Perle Feet VS Aurum Feet Under Capped Accuracy

Language: JP EN DE FR
2010-09-08
New Items
users online
Forum » FFXI » Jobs » Dark Knight » Perle Feet VS Aurum Feet under capped accuracy
Perle Feet VS Aurum Feet under capped accuracy
 Quetzalcoatl.Giazz
Offline
Server: Quetzalcoatl
Game: FFXI
user: Gias
Posts: 1308
By Quetzalcoatl.Giazz 2010-10-05 22:08:15
Link | Quote | Reply
 
Phoenix.Excelior said:
Fenrir.Nightfyre said:
Phoenix.Excelior said:
Asura.Andradi said:
Phoenix.Excelior said:
There is no difference beyond the reduction in delay for the purpose of determining TP return. Oh and its not subject to any cap.
My understanding was that the sword strap doesn't have the exponential increase that haste does, meaning the sword strap set will not be getting anywhere near an extra 11% damage at high haste over the 5-hit, which is kind of a big difference. I could be misunderstanding how it works, though.
I know its calculated separately from haste but I'll have to go find the formula to see exactly where. If its calculated before haste then its just a static 3% increase in damage. If its calculated after haste then its subject to increasing returns. I have a feeling its after because JA haste is -delay with no TP penalty. Then again Dual wield is -delay with a Tp penalty. Since sword strap is the only item of its type I'm not quite sure where it goes in the formula lol.
Place doesn't matter (associative property of multiplication). It also has increasing returns since both are technically delay reduction (3.09% increase), it just has the drawback of decreasing tp/hit unlike haste.

Oh, I wasnt sure if it was just straight multiplication in the formula or not. Based on what you're saying then my math would be correct anyways.

You started going with the assumption that with Sword strap and V-belt you will have a +4% haste advantage, which is not true. That would make your calculations incorrect...
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-05 22:21:42
Link | Quote | Reply
 
Quetzalcoatl.Giazz said:
Phoenix.Excelior said:
Fenrir.Nightfyre said:
Phoenix.Excelior said:
Asura.Andradi said:
Phoenix.Excelior said:
There is no difference beyond the reduction in delay for the purpose of determining TP return. Oh and its not subject to any cap.
My understanding was that the sword strap doesn't have the exponential increase that haste does, meaning the sword strap set will not be getting anywhere near an extra 11% damage at high haste over the 5-hit, which is kind of a big difference. I could be misunderstanding how it works, though.
I know its calculated separately from haste but I'll have to go find the formula to see exactly where. If its calculated before haste then its just a static 3% increase in damage. If its calculated after haste then its subject to increasing returns. I have a feeling its after because JA haste is -delay with no TP penalty. Then again Dual wield is -delay with a Tp penalty. Since sword strap is the only item of its type I'm not quite sure where it goes in the formula lol.
Place doesn't matter (associative property of multiplication). It also has increasing returns since both are technically delay reduction (3.09% increase), it just has the drawback of decreasing tp/hit unlike haste.
Oh, I wasnt sure if it was just straight multiplication in the formula or not. Based on what you're saying then my math would be correct anyways.
You started going with the assumption that with Sword strap and V-belt you will have a +4% haste advantage, which is not true. That would make your calculations incorrect...

Why is it not a 4% haste advantage? Redo the math how you think it should be done because I'm clearly missing your argument.
 Asura.Andradi
Offline
Server: Asura
Game: FFXI
user: Andradi
Posts: 91
By Asura.Andradi 2010-10-05 22:29:40
Link | Quote | Reply
 
Phoenix.Excelior said:
Why is it not a 4% haste advantage? Redo the math how you think it should be done because I'm clearly missing your argument.

As far as I can tell your maths assume a high level of haste meaning the "4% haste" from [goading -> vbelt and sword strap] gives you an 11% increase in damage, which is wrong. You'd be getting more like 6% at most (I think) with 1% haste and the reduction from sword strap.
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-05 22:36:39
Link | Quote | Reply
 
Asura.Andradi said:
Phoenix.Excelior said:
Why is it not a 4% haste advantage? Redo the math how you think it should be done because I'm clearly missing your argument.
As far as I can tell your maths assume a high level of haste meaning the "4% haste" from [goading -> vbelt and sword strap] gives you an 11% increase in damage, which is wrong. You'd be getting more like 6% at most (I think) with 1% haste and the reduction from sword strap.

I'm not sure what he's arguing tbh. If the two stack in any way then its subject to increasing returns, which means that 4% haste can yield more than 11% damage depending on relative haste values.
 Quetzalcoatl.Giazz
Offline
Server: Quetzalcoatl
Game: FFXI
user: Gias
Posts: 1308
By Quetzalcoatl.Giazz 2010-10-05 22:46:57
Link | Quote | Reply
 
Phoenix.Excelior said:
Quetzalcoatl.Giazz said:
Phoenix.Excelior said:
Fenrir.Nightfyre said:
Phoenix.Excelior said:
Asura.Andradi said:
Phoenix.Excelior said:
There is no difference beyond the reduction in delay for the purpose of determining TP return. Oh and its not subject to any cap.
My understanding was that the sword strap doesn't have the exponential increase that haste does, meaning the sword strap set will not be getting anywhere near an extra 11% damage at high haste over the 5-hit, which is kind of a big difference. I could be misunderstanding how it works, though.
I know its calculated separately from haste but I'll have to go find the formula to see exactly where. If its calculated before haste then its just a static 3% increase in damage. If its calculated after haste then its subject to increasing returns. I have a feeling its after because JA haste is -delay with no TP penalty. Then again Dual wield is -delay with a Tp penalty. Since sword strap is the only item of its type I'm not quite sure where it goes in the formula lol.
Place doesn't matter (associative property of multiplication). It also has increasing returns since both are technically delay reduction (3.09% increase), it just has the drawback of decreasing tp/hit unlike haste.
Oh, I wasnt sure if it was just straight multiplication in the formula or not. Based on what you're saying then my math would be correct anyways.
You started going with the assumption that with Sword strap and V-belt you will have a +4% haste advantage, which is not true. That would make your calculations incorrect...

Why is it not a 4% haste advantage? Redo the math how you think it should be done because I'm clearly missing your argument.

If you read my posts on page two you would see where I stated that -delay from sword strap does not stack with haste. Therefore, you cannot say you gain a 4% haste advantage. (I showed you an example with a weapon delay, and it shows how it does make a difference in final delay by stacking it and not stacking it)

That being said, let's throw in more numbers:

We have a 528 delay weapon,

Let's assume first that you could stack sword strap with haste, and that we can go with that +4% haste advantage that you suggested.

At 66% haste, adding 4% haste for a total of 70% haste will be a ~8.8% increase. (This is without taking into account stuff like JA delay though)

Now, we do know that you can't stack sword strap with haste(since they're calculated separately), therefore this is what you would do:

Add sword strap, to get a 512 delay.

Then, you go from 66% haste to 67% haste (by using V-belt over Goading belt)

Here, 66% haste to 67% haste would be a ~3.5% increase. (without taking into account JA delay)

See the difference now?



 Quetzalcoatl.Volkom
Offline
Server: Quetzalcoatl
Game: FFXI
user: volkom
Posts: 1294
By Quetzalcoatl.Volkom 2010-10-05 22:48:28
Link | Quote | Reply
 
this...i don't know if it would work
 Asura.Andradi
Offline
Server: Asura
Game: FFXI
user: Andradi
Posts: 91
By Asura.Andradi 2010-10-05 22:51:23
Link | Quote | Reply
 
Not sure if troll.
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-05 23:04:47
Link | Quote | Reply
 
Quetzalcoatl.Giazz said:
Phoenix.Excelior said:
Quetzalcoatl.Giazz said:
Phoenix.Excelior said:
Fenrir.Nightfyre said:
Phoenix.Excelior said:
Asura.Andradi said:
Phoenix.Excelior said:
There is no difference beyond the reduction in delay for the purpose of determining TP return. Oh and its not subject to any cap.
My understanding was that the sword strap doesn't have the exponential increase that haste does, meaning the sword strap set will not be getting anywhere near an extra 11% damage at high haste over the 5-hit, which is kind of a big difference. I could be misunderstanding how it works, though.
I know its calculated separately from haste but I'll have to go find the formula to see exactly where. If its calculated before haste then its just a static 3% increase in damage. If its calculated after haste then its subject to increasing returns. I have a feeling its after because JA haste is -delay with no TP penalty. Then again Dual wield is -delay with a Tp penalty. Since sword strap is the only item of its type I'm not quite sure where it goes in the formula lol.
Place doesn't matter (associative property of multiplication). It also has increasing returns since both are technically delay reduction (3.09% increase), it just has the drawback of decreasing tp/hit unlike haste.
Oh, I wasnt sure if it was just straight multiplication in the formula or not. Based on what you're saying then my math would be correct anyways.
You started going with the assumption that with Sword strap and V-belt you will have a +4% haste advantage, which is not true. That would make your calculations incorrect...
Why is it not a 4% haste advantage? Redo the math how you think it should be done because I'm clearly missing your argument.
If you read my posts on page one you would see where I stated that -delay from sword strap does not stack with haste. Therefore, you cannot say you gain a 4% haste advantage. (I showed you an example with a weapon delay, and it shows how it does make a difference in final delay by stacking it and not stacking it) That being said, let's throw in more numbers: We have a 528 delay weapon, Let's assume first that you could stack sword strap with haste, and that we can go with that +4% haste advantage that you suggested. At 66% haste, adding 4% haste for a total of 70% haste will be a ~8.8% increase. (This is without taking into account stuff like JA delay though) Now, we do know that you can't stack sword strap with haste(since they're calculated separately), therefore this is what you would do: Add sword strap, to get a 512 delay. Then, you go from 66% haste to 67% haste (by using V-belt over Goading belt) Here, 66% haste to 67% haste would be a ~3.5% increase. (without taking into account JA delay) See the difference now?

Nvm I'm seeing your argument. I should have calculated the two seperately because they are not a "unit". This makes the answer
significantly different.

528-66%= 179.52

528-67%= 174.24

174.24/179.52= 3%ish increase in speed

174.24-3%= 169.01

169.01/174.24 = 3.1% increase in damage

overall 6% increase instead of 11%.

However, I still don't "know" that you're correct. I'd have to go look at the formula itself to detirmine the order of operations that apply to it. It would make quite a difference.

 Quetzalcoatl.Giazz
Offline
Server: Quetzalcoatl
Game: FFXI
user: Gias
Posts: 1308
By Quetzalcoatl.Giazz 2010-10-05 23:33:22
Link | Quote | Reply
 
Nah...it's actually only a ~3.5% increase in DPS, like I said and showed in my previous post. (but this is without taking into accounting JA delay etc, which would lower this a little.)

Just as Nightfyre and I both stated in page 2, the order doesn't actually matter. In other words, you can do -delay first and then add the haste(how I personally calculate it), or you can do haste first and then add the -delay. You just cannot stack the -delay with the haste. They're just simply calculated separately. Therefore clearly giving different results.
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-05 23:43:15
Link | Quote | Reply
 
Quetzalcoatl.Giazz said:
Nah...it's actually only a ~3.5% increase in DPS, like I said and showed in my previous post. (but this is without taking into accounting JA delay etc, which would lower this a little.) Just as Nightfyre and I both stated in page 2, the order doesn't actually matter. In other words, you can do -delay first and then add the haste(how I personally calculate it), or you can do haste first and then add the -delay. You just cannot stack the -delay with the haste. They're just simply calculated separately. Therefore clearly giving different results.

Nvm I see what you're saying. I was thinking the formula looked like this:

(BaseDelay)-(Haste% + delay-% +JA%)

However you're saying it looks like this:

(BaseDelay)-(delay%)-(Ja%)-(Haste%)

Which means that the multiplication within the parathesis allow for stacking but not outside of parathesis due to the order of operations. Am I correct?
 Asura.Andradi
Offline
Server: Asura
Game: FFXI
user: Andradi
Posts: 91
By Asura.Andradi 2010-10-05 23:44:53
Link | Quote | Reply
 
Yes, JA haste stacks with normal haste for melee swings though.
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-05 23:46:32
Link | Quote | Reply
 
Asura.Andradi said:
Yes, JA haste stacks with normal haste for melee swings though.

K, do any of you have a link to a place where this ***is referenced? I'm curious.
 Asura.Andradi
Offline
Server: Asura
Game: FFXI
user: Andradi
Posts: 91
By Asura.Andradi 2010-10-05 23:59:01
Link | Quote | Reply
 
Here is a decent rundown.

Quote:
Job abilities: 25%

* Haste gained from job abilities does not reduce spell recast timers
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-06 00:03:12
Link | Quote | Reply
 
Asura.Andradi said:
Here is a decent rundown.
Quote:
Job abilities: 25% * Haste gained from job abilities does not reduce spell recast timers

That's not what I was asking for really. I know the types of haste. I'm looking for the mathmatical formula that shows the calculation done by the computer.
 Quetzalcoatl.Giazz
Offline
Server: Quetzalcoatl
Game: FFXI
user: Gias
Posts: 1308
By Quetzalcoatl.Giazz 2010-10-06 00:07:11
Link | Quote | Reply
 
Asura.Andradi said:
Yes, JA haste stacks with normal haste for melee swings though.

This would be correct, yes.

-delay from Job ability is counted as job ability "haste" ...therefore that actually stacks with other haste items towards the total haste cap of 80%

-delay from stuff like sword strap is not counted as "haste", and is actually calculated separate from haste.

Also note that -delay from Job abilities like desperate blows and hasso do not reduce your TP gain per hit. They work like haste; they are "job ability haste" after all. (Work like haste, as in: they reduce your weapon delay without reducing your TP per hit)

-delay from stuff like sword strap does not work like haste, and it actually reduces your TP per hit. From a logical stand point, it wouldn't really make much sense to stack it with haste, since it's not haste because it works different from haste.

Phoenix.Excelior said:
Asura.Andradi said:
Yes, JA haste stacks with normal haste for melee swings though.

K, do any of you have a link to a place where this ***is referenced? I'm curious.

Sorry I do not have, in hand, links to the original testing for this in particular. I know some were done on BG by Kirschy, along with other haste tests iirc.
 Fenrir.Nightfyre
Offline
Server: Fenrir
Game: FFXI
user: Nightfyre
Posts: 11680
By Fenrir.Nightfyre 2010-10-06 00:09:34
Link | Quote | Reply
 
Would be somewhat more accurate to write it as basedelay*(1-delay%)*(1-haste), recalculating TP/hit after calculating delay reduction.

Dunno about original testing (Hasso, Desperate Blows), but Kirschy's old Haste Samba data clearly points to that form of JA haste interacting as expected with other sources of haste.
 Phoenix.Excelior
Offline
Server: Phoenix
Game: FFXI
user: Excelior
Posts: 2093
By Phoenix.Excelior 2010-10-06 00:16:47
Link | Quote | Reply
 
Fenrir.Nightfyre said:
Would be somewhat more accurate to write it as basedelay*(1-delay%)*(1-haste), recalculating TP/hit after calculating delay reduction. Dunno about original testing (Hasso, Desperate Blows), but Kirschy's old Haste Samba data clearly points to that form of JA haste interacting as expected with other sources of haste.

Ahh. Idk how you'd even test delay interactions, it would require eyeballing unless you had a very accurate way to time swing rates. ?
 Fenrir.Nightfyre
Offline
Server: Fenrir
Game: FFXI
user: Nightfyre
Posts: 11680
By Fenrir.Nightfyre 2010-10-06 00:23:50
Link | Quote | Reply
 
Testing over an extended period of time gives you a fairly accurate answer, though not as precise as recast tests. Beating up walls in Lebros Cavern, Fortifications, or Fortalices with a low damage weapon and comparing expected number of rounds to actual results is the usual method I believe.
Log in to post.