Spirit Recast Timer Cap Found! i think >.>Follow

Summoner Spirit Casting/Recast timers.

THIS LOOKS A LOT BETTER HERE...


http://killingifrit.com/forums/index.php?showtopic=173985


Still will test more... kinda work in progress.

http://www.youtube.com/watch?v=HSqYN02g9t8

Just uploaded video... if it doesn't work give it sometime.

Preface:

I have always wondered if there was a cap to the recast timers on our spirits spells, or if we had unlimited reign to lower it so we could be casting spells every seven seconds (which is the theoretical maximum low possible). Theoretically you could achieve a seven second recast timer, Because of this I also wanted to test to see if during the cast of Ancient Magic you could use the "Assault exploit" to have your elemental cast another spell on top of the Ancient Magic. Its a long shot, but heres to hoping.

Testing Criteria:
This is what is needed to achieve the theoretical maximum lowest recast timer of 7 seconds.

* maximum smn magic skill build

* capped summoning magic skill

* 2hr ability

* Day of the week matching

* matching weather effect

* summoner spats


Actual Test Criteria #1:
This is how close to the cap I was for the theoretical maximum recast timer.

* maximum smn magic skill build= X (335 skill -4 off from absolute maximum using Smn Spats instead of Oracle Braconi or Marduk Pants)

* capped summoning magic skill= O

* 2hr ability= O

* Day of the week matching= O

* matching weather effect= O X **Only one weather effect actually counts towards recast timer, so double weather is not needed**

* summoner spats=O



Things to know about Spirits:

First I have to say that I couldn't do this test without Spira's guide on spirits. You can find it here if your interested where I got the equations for recast timers and other information.

http://wiki.ffxiclopedia.org/wiki/Spirit_Guide_by_Spira

The following bullets are key facts needed to test what i wanted to test. These are taken from Spira's guide.

* The specific duration of the spirit timer is from the time your elemental completes its first spell until the time it starts casting the next spell. Alternatively, if it is interrupted, from 1 second after it becomes interrupted.
* Before considering external factors or gear, the timer is 45s for Summoners with Summoning skill capped.
* For every 3 skill above or below cap, the timer decreases and increases by 1 second respectively.
* The current day can influence the timer by 3s. -3s if the day is of the same element as the spirit, +3s if it is of the element which the spirit is weak to.
* The timer reduces by 2s if the weather is the same element as the spirit. It increases by 2s if weather is of the element which the spirit is weak to. Weather from all other elements do not affect the timer. Dual weather effects have the same influence as single weather.
* The Summoner 2HR ability Astral Flow reduce 5s from the timer.
* The Summoner AF2 "Summoner's Spats" also reduce 5s from the timer.
* The spirit timer is dynamic and can change if any of the factors affecting it change at any time.
* All elementals when freshly summoned, have their spirit timer already counting down from half.
* The pet commands "Assault" and "Retreat" reset the spirit timer.
* Using "Assault" when your elemental starts casting a long spell starts its next timer prematurely, allowing it to cast its next spell sooner.



Event log:

22:21:27 Astral Flow used.
22:21:40 Assault command given to start timer to cast first spell.
22:22:05 Spirits starts to cast Freeze.
22:22:08 Assault command given to reset casting timer.
22:22:25 Ice spirit casts Freeze.
22:22:26 Ice weather wears off.
22:22:36 Ice spirit starts to cast Frost.
22:22:43 Ice spirit casts frost.
22:23:06 Ice spirit starts to cast Frost.
22:23:13 Ice spirit casts Frost.
22:23:33 Ice spirit starts to cast paralyze.
22:23:41 Ice spirit casts Paralyze.
22:23:54 Thundersday
22:24:00 Ice spirit starts to cast frost.
22:24:08 Ice spirit casts frost.
22:24:26 Astral Flow wears off.
22:24:28 Ice spirit starts to cast Blizzard IV
Buffalo dies with a nuke to da face!


Taking a look at the data:

Using this formula 48-(Current summoning magic skill - Natural skill cap)/3 (taken from spira's guide), I should have had an 8 second recast timer, and 10 second recast timer once weather effect wore off after freeze. Let's take a look at my raw data and see if this holds true, or if we are hitting a cap.

Time between Assault command and casting of first spell : 25 seconds
Assault command given 3 seconds after freeze started to be cast.
Ice spirit cast freeze 17 seconds after assault command given.
Ice spirit casts frost 11 seconds after freeze was cast. OR 28 seconds after the assault command was given.
Ice spirit casts frost 23 seconds after last frost was cast.
Ice spirit casts Paralyze 20 seconds after frost was cast.
Ice spirit casts Frost 27 seconds after paralyze was cast. (changed to thundersday so 3 seconds was added which might explain the 27 second cast time)
Ice spirit casts Blizzard IV 20 seconds after Frost was cast. (this 20 second lapse confuses me when I had 27 seconds prior.)

Conclusion:

Overall I am pretty confused. The exact cast and recast time I am unable to say for sure. It seems there are some variables that have yet to be found. I say this because to me it doesn't make any sense why it took 27 seconds to cast frost on thundersday, and then 20 seconds later on the same day Blizzard IV was cast. I did not see anything cast any lower than 20 seconds, so its possible that is the lowest obtainable. The only other way would be if your elemental casts Ancient Magic, and you reset the timer with the Assault command. If that is the circumstance you can see it cast a spell 11 seconds after it finishes casting the AM.

I was a bit disappointed with my results. I was really hoping that I was going to see an elemental cast a spell every 7 seconds, or close to this. Seeing that the lowest recast for a spirit is 20 seconds, that pretty much kills my hopes for Ancient magic to be casted on top of an existing Ancient magic cast. I know that was a long shot, but hey you got to dream right? lol...

Anyways.... To sum up the whole test. The absolute lowest you can recast a spell with your spirit is 20 seconds, unless you do the assault command during Ancient Magic, or possibly Blizzard IV. If that's the case you can see your spirit casting much faster possibly 11 seconds after the Ancient magic or tier IV spell lands.


I know this isn't game breaking for summoner, but gives you a little more insight with our job. If any of my data or tests are wrong, please feel free to PM me, and ill retest with other theories or fixes to what I previously did.


Thanks for reading my Wall O txt.

Darkrift
Hades Server
Aol IM: Bananachaos7
Smn 4 Life


ps.

I spent a decent chunk of time writing this up... If you feel I am dead wrong, please contribute something so i can retest and provide better data. I plan on testing more.. I guess I am kinda looking for some feedback.

Screenshots:

[img]http://i61.photobucket.com/albums/h71/rice7r/9b56c7a9.png[/img]

[img]http://i61.photobucket.com/albums/h71/rice7r/00127118.png[/img]

[img]http://i61.photobucket.com/albums/h71/rice7r/d49f9c12.png[/img]

Edited, Nov 23rd 2008 4:46am by Darkrifthades

Edited, Nov 23rd 2008 1:16pm by Darkrifthades

Edited, Nov 23rd 2008 1:17pm by Darkrifthades

Edited, Nov 23rd 2008 3:45pm by Darkrifthades

you don't, the Twilight Zone has control of the vertical and horizontal.. There is a little tag that says Markup Help when you are posting, it has what you can do in there.

k thanks ill fix the link, and im definitely not done... was planning on more testing, just wanted some feedback.

Link to the video is a little off, try this: http://www.youtube.com/watch?v=HSqYN02g9t8

Tried watching it, unfortunately was too fuzzy to make out the vana' clock timer by the compass. I'm wondering if possibly the variance between 20 and 23 seconds is due to how each tick isn't exactly 3 seconds (sorta how a Sam with AF helm can sometimes get lucky and get off a 140 TP Meditate rather than the usual 120). If that's the case, maybe the last used spell makes a difference to the recast? If you just adjust up to 3 seconds, for this, you get the following:

Assault
Start casting a spell 25/28 seconds after Assault
Assault
Start casting a spell 25/28 seconds after Assault
Start casting a spell 20/23 seconds after Frost
Start casting a spell 20/23 seconds after Frost
Start casting a spell 27 seconds after Paralyze (since day changed, this might have had an effect)
Start casting a spell 20/23 seconds after Frost

So, maybe there's a 20/23 (depending on exact vana'timer) recast cap after a previous spell, and 25/28 (vana'timer) recast cap after Assault?

It's also possible there is a variable cap based on what the previous spell used was. Perhaps nukes have a higher recast, paralyze has ~27 (could be as low as 21, if you deduct 3 seconds for the day change and 3 seconds for the timer), and Frost has ~20/23? Unfortunately, the mob didn't survive to get hit by another spell after that last nuke, so we don't know what the recast would have been after a regular IV nuke.

I've heard that the spell cast doesn't change the recast, but maybe those recasts weren't hitting the cap so we never would have known?

At any rate, thanks for testing this out!

lol..
funny no one ever pointed out that inconsistency in my guide.. and i never noticed it..

use 48, thats the correct one.. the 45 must have been a typo..
also i didn't upload the guide to wiki and don't manage it so if anything on it has been changed or edited by someone else i won't know of it anyway.

anyway i'm glad someone is still doing further tests on elemental spirits and smn magic skill in a capacity that i'm not able to perform at.
personally i've stopped doing any form of testing for anything (hell, i haven't really been playing the game for months now) so I can't help you..

only thing i can suggest is start with lower SMN skill for your normal and then increase 1 variable at a time.. try to leave weather, day and relic spats out of the equation as much as you can to hopefully establish the timer cap based on skill alone..

and only from that point add in the new variables 1 at a time, to see how each will affect it..

may altana guide your way

That would be interesting to find out. Maybe it's rounded up or down depending on the exact time the spirit's cast ended. If it ended before the halfway point, it's rounded down etc...

idk, just a theory.

I couldn't see this suggested anywhere but...

Ignoring the assault = reset recast, wouldn't the actual spells be fired at around 80% spell casting meter, and depending on total casting time, that last 20% can take different time.

I think that could explain a slight difference between different spells (but not between say frost and frost and frost)

Going to einherjar so had to post fast.

Too lazy to edit, but you should do some separate testing on light spirit. I *think* it casts faster, which means it might have a different recast min.

I don't typically post here, im a KI poster.

I was unaware you did any testing.... I would like to see your data though if you can find it. Post the link if you can? :D

If your data is good it could save me some time of having to test more stuff.

Some of the calculated numbers were off; I had to draw my results from the timestamps directly. If any of my own math is off, please feel free to correct (can be rather confusing to keep everything straight)

Cast times:

Protect IV: 7, 7, 7
Shell IV: 6, 6, 7
Haste: 7, 8, 8
Regen: 9, 8, 8

Protect III: 6, 6, 6, 6
Shell II: 6, 5, 6, 5
Haste: 8, 8, 8, 7
Regen: 7, 8, 8, 7

Preliminary conclusion: cast time is based on vana’timer ticks, not clock seconds. Two second variance on Regen indicates that at least up to 2 seconds variance is conclusively proven, unless the lower-level Light Spirit somehow casts Regen faster.

Time to start casting after previous spell completes:

After Protect IV: 14, 16, n/a
After Shell IV: 12, 12, 21 (wtf?)
After Haste: 14 (*), 19, 12
After Regen: 22, 13 (*), 16

After Protect III: 11 (*), 14, 16, 19
After Shell II: 12, 20 (wtf?), 13 (*), n/a
After Haste: 13, 14, 11, 12
After Regen: 17, 19 (*), 16, 14

(*) = all buffs were erased after spell was cast, this perhaps could have delayed it slightly.

Analysis: Wide variance in each category. Protect IV doesn’t have enough data, but Protect III has a variance of up to 5 seconds, including a 19 which is higher than either value listed for Protect IV. Shell IV has a variance of up to 9 seconds and Shell II has a variance of up to 8 seconds. Haste has a variance of up to 7 seconds, with the higher-level Light Spirit having a higher ‘spike’ than the level 50 values. Regen similarly has a variance of up to 5 points, with the higher-level Light Spirit having a higher spike.

Preliminary conclusion: difficult to believe 9 seconds variance can be explained by the vana’timer vs. earth clock difference, especially so frequently. More source data would help prove or disprove, but at this time evidence would suggest the determining factor is not when the previous spell is marked as ‘cast’.

Hypothesis: perhaps the timer is modified based on when the previous spell finished casting AND the cast time of the new spell? For example, after casting Regen, the spirit will wait for a base of, say, 10 seconds, and then it will add from 4-12 seconds depending on which spell is being cast? Would need more source data to test. For each level where all four spells are available, after each spellcast there are four possible options (any of the other three spells, or a repeat of the same spell). Levelcap down to a level where fewer options are available (such as down to a point where Haste is not an option, or below 21 so only Protect (I) and Shell (I) are available), or subbing Smn to another job (for example, if you have a job at 51, sub Smn to it and you can equip a Light Staff, Avatar Belt, and get Refresh and/or Ballad and/or Evoker’s Roll, letting you keep the spirit out indefinitely), would help determine this.

Start time to start time (level 50 Light Spirit):

Protect III -> Protect III: 17
Protect III -> Shell II: 22
Protect III -> Haste: 20
Protect III -> Regen: 25
Shell II -> Protect III: 19
Shell II -> Shell II:
Shell II -> Haste: 18
Shell II -> Regen: 25
Haste -> Protect III: 21
Haste -> Shell II: 22, 19
Haste -> Haste:
Haste -> Regen: 21
Regen -> Protect III: 24
Regen -> Shell II:
Regen -> Haste: 24, 21
Regen -> Regen: 27

Definitely need more evidence. What little evidence applies to this test shows an acceptable variance (3 seconds), but there is definitely too little to draw any definite conclusions.

Edit: adding in the spellcasts from the higher-level Light Spirit gives the following:

Haste -> Regen: 21, 27, 20
Regen -> Haste: 24, 21, 21

That's a 7 second variance on Haste -> Regen, and within the same tier Light Spirit too. {Hmmm.}


Edited, Nov 25th 2008 2:57pm by SylphiaOnFairyServer

Last I heard, buffs on the character has no effect on the character's pet, unless specifically listed in the buff. Thus, casting Haste on the Smn SHOULDN'T have any effect on the Spirit, nor should the possible presence of Fast Cast or Light Arts.

Really? If the Light Spirit casts Haste on itself, it will not benefit from it at all?

While I think that it would, I don't think I've ever seen the spirit cast haste on itself. I'm not saying it can't/doesn't/hasn't, just that to my personal recollection, I've never seen it.
Someone with more spirit use can tell us whether the spirit casts haste on itself?

You should all be aware that spirits only cast on what they see. They do not see themselves obviously. So at best you could try 2 SMN with 2 light spirit, make them look at each other and see if they can cast haste on each other.

Though since I doubt you can haste wyvern/bst pet/puppet, hasting spirits is probably out of the question.

man, im glad people are still looking into this formula. I remember getting information on this formula for the spiral guide (which is why there are different versions, since over time, people updated it with different formulas etc). I havnt had much time to look at this whole thread, but thought ill add that the who formula is an approx, since i think the proper formula has a power of 2 or 3 chucked in there somewhere, either that, or itat certain time, the formula changes, since i noticed that the lower you go (timer wise) the less accurate the formula is.

Anyway one funny thing to add, is that Spirits do have a limit on MP. Since i left out my Light spirit in limbus (light weather), and after about 20 or so minutes, the spirit stopped casting spells and i could not gain anymore mp from it. After a recast, all is well though.

Anyway, i think the biggest thing we could possibly learn, would predicting which spell it will use, in which i imagine has many variables

This is true, CM proved it a while back, and was quite interesting.
Somewhat disappointing, but this is the same thing that has been mentioned added to Avatars. A longer timer, where they free cast a 'bloodpact' at their own mp pool cost not ours, that promotes leaving avatars out.