Katana Combos 70+

[Zempten]

Hello everyone

I was curious as to a few Katana Combos and wondered if you could answer some of my questions. The following Katana Combinations are what I have in mind

A) Unji &Unsho

B) Fudo & Rai Kunimitsu

C) Fudo & *anything else 70+*

Type A combination for the type of player that likes to hit fast.
Type B combination for the type of player that likes to hit hard
Are these correct assumptions?

What other combinations are there? Which do you guys use and why?

Thanks ^ ^ Appreciate all responses.

[LyonheartLakshmi]

There's also Senjuinrikio (http://wiki.ffxiclopedia.org/Senjuinrikio). Break the latent on that sucker, and you could probably do Senju/Unsho (low delay offhand katana to try and take max advantage on the Senju).

[little ninja]

Unji & Senju

[Zempten]

Why Senju and Unji? Wouldn't Senji & Rai give like 11% critical hit and have the highest base dmg katanas?

[Omni]

what about good ol fudo + senju? crit+ wins. even though that 6% only applies to the hand in which it's in, nin has very good crit %.

and lower delay weapons arent as crazy as what they used to be with the removal of the tp floor.

[Zempten]

So U-U is out now?

[little ninja]

Hmm not sure why i went Senji/unji. It has been so long since i started using that combo. I know when i play with my haste/D.W gear i will go U.U.

[Aeolus]

Would'nt Senji/Unji be better as it has a lower delay overall for more crits/tp and the Str+2 and Dex+4 will help the chances of your Blade:Jins doing alot more dmg? Especially if youve already got a Str/Dex ws set up.

[Hankthetank]

I personally, use Senju/Unji. I love my Senju and use a low delay offhand weapon to get maximize the attacking innings on my Senju. The bonus stats are nice for WSs, but I wouldn’t consider them the main selling point. WS damage wise I didn’t notice much of a difference when I switched from Senju/Fudo to Senju/Unji.

I have not played with Unsho/Unji yet. Unsho costs 2.5Mil currently which is more than I am willing to spend on a test weapon. There is a noticeable difference in damage between my Senju and Unji and I am not sure I want to lower my hit for hit damage by that much, especially on higher defense mobs like Mamools or Trolls. In Mire the difference isn’t as drastic and I can see U/U pulling ahead, but truth be told I hate meriting in mire.

I played with a Rai and didn’t care for it. Its only a 2% increase in hit rate for the offhand, at the cost of significantly lower base damage without the benefit of a low delay. I wouldn’t recommend it.

In the end, I am of the opinion that between Fudo, Unji, Unsho, and Senju it doesn’t matter too much. Get a Senju since it’s free, then pick any katana to offhand. Fudo is perfectly acceptable until you get a Senju. While I feel Unji is an improvement for the offhand, but at the cost of 1mil on my server there are better things to spend you money on that will make a more noticeable improvement. Fudo gets and unfairly bad rap, because its so cheap all bad Nins tend to use it.

Bad nins use fudo, but not all fudo users are bad nins.

Sincerely,
Hankthetank

[little ninja]

I was lucky i got mine from the BCNM. At the time it was a bunch of us going to get it for the ninja's. I was lucky to use my seals when it dropped so i got it automatically. It's already dropping in prices now going for 900k. So its bound to go allot lower.

As for the other question. I think senji is 227 and unji is 201. thats 428 total delay. While U.U is 402. Now ad in Ninja AF +1, Ninja AF2 pants, and sepp earring. Now add haste on my head, hands, feet, waist and sometimes legs.

As for people who talk about crits. Senji has +6 if im not mistaken. My crit merits are capped, an i have around 20+25 dex. So im pushing as much crits as i can.

[Zempten]

I thought low delay weapons are rather useless since the TP floor removal thingy. So why is U-U still a popular choice?

[Hankthetank]

Just for fun. I’ve never calculated the different swing speeds of the different Katana sets.

60 Delay = 1 Second

Senju/Fudo Delay = 454
Senju/Unji Delay = 428
Unji/Unsho Delay = 402

Using Hauby:

Dual Wield IV + Suppa = 30% Delay Reduction
Senju/Fudo Delay = 454 * .70 = 317.80
Senju/Unji Delay = 428 * .70 = 299.60
Unji/Unsho Delay = 402 * .70 = 281.40

20% in Haste Gear + 15% Haste from Spell
Senju/Fudo Delay = 454 * .70 = 317.80 * .65 = 206.57
Senju/Unji Delay = 428 * .70 = 299.60 * .65 = 194.74
Unji/Unsho Delay = 402 * .70 = 281.40 * .65 = 186.55

Adding AF body instead of Hauby:

Dual Wield IV + Suppa + AF Body = 35% Delay Reduction
Senju/Fudo Delay = 454 * .65 = 295.10
Senju/Unji Delay = 428 * .65 = 278.20
Unji/Unsho Delay = 402 * .65 = 261.30

20% in Haste Gear + 15% Haste from Spell
Senju/Fudo Delay = 454 * .65 = 295.10 * .65 = 191.81
Senju/Unji Delay = 428 * .65 = 278.20 * .65 = 180.83
Unji/Unsho Delay = 402 * .65 = 261.30 * .65 = 169.84

Now lets say we swap out Byakko’s for Koga for a Max Dual Wield Build.

Dual Wield IV + Suppa + AF Body + Koga = 40% Delay Reduction
Senju/Fudo Delay = 454 * .60 = 272.40
Senju/Unji Delay = 428 * .60 = 255.00
Unji/Unsho Delay = 402 * .60 = 241.20

15% in Haste Gear + 15% Haste from Spell
Senju/Fudo Delay = 454 * .60 = 272.40 * .70 = 190.68
Senju/Unji Delay = 428 * .60 = 255.00 * .70 = 178.50
Unji/Unsho Delay = 402 * .60 = 241.20 * .70 = 168.84

Well there are some numbers for anyone who is interested in the delay differences. I think I did that right.

I thought low delay weapons are rather useless since the TP floor removal thingy. So why is U-U still a popular choice?

1. It's still offers increased TP gain, just not nearly as fast as it used to.


While the TP advantage isn't as pronounced as it used to be, U/U still makes for faster hits, and TP gain/delay is skewed towards the ends, with very low delay and very high delay weapons having a TP gaining advantage.


2. Swinging faster is often seen as better than hitting harder.


Comparing S/U to U/U, the delay differance is 180.83 Vs 169.84 or a differance of only about ~11 delay. While it doesn't seem like much, thats effectivly 5% haste in terms of swing speed.


In a camp like mire where the mobs are already very squishy, the differance between a Senju hit and a Unji hit might not be that big. So hitting faster might be more beneficial. At other camps, the differance might be more noticable and using stronger Katanas might be the way to go.


Sincerely,
Hankthetank

[Zempten]

Hm....from those calculations. It seems like 5% in Haste is greater then 5% in Dual Wield then? Since Haste also affects spell timers and the difference between them is minimie. So If I had an option of Byakko Haidaite vs Koga. I would get Haidaite 1st?

[Hankthetank]

Yeah, for the most part Haste is better due to the extra benefits of haste and the fact it doesn’t lower TP gain like Dual Wield does.

However Koga does offer a slightly faster swing rate so from a raw DD stand point, it is technically better, if you ignore the 15 Dex and the Acc Byakko provides.

It is also fairly important to note that it works out to nearly the same in situations where you can get about equal amounts of Dual Wield or Haste. When you can get significantly more of one it generally pays to keep stacking that.

Lets look at a soloing situation where you don’t get haste from a mage. Lets also say you have only the Turban and are deciding between Byakko and Koga:

Dual Wield IV + Suppa + AF Body = 35% Delay Reduction
Senju/Fudo Delay = 454 * .65 = 295.10
Senju/Unji Delay = 428 * .65 = 278.20
Unji/Unsho Delay = 402 * .65 = 261.30

10% in Haste Gear (Turban + Byakko)
Senju/Fudo Delay = 454 * .65 = 295.10 * .90 = 265.59
Senju/Unji Delay = 428 * .65 = 278.20 * .90 = 250.38
Unji/Unsho Delay = 402 * .65 = 261.30 * .90 = 235.17

Now lets say we swap out Byakko’s for Koga for a Max Dual Wield Build.

Dual Wield IV + Suppa + AF Body + Koga = 40% Delay Reduction
Senju/Fudo Delay = 454 * .60 = 272.40
Senju/Unji Delay = 428 * .60 = 255.00
Unji/Unsho Delay = 402 * .60 = 241.20

5% in Haste Gear (Just Turban)
Senju/Fudo Delay = 454 * .60 = 272.40 * .95 = 258.78
Senju/Unji Delay = 428 * .60 = 255.00 * .95 = 242.25
Unji/Unsho Delay = 402 * .60 = 241.20 * .95 = 229.14

The difference is far more pronounced here, because of the lack of haste options. If you remove the turban and compare just the Byakko Vs Koga you get the following:

Dual Wield IV + Suppa + AF Body = 35% Delay Reduction
Senju/Fudo Delay = 454 * .65 = 295.10
Senju/Unji Delay = 428 * .65 = 278.20
Unji/Unsho Delay = 402 * .65 = 261.30

5% in Haste Gear (Byakko)
Senju/Fudo Delay = 454 * .65 = 295.10 * .95 = 280.34
Senju/Unji Delay = 428 * .65 = 278.20 * .95 = 264.29
Unji/Unsho Delay = 402 * .65 = 261.30 * .95 = 248.23

Now lets say we swap out Byakko’s for Koga for a Max Dual Wield Build.

Dual Wield IV + Suppa + AF Body + Koga = 40% Delay Reduction
Senju/Fudo Delay = 454 * .60 = 272.40
Senju/Unji Delay = 428 * .60 = 255.00
Unji/Unsho Delay = 402 * .60 = 241.20

0% in Haste Gear
Senju/Fudo Delay = 454 * .60 = 272.40 * 1 = 272.40
Senju/Unji Delay = 428 * .60 = 255.00 * 1 = 255.00
Unji/Unsho Delay = 402 * .60 = 241.20 * 1 = 241.20

As a general rule Haste is better than Dual Wield enhancement due to ancillary benefits. But from a pure how fast can I swing standpoint, Dual Wield is generally better especially if you have access to very little haste gear.

Its also important to note if you are not 75 yet, Koga is a 72 piece of gear and Byakko is 75. So for people leveling Koga might be the better choice since they can use Koga while they level and eventually get Byakko at 75. But if you are 75, I assume you have access to a number of haste pieces and in this case Byakko will serve you much better.

Sincerely,
Hankthetank

[Amele]

something I've always been curious about (and wondering if you know, hank)

I think it's pretty clear that delay reduction is applied as:

(delay * (1.0 - reduction) / 60

but what I've never been sure about is where haste fits. (this is a subtle question)
the first possibility is that haste is a reduction, like delay reduction.

so the equation would be: (delay * (1.0 - haste) / 60

the other possibility is that haste increases the rate that 'frames' are counted.

so the formula becomes, delay / (60 * (1.0 + haste) )

this is a subtle difference because these two formulas are not equivalent (example: 240 delay weapon and 15% haste - 240 * .85 / 60 = 3.4 :: 240 / (60 * 1.15) = 3.47~ )

my suspicion is that it is the first formula, since that allows the delay timer to be an RTC rather than software timed (making it 'immune' to lag) but I'm not certain, since the latter is probably the 'more accurate' way to depict an increase in speed.

[Hankthetank]

Its great question and I really don't know. I have always seen both dual weild and haste calculated as reductions.

Hmm, I think I got an idea.

[Hankthetank]

/derail

Ok making a few assumptions I think I got a test for you.

Assumptions:

Haste Spell is 15% Haste
The Calculations for haste’s effect on melee swings is the same for Spell recast. Raise III is a 60 second recast or 3600 Delay

1st method
15% haste (Haste Spell)
3600 * .85 = 3060
51 second recast or 3060 Delay

2nd Method
60 second recast 0r 3600 delay
3600 / ((60 *(1+.15)) = 52.17 Second recast

So in theory, if a WHM casts haste on himself then casts Raise III if the recast is 51 seconds it means the game calculates using the first method of (delay * (1.0 - haste) / 60

If the recast is 52 seconds its is calculated using the second method of delay / (60 * (1.0 + haste) )

My money is on the first method, but who knows. I think this sounds like a viable test to get your answer. At work can't test it myself, but if anyone else does I'd love to hear the result.

/derail off

Sincerely,
Hankthetank

[Amele]

it's a good test.

I'll run it tonight after work and report my findings. although it is a big assumption that recast times are handled by the delay calculations and not a time_millis() comparison. - (it's a fair assumption).

the other tests all involve estimating averages from swinging a whole bunch at walls in excavation duty (hey! there's another idea)

see if in the time it would take you to get 100 swings, you get 117 or 115 swings (if you get 117 swings, then it's the first formula, if you get 115 it is the second formula)

use a lower delay weapon for the test, go on rdm/blm or something else with no delay reduction or innate double attack if you can, and report back! (with a 180 delay dagger, it should take approximately 5 minutes per 100 swings) you should have someone else along to recast haste on you at about the 2.5 minute mark so it doesn't wear off mid trial.

I'll try to run a similar test on sam/thf soon too with a different haste value (10 and 25%) to see if the numbers line up as well.

[little ninja]

Dusk gloves = +3 haste. Sun-ate = +3 haste. turban = +5 haste, speed belt = +6 haste. Thats +17haste. Now if i get a haste spell cast on me. Thats +32haste which puts me over the cap. Then i add in my Sup earring, my koga pants, Ninja af body, and D.W from being a 75 ninja. I love to wear that stuff when i fart around, skill up, or farm. MY ninja just stands there and swings away.

Also what are you guys refering to when you talk about the "TP FLOOR" Its been a long time since i tp burned.

[LyonheartLakshmi]

Also what are you guys refering to when you talk about the "TP FLOOR" Its been a long time since i tp burned.
Back before an update that happened a year ago, there was a minimum amount of TP you would get back from a normal melee swing. If you look at wiki's info on TP calculations (http://wiki.ffxiclopedia.org/Tactical_Points), you'll see a blurb down at the bottom about TP gain before that update. Normally, TP will scale down so that the faster the weapon, the less TP you get. But once you hit 180 delay, you got 5.0 TP per swing, no matter how far below 180 delay that you reached. Hence, 5.0 TP was the "floor".

Today, you'll get less than 5.0 TP per swing when your delay is below 180.

[Srxjo]

Hankthetank please use the set font it was a pain in the ass to read what you wrote

[Hankthetank]

Dusk gloves = +3 haste. Sun-ate = +3 haste. turban = +5 haste, speed belt = +6 haste. Thats +17haste. Now if i get a haste spell cast on me. Thats +32haste which puts me over the cap. Then i add in my Sup earring, my koga pants, Ninja af body, and D.W from being a 75 ninja. I love to wear that stuff when i fart around, skill up, or farm. MY ninja just stands there and swings away.

1. The haste cap for gear is 25%. Not including DW enhancements.
2. The haste cap for Spells recast is 50%
3. Hard to prove, and I haven’t seen any definitive numbers on it, but the haste cap for melee swings is said to be around 70%.

I can personally vouch for 1 and 2.

1 can be demonstrated by using a Shinobi earring, which gives 20% haste when in Red. This is almost a null point however, since other than BB monks most jobs will have an incredibly hard time getting 25% haste in gear.
2 can be proven with the fact I cannot get my Flash or Utsu timers to drop below, 22 seconds dispite wearing full haste gear, including haste belt, while getting Haste Spell and dual marches. Which is well over 50% haste.
3 I believe because I can confidently say that 20% in haste gear, haste (spell) and dual marches is an improvement over 20% in haste gear, haste (spell) and a single march. I can’t put and exact number on it, but I know it’s over 50%.

Also what are you guys refering to when you talk about the "TP FLOOR" Its been a long time since i tp burned.

LyonheartLakihmi said it perfectly, it was the minimum TP you get per successful swing.

The TP floor was also the main reason U/U was so powerful. U/U was so far below the floor yet by still getting 5TP a hit it was able to WS and get back to 100TP in about 25 seconds (less if you got a string of DA procs) pre TP patch. U/U users would spam Jin, which more than made up for the low base damage of the katanas. U/U is still a fast hitting combination, but now that the TP gain from them has been corrected, they seem lackluster when compared to their former glory.

Sincerely,
Hankthetank

[LyonheartLakshmi]

I thought low delay weapons are rather useless since the TP floor removal thingy. So why is U-U still a popular choice?
Just to put things into perspective, the following is the TP you would get per swing for the following delay values:
120: 4.5 TP
180: 5.0 TP
450: 11.5 TP
480: 13.0 TP
530: 14.5 TP

Since 60 delay equals one second, we can calculate TP-per-second:
120: 2.25 TP/sec
180: 1.67 TP/sec
450: 1.53 TP/sec
480: 1.63 TP/sec
530: 1.64 TP/sec

As you can see, TP gain is much faster at 120 average delay than at the other delay values listed. Past 180 delay, the lower you can push your delay, the faster you can gain TP. (http://wiki.ffxiclopedia.org/Image:Tpgain.jpg) Actually, this is true once you go below 450 delay. But between 450 and 180 delay, the change is gradual. Once you hit 180, TP gain changes very steeply.

TP burn efficiency is mostly about getting TP really fast without compromising your DoT. This is why Joyeuse and Ridill are such great offhand weapons for WARs while Kraken Club is viewed as more of a toy for WARs.

U/U manages to accomplish this quite well. Both weapons have decent DPS and the average delay is 121 with Dual Wield IV + Ninja Chainmail + Suppa.

Rai Kunimitsu ends up being perhaps choice #4 behind Senju, Unji and Unsho from that perspective. It's not exceptionally fast, and it actually has lower DPS than either Unji or Unsho, even if you account for the Firesday damage bonus raising base damage 1/8th of the time.

Hm....from those calculations. It seems like 5% in Haste is greater then 5% in Dual Wield then?
That's probably true just from the fact that increasing dual wield speed decreases TP per hit, while increasing haste doesn't. While increasing dual wield increases the speed of your TP gain, haste increases it even moreso.

[Amele]

1 can be demonstrated by using a Shinobi earring, which gives 20% haste when in Red. This is almost a null point however, since other than BB monks most jobs will have an incredibly hard time getting 25% haste in gear.

yeah, most DD will use a haste belt (4 or 6%) b.haidate if they can wear (5%) walahra turban 5%. several jobs can also wear dusk gear, BRD/BST/NIN/RNG/SAM/WAR , who will gain a further 5% with hands and feet.

this is about 19-21% haste - you can hit a similar number with blu/drk/drg/pld/thf by using homam in addition to w.turban and a haste belt,

so 21% is 'typical' max a job can reach without resorting to multiple +1's, which will only get you to 23% at the outside.

mnk gets 25% with BB, sam/drg gets 25% with wyvern earring, and (amusingly) whm gets to cap with the right mix of hq blessed and ex gear. brd gets 25% by stacking a manteel ontop of the above 23%.

I don't *think* any other jobs can currently hit 25% but I wouldn't bet much on it.

[little ninja]

So Is the TP floor figured in before or after the reduction of the weapons delay due to Haste, and D.W gear ?

Since it was corrected in 06. I was using U.U just before the tp floor was added. Nowadays i use it to mainly farm an goof around outside white gate. Its been so long, i probably don't even remember much of the original TP gains.

I almost never use Ninja anymore for LS events or meriting. But every once in awhile i will bring it out of retirement to shut up the noob nin's in my shell.

[Hankthetank]

TP gain is calculated after Dual Weild reduction is, so dual weild does lower TPO per a hit. Haste however doesn't effect TP gain.

Sincerely,
Hankthetank

[little ninja]

TP gain is calculated after Dual Weild reduction is, so dual weild does lower TPO per a hit. Haste however doesn't effect TP gain.
Sincerely,
Hankthetank

Ahh so another gimp for ninja i have overlooked all these months.

[LyonheartLakshmi]

Ahh so another gimp for ninja i have overlooked all these months.
Don't feel bad. This was one of the more subtle changes made during a particular update. In the update notes, it was simply listed as, "The amount of TP accumulated when attacking a monster has changed for both the player and the monster." Either you read that and got a burst of curiosity to go and figure out exactly what changes, or you read that and went "huh?" and moved on with your life. It was easily overlooked since the claim mechanics change was the big controversy in that update. I only found out about the TP floor removal because of some random thread either in my server's forum or the WAR forum.

If it's any consolation, THF got a much bigger whack from the nerf stick out of this change than NIN. THF really needed to spam WSs faster to compete. Thankfully, SE helped restore some power to THF by increasing base damage on many daggers in a subsequent update 3 months later.

[Amele]

little ninja, just to clear some stuff up since I'm not sure we're all entirely on the same page.

in the beginning, all weapons below 180 delay gave 5 tp / hit (the same as 180 delay weapons) this is the 'TP' floor.

ninja dual-wield reduction has always reduced the tp return of weapons dual wielded but once the DW-reduction was high enough to take a weapon -below- the 180 delay point, additional dw-reduction had no further effect on the tp / hit, only making the weapons swing faster.

as part of the trend to reduce the ascendancy of x/nin, and rebalance the 2h weapons tp gain, the tp gain rates were adjusted in 2006, among some of the adjustments was the removal of the TP floor.

gain rates are still 'better' under 180 than at 180, but it's not as obvious as it once was. so basically, nothing got gimped, SE just fixed the tp floor advantage that certain jobs had over dark, dragoon, samurai, and great axe warrior.