450 Additional effect:Weakens defence +15

vs.

DMG:91 Delay:464 Store TP +5

I have Rajas and Usukane Haramki, but not Sune-ate. From what I can see, I can get a 6-hit build with the first weapon if I WS in Haramaki, and with the second I can use Sword Strap (for 451 delay) and still 6-hit without having to WS in Haramaki. So basically it boils down to an extra 2 STR on WS from Osode vs. 1 more base damage, 1 less delay, 2% DA from Pole Grip for both TP and WS, and the fairly negligible additional effect. Thus, the first weapon seems superior. I thought I'd check since people were drooling over the STP+5 GKT when it came out, so please correct me if I'm wrong on anything.

The point is that you can make a viable 5-hit build off the second one. All pales after that.

[buddha_masta]

The point is that you can make a viable 5-hit build off the second one. All pales after that.

Anyone who has completed 5 sTP Keitonotachi, how do WS numbers compare with Hagun? On the average merit/exp, Dyna, Abyssea, etc. lvl mob. Be nice to see how much we're giving up to attain 5 hits.

Another question: What kind of situations will still allow for use of 6-hit Hagun? I'd assume the TP bonus will still be very worthwhile, compared to 5-hit Keitonotachi, on really high DEF mobs?

[Raen Ryong]

How does X's knife compare to the STR+6 augmented Kila for Mandau offhand?

X's Knife's DoT is far inferior to the augmented Kila so I'd choose the Kila easily.

[Haborym]

Ares's Flanchard or Onyx Cuisses for Guillotine?

[Mercilessturtle]

Onyx unless you really need the 4 acc.

[Medelus]

X's Knife's DoT is far inferior to the augmented Kila so I'd choose the Kila easily.

I'm thinking so too. Sort of hoping that the crit damage increase stacked with Mercy Stroke's after affect would make up for that, but I think such a large difference in DoT and STR is hard to beat for basically every situation. I just don't want to say goodbye to X's Knife.

[Atoreis]

Anyone who has completed 5 sTP Keitonotachi, how do WS numbers compare with Hagun? On the average merit/exp, Dyna, Abyssea, etc. lvl mob. Be nice to see how much we're giving up to attain 5 hits.

Another question: What kind of situations will still allow for use of 6-hit Hagun? I'd assume the TP bonus will still be very worthwhile, compared to 5-hit Keitonotachi, on really high DEF mobs?

STR pdif noDA noDA(Overhelm) DA DA(Overhelm)
130 1 339(315) 404(375) 511(505) 608(601)
130 2 679(631) 808(751) 1022(1011) 1217(1203)
130 2,7 916(852) 1090(1014) 1380(1365) 1643(1624)
150 1 364(336) 433(400) 548(538) 652(641)
150 2 728(672) 866(800) 1097(1077) 1305(1282)
150 2,7 983(908) 1169(1081) 1480(1454) 1762(1731)
178 1 398(365) 474(434) 600(585) 714(696)
178 2 797(730) 948(869) 1200(1170) 1428(1392)
178 2,7 1076(986) 1280(1174) 1620(1580) 1928(1880)

normal numbers - HAGUN
number in () - 91 dmg GKT
all 100TP, all with gorget

Hmmm Its auto editing and its not clear :/ Well i supose if someone realy want to read he will anyway but in %

130str Hagun beats 91gkt by: 7,5% noDA and 1% DA
150str Hagun beats 91gkt by: 8,2% noDA and 1,7% DA
178str Hagun beats 91gkt by: 9% noDA and 2,6% DA

regardless of pdif advantage is the same(which is kinda obvious).

[lolwut]

Cobra Harness or Aurore Doublet for Joyeuse TPing?

If it saves a hit on average to 100 (which it should), Cobra wins while giving a bit of extra acc as well.

Aside from the legs, Aurore pieces are typically beaten by the older stuff, which is all pretty easy to get already.

[mdkuser]

I'd like to see the math on this. While it's obvious that it's better to WS in fewer hits, all things being equal, I tried to recreate what led you to these particular tiers and I'm missing something.

Quick answer :

list all TP return from asuran fists, for each one calculate the number of hits required both with 19 stp and 21 stp for 100% TP or more :

format : return from asuran fist with 12 stp, number of hit needed with 19 stp, number of hits need with 21 stp

                              0., 18, 17  *
                             1.1, 18, 17 *
                             2.2, 17, 17
                             3.3, 17, 17
                             4.4, 17, 17
                             5.5, 17, 17
                             6.6, 17, 16 *
                             5.4, 17, 17
                             6.5, 17, 16*
                             7.6, 16, 16
                             8.7, 16, 16
                             9.8, 16, 16
                             10.9, 16, 16
                             12.0, 16, 15 *
                             10.8, 16, 16
                             11.9, 16, 15 *
                             13.0, 15, 15
                             14.1, 15, 15
                             15.2, 15, 15
                             16.3, 15, 15
                             17.4, 15, 14 *

the '*' are the situations where you do save a hit. All you have to do after that is to check the probabilty of those situations to happen and you'll find that it saves a hit ~73.5% of the time and that the average number of hits goes from 15.10 to 14.36.

15.10/14.36=1.051 so that's ~5% ws frequency boost. Now this math was 'quick' and does not account for over TP. If you want that see below but you find about the same results.

LONG MATH (and exact math):

Spoiler: show

If you are familiar to independance, conditionnal independance, or markov properties you can find the equations leading to the exact value. Exemple :


Call F(x) the number of round needed on average starting from x TP - in other words, engage a mob with x TP several thousands of time with 95% acc, 17.5% KA, 5% DA, and calculate the number of rounds for each time you engage, then divide to get the average.

Call 'tp' the TP you get per hit (same for KA and normal hits).

Conditionnally to the first round returning 0 TP, you need this round with zero tp + F(x) rounds => F(x)+1
Conditionnally to the first round returning 'tp' TP, you need this round with tp + F(x+tp) rounds=>F(x+tp)+1
Conditionnally to the first round returning '2*tp' TP, you need this round with 2*tp + F(x+2*tp) rounds=>F(x+2*tp)+1
...
Conditionnally to the first round returning '5*tp' TP, you need this round with 5*tp + F(x+5*tp) rounds=>F(x+5*tp)+1

So the equation looks like

F(x)=p_0 *( F(x) +1 ) + p_1 * ( F(x+tp) + 1 ) + p_2 * ( F(x+2*tp) + 1 ) ... + p_5*( F(x + 5*tp) +1)

where p_0=probabality that a round has zero hits
p_1=proba of 1 hit
... p_5=proba of 5 hits


This a recursive equation :

If you want to know F(x), you need to know F(x+tp),F(x+2*tp) ... F(x+5*tp).
If you to know F(x+tp), you need to know F(x+2*tp),F(x+3*tp) ... F(x+6*tp).
and so on ...
Knowing that F(y)=0 when y >=100 tp, you can probably guess you will quickly find how to calculate F(x) for any x. The only difficulty is p_0,p_1 ... p5 which is not too difficult. A simple way is to introduce the generatrice function of N, the number of hit per round and calculate E(r^N) : this is a polynomial and it's coefficients correspond to p_0, p_1, etc ...

For instance with normal H2H, 5% DA, 95% acc, 17.5% KA
p_0=0.189%, p_1=7.24%, p_2=70.42%, p_3=20.58%, p_4=1.54%, p_5=0.003%


exemple : how to calculate F(17.6) ? 17.6 is an asuran fists with every hit connecting and 13 store tp. If you TP with 19 store tp, tp=5.8 in the above explanation.

you need to know

F(23.4), F(29.2), F(35.0), F(40.8), F(46.6), F(52.4), F(58.2), F(64.0), F(69.8), F(75.6), F(81.4), F(87.2), F(93.0), F(98.8).

Fist calculate F(98.8) :

F(98.8)=p_0 *( F(98.8) +1 ) + p_1 * ( F(104.6) + 1 ) + p_2 * ( F(110.4) + 1 ) ... + p_5*( F(127.8) +1)

Of course F(104.6)=0, F(110.4)=0 ... F(127.8)=0 because if you already have more than 100 TP you need no more rounds !

*This leaves simply F(98.8)=1/(1-p_0)= 1.00189 rounds.
Now F(93.0)=p_0 *( F(93.0) +1 ) + p_1 * ( F(98.8) + 1 ) + p_2 * ( 0 + 1 ) ... + p_5*( 0 +1)
*which leaves F(93.0)=(1+p_1*F(98.8))/(1-p_0)=1.0745 rounds
and so on ...
F(87.2)=1.786
F(81.4)=2.096
F(75.6)=2.65 ...
F(17.6)=7.24

You need 7.24 rounds on average after a full asuran fists. Of course you have to calculate this for every possible TP return from asuran fists, and then ponderate by the probaility of this to happen. The final value is 7.290 rounds on average.

You can easily input this computations in a spreadsheet, and adapt it to OAT, footwork by ajusting the values of p_0, p_1, etc.. (up to p_2 with footwork, p_7 with faith).

form pchan

[CDF]

trying to decide which is best sub katana now, assuming mozu(kannagi) main

mozu/oat (vnm)
/oa2-3
/dmg+23 double attack +7

OA2-3 looks like the most fun, but the math might say differently. any thouhts?

Enchu "occasionally attacks 2-3 times" is not a good offhand katana (in its current form anyway, allowing for the slim possibility SE cares). The attack distribution ("50:30:20" for 1:2:3 attacks) doesn't offset the low base damage. Enchu "occasionally attacks twice" can be a viable offhand katana, but the Enchu DA +7% should be better for that purpose given /WAR DA and Brutal. Actually Enchu DA +7% should be the main hand in a Mozu pairing.

Trying to figure out Ninja weapons at 80...

I am wondering what people are thinking is the best Main/Off setup at 80.

I am working on the DMG weapon but not being in having an LS that is routinely killing VNM at a time I am available is hampering that process.

The STP katana is almost done, but is it worth it? Basically with my setup its going to turn 4.7 tp a hit to 5 tp a hit. Taking away 2 hits from the build is nice, but im not sure if that brings it above other katanas.

I am at the point in the Elemental to choose between STR/Attack or Dex/Accuracy. Which is better? Is it worth it even to use either one over other available weapons? ie, if they arent better than the others then i will make the Agi/Evasion katana instead.

Right now Im using Sen/Fudo (When I came back I didnt have the IN for Perdu and now that I am do... Im wondering if its worth buying still over one of more of these katanas.)

"On paper," Kannagi/Enchu DA +7% is the best given 95% hit rate (even better than Kannagi/Tobi, this under the supposition that DA +7% is not weapon-specific (you would hope); if accuracy isn't capped and the full benefit of accuracy +15 is realized, Kannagi/Sekka acc/DEX is. Uzura is a good offhand, but Enchu DA +7% and Enchu OAT can be shown to be better options. Given "spamming" use of Blade: Jin, increasing TP/hit from 4.7 to 5.0 decreases the mode from 19 hits (to 100 TP) to 18, not 19 to 17.

Perdu may not be worth getting now if you will actually get two post-75 katanas as long as one of them is not OA2-3T.

I am basing all of the claims above on my own 'rithmetic if you can suffer looking at a cryptic table: http://img828.imageshack.us/img828/1895/katana.png

The above is based on NIN/WAR. Things could be different with NIN/DNC, but that's not about "optimal" damage-dealing.