Katana Comparison -- Unji/Unsho vs Fudo/Senjuinrikio

[decidence]

DWIV is 30%, it follows the pattern of 10% - 5% - 10% - 5%.

Haste is 15% as proven about 2hrs ago when I completely stripped my gear and had Haste cast on me. Cast Utsusemi: Ni and had 38s recast which when you do the math 38/45 is 85%... or 15% reduction.

There are many holes in that source your quoting, I don't think the calculations that guy is making is true, nor are his formula's. Judging by what he knows about DW and Haste in the very basic of instances.

It is very possible to get nearly 0s delay (I said nearly).

Recast Delay Haste caps at 50%
Melee Delay Haste caps at 90%

This was proven a long time ago when Hundred Fists used to be programmed differently. It used to just simply increase Haste to the maximum of 90% which gave instant 50% recast etc which was since patched. 90% was found using Hundred Fists when tested with a high delay Staff and using the number of attacks made with the delay of the weapon.

With the highest level of DW (45%)
The highest Haste possible (using Soul Voice March x2, Speed Belt, Dusk Gloves +1, Panther +1, Byakkos, Haste Spell, Wyvern Earring = over 90%)

Unsho/Unji would be...

402 * 0.55 = 221.1
221.1 * 0.1 = 22.1
22 / 60 = 0.36s

/me creams pants

 

[Bajaratt]

1. DW is calculated first.
2. Haste is 15% not 20%
3. March2 is 12% and March1 is 8% (but together they equal 21% Im guessing because of fractions of a percent).

I had always heard that haste was 10% and I had heard the March songs totaled up to haste(10%), but with Advancing March(March 1) only being 2-3% with Victory March(March 2) making up the other 7-8%. Glad to learn I was wrong and that Advancing March isn't a complete waste of time to sing anymore.

 

[Trippyhippy]

I had always heard that haste was 10% and I had heard the March songs totaled up to haste(10%), but with Advancing March(March 1) only being 2-3% with Victory March(March 2) making up the other 7-8%. Glad to learn I was wrong and that Advancing March isn't a complete waste of time to sing anymore.

Haste % is based on the casters enhancing magic isn't it?

 

[Kire]

Haste % is based on the casters enhancing magic isn't it?

no do not think so i think the % is set but how long it lasts is on the casters skill in that lv

 

[Tagrineth]

Sorry to derail for a moment here but

What the fuck? A completely readable post by Blazn?

ALERT THE INTERNET

 

[Bajaratt]

no do not think so i think the % is set but how long it lasts is on the casters skill in that lv

No, Haste is always 3 minutes, regardless of your enhancing magic skill.

 

[Stouter]

Yes, I see Unji going up in price because of findings like this.

Unji was 2M with none up at AH last night, glad I bought mine for 800k a couple months ago.:wink:


On the haste % discussion, with a March +2 instrument using the spell drain as a brd/blm:
Recast with no march - 60 second
Recast with March1 - 54 seconds
Recast with March2 - 53 second
Recast with March1+2 - 48 seconds
Recast with SV-March1 - 48 seconds
Recast with SV-March2 - 47 seconds
Recast with SV-March1+2 - 36 seconds

I'm starting to doubt that Rones's way of figuring haste is correct.

 

[Stouter]

It will take about 18 seconds for Unsho/Unji to hit 100% TP @ an average of 10% TP per attack.
It will take about 19.38 seconds for Fudo/Senjuinrikio to hit 100% TP @ an average of 10.5% TP per attack (10.5 because sometimes its 11 sometimes its 10... seem's random).

After 50 no Katana combination (besides 999 delay relic) will give you more than 5.0 TP per hit unless you have some Store TP on. DWIII is that powerful.

 

[decidence]

Yea I found that out already too hehe... it was Store TP that was being confusing and giving 11 TP per attack sometimes rather then 10 TP.

 

[Stouter]

Side note on Store TP, it's rounded down to the nearest tenth, so Store TP +5 takes you from 5.0 to 5.2 (rounded down from 5.25) but if you toss on a brutal earring you're up to 5.3 per hit.

 

[Claxton]

Someone get teh bow with +6%. Should be easy to find out if crit% stacks or not.

 

[Bajaratt]

Someone get teh bow with +6%. Should be easy to find out if crit% stacks or not.

That was already determined earlier on in the thread and no, they do not stack. The +crit % applies only to damage from that specific weapon, not to all attacks performed.

 

[Rones]

There are many holes in that source your quoting, I don't think the calculations that guy is making is true, nor are his formula's. Judging by what he knows about DW and Haste in the very basic of instances.

The formules were provided by Applepie from ffxionline.com. He got them out of the JP game guides that have all the hidden formuleas and info. Unlike our Brady game guide, they get game guides with inside information.

It is very possible to get nearly 0s delay (I said nearly).

Recast Delay Haste caps at 50%
Melee Delay Haste caps at 90%

This was proven a long time ago when Hundred Fists used to be programmed differently. It used to just simply increase Haste to the maximum of 90% which gave instant 50% recast etc which was since patched. 90% was found using Hundred Fists when tested with a high delay Staff and using the number of attacks made with the delay of the weapon.

This was mentioned in the link, but hundred fists breaks all the rules. Outside hundred fists, best you will get with 100% haste is half delay. This is also why haste has deminishing returns like everything in this game, unlike your formulea which doesn't.

I'm starting to doubt that Rones's way of figuring haste is correct.

How so? Your example didn't even deal with my way of figuring. My method is calculating melee delay, not recast delay. Additionaly, your tests show deminshing returns for larger amounts of haste. Under Decidence's way, that doesn't happen. I can point it out in your data if you want, but I'm sure you are capable of that yourself.

 

[decidence]

Adding Senjuinrikio/Unji since thats a popular setup for awhile.

SENJUINRIKIO+UNJI
STR: 111
DEX: 111
ATK: 424
ATK w/Berserk: 530

SENJUINRIKIO+UNJI
WSC: 54.78
fSTR: 12
PDIF: 1.00
PDIF w/Berserk: 1.26
fTP: 1.0

SENJUINRIKIO+UNJI Formula's
(( 38 + 12 + 54.78 ) * 1.0) * 1.00 = 104.78 * 1.00 = 104.12
(( 38 + 12 + 54.78 ) * 1.0) * 1.26 = 104.78 * 1.26 = 132.02

SENJUINRIKIO+UNJI
With No Haste Equip / With 22% Haste Equip + DWIV + Suppa / Previous + Haste Spell
7.13s / 3.61s / 2.92s

Although these numbers seem way better, so many things come in to play with Blade:Jin that in the end you just need to do what parses best in each situation depending on the difficulty of the mob. What does this mean? Lots of testing, lots of gilspending.

 

[decidence]

What does this mean in a normal PT for TP Build?

It will take about 18 seconds for Unsho/Unji to hit 100% TP @ an average of 10% TP per attack.
It will take about 19.38 seconds for Fudo/Senjuinrikio to hit 100% TP @ an average of 10.5% TP per attack (10.5 because sometimes its 11 sometimes its 10... seem's random).

Which means it will take approximately 18-20 WS's used before Unsho/Unji is only 1 WS ahead of Fudo/Senj which if your keeping track is a long time considering the higher damage output of Fudo/Senj.

Just thought I would mention that because of the TP Nerf this shit has changed. Also there was some problems with the original calculations. I was calculating DW and Haste combined instead of seperately which is indeed how it is supposed to be done, so some numbers are messed up. So I edited and fixed the original delays in seconds with the actual delays.

As for how long it will take to get 100TP with the 3 different combinations. Assuming all the Haste and shit is applied in a normal Merit PT.

Fudo/Senj @ ~4.9 TP per swing = 20.40 swings or 10.2 attack volleys
Senj/Unji @ ~4.7 TP per swing = 21.27 swings or 10.5 attack volleys
Unji/Unsho @ ~4.5 TP per swing = 22.22 swings or 11.1 attack volleys

Now how long does it take to do 20-23 swings is the hard part to find out. Because now I have to calculate the fact that each swing has a 15% (1 in every 6.67 swings) chance to be two swings which would mean less attacks overall to get to 100TP.

20.40 Swings = 3.05 DA's
21.27 Swings = 3.19 DA's
22.22 Swings = 3.33 DA's

So overall we are looking at the following numbers which have all be rounded up to the nearest swing because theres no such thing at 18.08 swings.

Fudo/Senj: 18 swings or 9 attacks to get 100TP
Senj/Unji: 19 swings or 9.5 attacks to get 100TP
Unji/Unsho: 19 swings or 9.5 attacks to get 100TP

And the moment you have all been waiting for...

Post TP Nerf
Fudo/Senj @ 3.09s per attack = 27.81s for 100TP
Senj/Unji @ 2.92s per attack = 27.74s for 100TP
Unji/Unsho @ 2.74s per attack = 26.03s for 100TP

So did the TP Nerf hurt U/U quite a bit, I guess I will just have to parse it and see for myself, but just by looking at the numbers (if I have the TP per swing correct which I'm not sure I do) then it's not much of an advantage as it used to be for sure as you can see below.

Pre TP Nerf:
Fudo/Senj @ 3.09s per attack = 26.27s for 100TP
Senj/Unji @ 2.92s per attack = 24.82s for 100TP
Unji/Unsho @ 2.74s per attack = 23.29s for 100TP

Although now the gap between U/U and Senj/Unji is slightly bigger then it used to be the gap between 201 delay weapons and 227 delay weapons has shortened significantly.

 

[Stouter]

How so? Your example didn't even deal with my way of figuring. My method is calculating melee delay, not recast delay. Additionaly, your tests show deminshing returns for larger amounts of haste. Under Decidence's way, that doesn't happen. I can point it out in your data if you want, but I'm sure you are capable of that yourself.

This isn't diminishing returns, it's decimal dropping. I tested further than simple recast, here you go:

First test, spell recast. Brd/blm, measuring recast of drain and aspir, going via menu, using March +2 instrument.
No March or Haste - 60 seconds.
Victory March - 53 seconds (roughly a 12% decrease in recast, or roughly 13% "MMO Haste")
Advancing March - 54 seconds (roughly a 10% decrease in recast, or roughly 11% "MMO Haste")
Haste Spell - 51 seconds (roughly a 15% decrease in recast, or roughly 17% "MMO Haste")
SV Victory March - 47 seconds (This puts Victory March at closer to an 11% decrease in recast)
SV Advancing March - 48 seconds (This keeps Advancing March at 10%)
This means you will either get a 36% decrease in recast with all 3, or 41% "MMO Haste"
Recast with double March and Haste - 39 seconds, this is about a 35% decrease in recast, or about 53% "MMO Haste"

Test two, attack delay. For this test I used the /clock function. I hit /clock as soon as a strike appeared in my text box, and then hit it again as soon as the 20th hit after that one by myself appeared in my text box. Again using a March+2 instrument, and an Earth Staff (366 Delay) Three tests each setup
No Haste - 123, 123, 123 seconds
Victory March - 110 seconds (this is about a 10.5% decrease in delay)
Advancing March - 112 seconds (This is about a 9% decrease in delay)
Double March & Haste - 81 seconds (This is about a 34% decrease in delay)
Double March & Haste lowered attack delay by roughly 34%.

Any further questions?

 

[Rones]

Any further questions?

Why did you wait over two months to reply?

What is this "MMO Haste" and how are you calculating it?

No Haste - 123, 123, 123 seconds

This is off by 1 second. 366delay/60 * 20 attacks = 122 seconds. Your other calculations would be off slightly if you used 123.

Also curious on your thoughts as to why melee haste percent differed from recast percent change. Earlier Decidence was using the recast haste percent to state the melee haste percent as the same. Additionaly, if the two differ, then it might be a mistake to assume haste is a 15% melee haste just because it affects spells by 15% since neither of the marches follow that. You skipped that part in your melee testing.

 

[Stouter]

I didn't see your post 2 months ago, and deci's bump of this is when I caught it. "MMO Haste" is the method you use to calculate haste, Delay / (1 + Haste%) where 100% Haste would equal a 50% drop in delay.

I'll admit my testing isn't exact, but it's close. You're not gonna get exact with the stopwatch method unless you take a larger string of attacks, which isn't all that easy when you need to stop to recast marches on yourself. Getting 33% instead of 35% is understandable with my methods, but getting 33% when it should be 50% isn't.

 

[Rones]

I didn't see your post 2 months ago, and deci's bump of this is when I caught it. "MMO Haste" is the method you use to calculate haste, Delay / (1 + Haste%) where 100% Haste would equal a 50% drop in delay.

The method I linked was for melee delay only. It didn't have any equation for recasts.