Dual Wield Delay Reduction + Haste

[friedfunk]

Does the Dual Wield Delay reduction stack with Haste towards the haste cap ?

So a WAR/NIN with Suppa has 20% delay reduction, with Soul Voice Marches, Haste spell, and perfect Haste gear (total 68.75% Haste), does that make it 88.75% total delay reduction? Or capped at 80% ?

[Radec]

No. Dual wield reduces delay (and TP), haste increases attack speed, they are separate.

Effective_Delay = Base_Delay*(1-Dualwield reduction)*(1-haste)

With given conditions, Effective_Delay is 0.25*Base_Delay.

[Amele]

No. Dual wield reduces delay (and TP), haste increases attack speed, they are separate.

Effective_Delay = Base_Delay*(1-Dualwield reduction)*(1-haste)

With given conditions, Effective_Delay is 0.25*Base_Delay.

there is a global attack speed cap now - I suspect that the global attack speed cap does change the interaction of dual wield and haste at the upper end.

(that said, unless you're a zerging dark you probably won't run into it.)

[orson]

I don't think it's actually possible to test whether dual wield goes over the haste cap because of individual haste caps. It would make sense that it wouldn't though based on how it's -delay and I don't believe the sword strap can break through the 80% haste cap.

[Ruiner]

Slightly off-topic: Do Suppanomimi and Ninja Chainmail count towards the 25% Haste-from-gear cap? I was under the impression they didn't, and I just wanted to make sure.

[friedfunk]

No I believe the Ninja Chainmail and Suppa are Dual Wield bonuses so they are separate from the gear haste.

I guess the only job that can break thru the 80% is NIN...

Dual Wield IV for 30%, + 5% Suppa + 5% Ninja Chainmail for 40% Bonus delay reduction

Max Haste from gear + SV marches + Haste spell total 68.75%

So the multiplier is .60 * .3125 = .1875 which just passes the .20 cap.

Not sure if NIN can take a spot in a zerg, but it would be fun to see one with dual Mamushitos or Stun Knife +1's...

[Ruiner]

No I believe the Ninja Chainmail and Suppa are Dual Wield bonuses so they are separate from the gear haste.

I guess the only job that can break thru the 80% is NIN...

Dual Wield IV for 30%, + 5% Suppa + 5% Ninja Chainmail for 40% Bonus delay reduction

Max Haste from gear + SV marches + Haste spell total 68.75%

So the multiplier is .60 * .3125 = .1875 which just passes the .20 cap.

Not sure if NIN can take a spot in a zerg, but it would be fun to see one with dual Mamushitos or Stun Knife +1's...

Edit: Nevermind, just realized you're talking about present day, not pre-nerf. My mistake.

[Coolmints]

Slightly off-topic: Do Suppanomimi and Ninja Chainmail count towards the 25% Haste-from-gear cap? I was under the impression they didn't, and I just wanted to make sure.

Delay reduction from DW (suppa and chainmail) = 0.6
Delay reduction from Haste (15%), SV marches (40%~), haste gears with shinobi ring, and speed belt (25%) = 0.2

If suppa and chainmail don't count towards the 25% haste-from-gear cap, you could have
the weapon delay multiplier of 0.2*0.6 = 0.12

If DW delay reduction doesn't count toward haste cap with Senji/Fudo, you should be able to swing 132 times in a minute.

[Kirschy]

Delay reduction from DW (suppa and chainmail) = 0.6
Delay reduction from Haste (15%), SV marches (40%~), haste gears with shinobi ring, and speed belt (25%) = 0.2

If suppa and chainmail don't count towards the 25% haste-from-gear cap, you could have the weapon delay multiplier of 0.2*0.6 = 0.12

If DW delay reduction doesn't count toward haste cap with Senji/Fudo, you should be able to swing 132 times in a minute.

The problem with the above is that you're neglecting to account for magic haste cap. Haste + SV Marches is going to cap at out at 43.75% (7/16).

As it was stated above, only a Ninja could reach that kind of gain. DWIV(30%) + AF1 Body(5%) + AF2 Legs(5%) + Suppa(5%). Haste + SV Marches + 20% Earring + Speed Belt.

.3125 * .55 = .171875 (82.8% Haste)

I don't have a Suppa... so if I tried to test /w 999 Delay weapon, It'd come out to be around 81.25% Haste. I'm not even sure if that would show up in animations. I planned to test it sometime but never got around to it.

Since Dual Wield and Haste act similar on attack delay when compared to Fast Cast and Haste on recast. It would seem the Hard Cap of 50% on Recast Delay would be analogous to the ~80% hard cap on attack delay.

This kind of scenario is rare though.

[Maxxthepenguin]

Wait I always thought DW and Haste were similar effect but separated so that final delay of a 100 delay weapon with 30% DW reduction and 50% haste was 100*(1-.3)*(1-.5)=35, not 100*(1-.3-.5)=20. This being true, haste cap wouldn't effect NIN anyways.

[Kirschy]

Wait I always thought DW and Haste were similar effect but separated so that final delay of a 100 delay weapon with 30% DW reduction and 50% haste was 100*(1-.3)*(1-.5)=35, not 100*(1-.3-.5)=20

You're right, they're seperated. But you can use associative property(assuming there's no rounding) and multiply (1-.3) and (1-.5) first and get .35*100. .35 being equivalant to 65% haste.

[Maxxthepenguin]

Ok, I see your point now.

I'm still doubtful that there is such a cap, mostly because of the TP reduction.

If there is no cap, than the highest you can get (w/ senjin+perdu) seems to be 98.1 attacks per minute. If there is, it looks like 86.4 attacks per minute.

[Amele]

Ok, I see your point now.

I'm still doubtful that there is such a cap, mostly because of the TP reduction.

If there is no cap, than the highest you can get (w/ senjin+perdu) seems to be 98.1 attacks per minute. If there is, it looks like 86.4 attacks per minute.

It is highly probable that the cap is enforced on attack speed and not just 'total haste' we just need someone to go test it lol

[Coolmints]

The problem with the above is that you're neglecting to account for magic haste cap. Haste + SV Marches is going to cap at out at 43.75% (7/16).
....

Opps, I was wrong. Well, I blame it to wiki http://wiki.ffxiclopedia.org/wiki/Soul_Voice :/

I know you could use shinobi's earrings but you can't use -HP gears to force the latent unless you just turn unweaken. Only the shinobi ring I find easy to trigger with some cheap -HP gears.

Based on your haste research posts, we see
Haste attainable from magical sources is capped at 448 /1024 (43.75%)
Dusk Hands = 30/1024
Turban = 50/1024
Swift belt/Shinboi ring (assuming it is the same as Ace's helm) = 40/1024
Speed Belt = 61/1024
Haste Spell = 150/1024

Let say you are going as NIN75/WAR37 for a zerg fight,

A NQ/HQ NIN could have:
Turban (50)
Dusk gloves/+1 (30/40)
Swift belt/Speed belt (40/61)
Shinobi ring (40)
Fuma kyahan (30)
Byakko haidate (50)
magical haste cap (448 )

Add them up we have
NQ = 668/1024 = 0.6523
HQ = (256 being gear haste cap + 448 )/1024 = 0.6875

As for the DW gears, let say you have
Ninja chainmail (5%)
Suppanomimi (5%)
DW trait (30%)
= 0.3906 (assuming 400/1024 for truncation)

Pairing up with Senju/Fudo for purely dps or Mamushito+1/Mamushito+1 for stunning,
NQ = 454*(1-0.3906)*(1-0.6523) = 454*(0.2118) roughly between 95 to 97 = 74 to 75 swing in 1 min.
HQ = 454*(1-0.3906)*(1-0.6875) = 454*(0.1904) = 86 to 87 = 82 to 83 swing in a min.
If there is a hard delay reduction cap, then HQ = 454*(0.2) resulting 79 swings in a min


with /war and brutal earring with 95% hit rate from feint,
NQ = (74.5)(1.15)(0.95) = 81 swings
HQ = (82.5)(1.15)(0.95) = 90 swings or 86 swings with hard delay cap.
In conclusion, the difference for a hard delay cap is only 4 less swings or roughly 4-5% less dmg.


Feel free to correct me if I am wrong again.

[Kirschy]

The math looks alright. The numbers we use for dual wielding are very accurate, but possibly not exact. Even if the test were performed, it would be hard to say for sure if a 4 swing difference over a minute is due to our estimation of dual wield being slightly off or caused by the attack speed cap. With a 999 Delay weapon dualwielded /w another high delay weapon it might raise the accuracy a bit, but it seems not worth the time investment. Would be cool to know though.

I also think there might be complications or problems with how delay/attack animation is done. When I tested attack delay from mobs at different slow %, there was some issues with the animation changes occuring in tiers.

[rog]

this is something i've also been wondering. if any brd+rdm/whm on sylph wants to test it, i have suppa+relic pants, and a shinobi earring.