Alternator

言語: JP EN DE FR
2010-06-21
New Items
users online
フォーラム » FFXI » Jobs » Puppet Master » Alternator
Alternator
First Page 2 3 ... 9 10 11 ... 16 17 18
 Cerberus.Balloon
Offline
サーバ: Cerberus
Game: FFXI
Posts: 424
By Cerberus.Balloon 2013-05-14 19:32:35  
Fenrir.Sylow said: »
As an aside Armor Shatterer's additional effect is -40/256 (~16%) and is on the same level of priority as Frightful Roar and Corrosive Ooze

Also it seems to be very accurate

I know that it always hits four times or never, but any idea how that works? It's a very odd mechanic.

I mean, the acc check? Does it only check one hit? Or if it misses one does it miss them all? Or since the accuracy seems quite high, perhaps it only misses if all 4 hits miss? Is it a 4 hit ws at all?
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 19:36:43  
It's just a 1 hit 4 ftp WS
 Valefor.Prothescar
Guide Master
Offline
サーバ: Valefor
Game: FFXI
Posts: 19647
By Valefor.Prothescar 2013-05-14 19:47:02  
Fenrir.Sylow said: »
The raptor may be a bad example because Raptors are generally more evasive than the average mob of a similar level (some raptors are THF type but I don't know if the Raptor NM is)

The Raptor NM is - for example - extremely evasive compared to the Fish or Eft NM and even the Orobon NM

pretty positive it is a thf, it's triple attacked me on many an occasion.

I'd assist in data collection but my PUP is lv 85. Be interested in seeing the results tho
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 19:57:09  
Actually I need to see what a non-alternator maneuver bust looks like so PUP 85 is fine '~'

I'd do it myself but I'm standing in the middle of Bhaflau Thickets for some reason and I don't feel like going to town and buying an animator

Basically activate a naked automaton (disable any gear changes that will equip for maneuver gear and be sure you're not wearing any) - wait 3 minutes then pick a stat the auto will never always be higher than you for even after 3 maneuvers - spam maneuvers at recast (spamming enter on the menu works) until you overload and report

1) Number of maneuvers before overload
2) Duration of overload

It might be nice to report both the status timer duration and the finishtime-starttime timestamp duration


My new model is back to perfect prediction!
 Valefor.Prothescar
Guide Master
Offline
サーバ: Valefor
Game: FFXI
Posts: 19647
By Valefor.Prothescar 2013-05-14 20:21:53  
Will do it once I'm done with this Delve stuff. I don't normally use plugins for timing things as I find them to be fairly inaccurate due to latency or not being fast enough to report things on the screen, I use recast timers and my cell phone.
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 20:57:14  
Being super accurate doesn't really happen - since the automaton cools in 3 second tics I generally need multiple data points and I pick the most common 3-second bin for analysis!

But I really only need one value for non-alternator because it's just an offset and I'll be able to see where it falls immediately
 Cerberus.Pleebo
Offline
サーバ: Cerberus
Game: FFXI
user: Pleebo
Posts: 9720
By Cerberus.Pleebo 2013-05-14 21:14:46  
Ninja edits >O

Only did 2 trials though. Using turbo animator unless that's a problem.
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 21:37:25  
Turbo animator is fiiine
 Cerberus.Pleebo
Offline
サーバ: Cerberus
Game: FFXI
user: Pleebo
Posts: 9720
By Cerberus.Pleebo 2013-05-14 22:08:21  
Code
	Master	Pet
INT	80	94
		
		
# of man.	Duration	
3		21	
2		6	
3		22	
4		38	
5		54	
6		72	
5		55	
4		39	
6		71	
5		56	
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 22:15:00  
Looks to be the same as Alternator
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 22:24:58  
Code
[evaluate maneuvers.py]
N_man     t_min  t_max  
2         3    6    
3         21   24   
4         36   39   
5         54   57   
6         69   72   
7         87   90   
8         102  105  
9         120  123  
10        135  138  
11        153  156  
12        171  174  
13        186  189  
14        204  207  
15        219  222  
16        237  240  
17        252  255  
18        270  273  
19        285  288  
20        303  306
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 22:26:10  
If anyone wants to look at data!

https://docs.google.com/a/email.wm.edu/spreadsheet/ccc?key=0AoEfIbdoAw7adExWVTRyUDk4a0JIbmdPRkRyLXdya2c#gid=0
 Sylph.Knala
Offline
サーバ: Sylph
Game: FFXI
user: nala
Posts: 678
By Sylph.Knala 2013-05-14 22:28:33  
Probably not the best of places to ask this but since it's about some of the newer content from SoA.
Are they plan on buffing the auto's base damage? that is i get the alternator is beastly however from what i've read by checking up on this thread no one has mentioned if there is a boost to base damage.

Other than that know they are going to buff wyverns via pole arms... (i'm hoping that's just a portion of weapon damage added/scaled to wyvern damage tbh)

And probably new jug pets for bsts which i'm assuming will have higher base damage...

And that summoners are still SOL.
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 22:31:06  
JP wiki says Turbo animator may reduce overload rate so I'll test the regular animator Q_Q
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-14 22:51:10  
None of the three tested animators reduce overload rate! So either they ninja reduced overload rate when they added animator or JP numbers were just slightly off
 Lakshmi.Byrth
VIP
Offline
サーバ: Lakshmi
Game: FFXI
user: Byrthnoth
Posts: 6184
By Lakshmi.Byrth 2013-05-15 07:35:25  
How many minutes do you guys wait after overloading before you can test again?
Offline
Posts: 1018
By kenshynofshiva 2013-05-15 08:16:02  
pchan said: »
change job to 99 and cap your fists with formless strike while farming airlixir on the matamatat.

Also I didn't follow, did Jkun actualy time her puppet against a empty reive's rock to check if altenator is game breaking as she said ?

For info to wanabe mnks, one the easiest delve NM (the raptor) has 730 evasion and requires +285 accuracy on monk to cap. MNK can do it with 2 x madrigals, RCB, salvage/delve gear. What about pup ?

With the alternator the pup has a few advantages the difference in accuracy is basically the h2h skill and focus? So we use aggressor and have to eat sushi the puppet doesn't appear with the alternator so far to have any issues hitting and its 3k ws are pretty nice or 3k nukes. Pup with the alternator and rigors is 100% sweeter than ever. Parsing a delve run the puppet was at 9% damage add the master together and yeah its pretty fin good not king of all the things but not bad at all. The alternator is so good atm I fear its butts 2.0 and SE will be like sorry we made a typo and it got by test and yeah we all know that fin story...
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-15 08:16:46  
Lakshmi.Byrth said: »
How many minutes do you guys wait after overloading before you can test again?

You can just do another maneuver
 Lakshmi.Byrth
VIP
Offline
サーバ: Lakshmi
Game: FFXI
user: Byrthnoth
Posts: 6184
By Lakshmi.Byrth 2013-05-15 08:30:05  
I used 5 minutes just to be sure. My mule is using Fire Maneuvers every 12 seconds (8 times) and then waiting 480 seconds and doing it again. He's level 1 with a level 1 animator, and the puppet's STR is greater than his own.

I'll repeat this test with Thunder Maneuvers, because the puppet's DEX is lower than the master's.

Preliminarily:
* Someone said that puppets start with no burden. That's not correct I think.
* Currently I have gotten these numbers multiple times:
** 2 - 6 seconds (1x so far)
** 3 - 22 seconds (or 33 seconds? Probably a result of not-full-decay)
** 4 - 38 seconds (many times)
** 5 - 54 seconds (the most times)
** 6 - 70 seconds (several times)
* The fact that I'm using exactly 12 seconds between my maneuvers instead of ~10 and still getting the same values for 5/6 maneuvers indicates the decay may be pretty different from what we think. Perhaps it decays every minute (5 maneuvers = 60 seconds) instead of every 3 seconds?

I measure the overload time based on timestamps from the overload to wears off message using battlemod, but I'm probably underestimating overload times by a fraction of a second.


Edit: It looks like 5 minutes wasn't enough. Extended it to 480 seconds (8 minutes) and trying again.

Edit2: Resetting my count of number of overloads due to the change. Re-verified 4/5/6 maneuver overload durations. Overload 4 is always 38 seconds, not a multiple of 3.

Edit3:
There is exactly 16 seconds between my overload durations. Combined with my similarity to Sylow's data, which used a different time between maneuvers, this shows no decay at all. This does not mean no delay occurs, and I'll note that my numbers are 100% consistent while his are not. It probably means that the delay does work on a "tick" system and because I chose a multiple of 3 for my delay between maneuvers, I end up with a completely consistent value.

To take it a little further, 16 seconds is when I use one maneuver every 12 seconds. If 1 second = 1 degree when overloaded and heat decays at 1 degree per tick when not overloaded, then each maneuver would be giving 20 degrees of heat. Each maneuver lasts 60 seconds, so the heat from one maneuver would completely decay by the time it wore off. SE loves doing this ***, so I bet this is the system. Initial overload threshold would be 30 degrees.
 
Offline
Posts:
By 2013-05-15 09:50:44
 Undelete | Edit  | Link | 引用 | 返事
 
Post deleted by User.
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-15 09:54:22  
Puppets start with (apparently) 105 seconds worth of Burden according to JP model

It's worth noting that I get 105 as the intitial threshold with 60 as the starting value for Maneuvers with a decay of 3/sec

With all gear the threshold seems to bump to 255 (+15 +15 and +120)
 Lakshmi.Byrth
VIP
Offline
サーバ: Lakshmi
Game: FFXI
user: Byrthnoth
Posts: 6184
By Lakshmi.Byrth 2013-05-15 09:57:29  
Just got an 8-maneuver overload somehow. 102 second duration. So it's consistently +16 seconds per maneuver for me. (6+(n-2)*16) I seem to have removed the decay process.

As I said before, I suspect that the duration of Overload (which is not necessarily a multiple of 3 and therefore is not related to ticks) is determined by the number of degrees you were over threshold. It's just a debuff with a variable duration.
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-15 09:59:04  
Code
[evaluate maneuvers.py]
N_man     t_min  t_max  
2         3    6    
3         21   24   
4         36   39   
5         54   57   
6         69   72   
7         87   90   
8         102  105  


!
oh this is the formula I'm using to determine which tic the overload will wear off
Code python
from math import floor as floor
from math import ceil as ceil


heat = 60
opTemp = 105

print("%-10s%-7s%-7s" %('N_man', 't_min', 't_max'))
for n in range(2,21):
    
    totalHeat = heat * n
    overHeat = totalHeat - opTemp
    
    decay =  3* floor( 10*(n-1)/3 )
    
    durationEstimate = 3 *ceil( ceil( (overHeat - decay) / 3 )/ 3)
    

    print("%-10s%-5s%-5s" %(n,durationEstimate, durationEstimate+3))
 Lakshmi.Byrth
VIP
Offline
サーバ: Lakshmi
Game: FFXI
user: Byrthnoth
Posts: 6184
By Lakshmi.Byrth 2013-05-15 10:00:26  
Your t_min and t_max imply Overload works on a tick system, which it doesn't. It may decay on a tick system when not overloaded, but when overloaded it does not.

Going to leave my mule overloading for a while and do some work. See you!
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-15 10:02:51  
I suppose it could decay at 1/sec! I'll see if I can make a less annoying model that way!
 Quetzalcoatl.Lucieus
Offline
サーバ: Quetzalcoatl
Game: FFXI
user: Falsetto
Posts: 45
By Quetzalcoatl.Lucieus 2013-05-15 10:04:26  
All this research really makes me want to write up a PUP Guide for these forums so we can put all our findings there, instead of scattered in an Alternator thread lol.

EDIT: What happened to the one Kuching was supposed to write like a month ago??
 Leviathan.Kaparu
Offline
サーバ: Leviathan
Game: FFXI
user: Tsuko
Posts: 949
By Leviathan.Kaparu 2013-05-15 10:25:27  
Copy/paste Jinjo's MNK guide with a small addendum for attachments and a massive reminder that maneuvers are for solo play or when you're not engaged '~'
[+]
 Fenrir.Sylow
Offline
サーバ: Fenrir
Game: FFXI
Posts: 6862
By Fenrir.Sylow 2013-05-15 11:03:26  
It's probably worth it now to use maneuvers while engaged on a continuous target while using RNG automaton, but for plasm farming you can just use them between mobs.

Also worth mentioning: I suspect Armor shatterer has a significant amount of ratio correction on it
 Lakshmi.Byrth
VIP
Offline
サーバ: Lakshmi
Game: FFXI
user: Byrthnoth
Posts: 6184
By Lakshmi.Byrth 2013-05-15 11:15:59  
At the moment I have:
3 Maneuvers - 22 seconds - 3
4 Maneuvers - 38 seconds - 5
5 Maneuvers - 54 seconds - 8
6 Maneuvers - 70 seconds - 1
8 Maneuvers - 102 seconds - 1

I'll let him spam until the servers shut down, but this would take forever to actually make a probability distribution (which is why I haven't done it.)
 Cerberus.Kuching
Offline
サーバ: Cerberus
Game: FFXI
user: Kuching
Posts: 40
By Cerberus.Kuching 2013-05-15 11:24:57  
Quetzalcoatl.Lucieus said: »
What happened to the one Kuching was supposed to write like a month ago??

Finishing University currently. The addition of Adoulin requires me to relook all of the gearsets and now we gotta sort out this Overload/Maneuver/Animator relationship. Guide on hiatus - Please help.
First Page 2 3 ... 9 10 11 ... 16 17 18
Log in to post.