I don't know spike calculations well, but couldn't someone use blaze spikes and gather say, 20 of a mob to populate a large sample size very fast?
I don't know spike calculations well, but couldn't someone use blaze spikes and gather say, 20 of a mob to populate a large sample size very fast?
Assuming that Blaze Spike and Enspell accuracy is related to player Magic Accuracy, yes.
With the introduction of MAcc Atmas, this should be a very easy link to show.
Edit: I went out to play with it a little. 95% of the time I did 27 damage. 5% of the time I did 13 damage. RDM/SCH with 356 Enhancing skill (Light Arts Up) and 90+100INT. Specifically, it was resisted 89/2420 times (3.7% of the time +/-.75%),
I got a lower sample size with Dark Arts Blaze Spikes up (310 skill), but bottom line is that resist rate was still floored (5/206).
All of these tests are exclusively against EM black mandies in Attohwa, btw.
I'm going to try swapping out all my MAcc stuff (was wearing a smattering of AF3+2) and dropping Minikin to see if I can uncap my resist rate.
DCs vs. level 295 Enhancing magic with 0 MAcc in gear and 92 INT = 23 Resists / 509 Attacks. 4.5%. If there is a dependence of spike MAcc on gear, it hides itself quite craftily. I guess my last test could be on DNC/RDM
Quantifying Movement Speed
I've chosen to do it this way:
1) I make a script that looks like this:
2) I fill the ____ in with a time value, and find an NPC on relatively flat ground.Code:input /follow <t>; input /echo Following <t>; wait _____ input /echo There!
3) I set my starting distance (49') and position using mostly local landmarks/eyeballing, then I activate the script.
4) I watch the Distance plugin count down until it does the brief "you've hit the NPC's body" pause.
5) I adjust the time value by increments of 0.1 until I have one point before the pause and one point during the pause.
49'-1' for the NPC's body at the pause = 48' traveled.
Wash, rinse, and repeat with different test cases. Does this sound reasonable?
I remember a SS from a while back, right when Ixion was new from Aikar with movespeed of yourself and target visible. It must've been a private one, but 99% sure player speed is 4 distance/second.
Instead of using /follow (It messes with speed if your center is at a different z-axis value than the target) I'd lock on, then unlock and use "setkey numpad_8 down" until you get the clipping effect by running through the target.
It is 4 distance/second (minus time you waste going up and down), I actually did tests with this already. I figured going up and down wouldn't really affect the results, as long as it was relatively minor and I kept my path the same. Important, because there aren't a whole lot of perfectly flat places I could think of in Abyssea.
If you can use shorter distances, you could make a JA spam spellcast.
Macro:
/sc var set Movement 1
/follow <t>; wait 1
/ja Provoke; wait 1
/echo START
Spellcast:
<var name="Movement">0</var>
<if advanced='"$Movement"=="1"' spell="Provoke">
<command>sc var set Movement 0</command>
<command>wait 2; input /ja Provoke; wait .5; input /ja Provoke; wait .5; input /ja Provoke</command>
</if>
You then start from a set distance beyond provoke range (Like Provoke Range + 10?). Start the run by hitting the macro (Timestamp will log the start). When your character can finally voke, you'll see the timestamp within the log. That will give you your time and you'll have a set distance from when that time started.
Alternatively, using any command that shows "Target out of range" until you reach that target is spammable for close to accurate measurements. I assume this is for movement values not listed on any gear, which would then just be a /follow race/comparison.
Edit: Abyssea-Paradox should be close to flat I imagine.
Paradox is flat, but I'm not sure I can get 50' between Prishe and I. I'm trying to aim to do as close to 50' as possible so that I can get the maximum possible accuracy. Here are the three things I tested so far:
https://www.bg-wiki.com/bg/Raptor_Mazurka
https://www.bg-wiki.com/bg/Chocobo_Mazurka
https://www.bg-wiki.com/bg/Atma_of_the_Master_Crafter
I'll need to do Atma of Ambition (kind of, it's certainly the same as Crafter) and then probably Dusk gear/Retaliation too, but my character that can do that is busy doing multi-attack Atma testing!
I've been thinking about this whole MAcc vs. MEva things some more, and I might end up doing some testing with it tonight (if all my other plans fall through.)
Here is my idea:
1) Charm is affected by Light/Apollo's staff
2) Maybe it is affected because it depends on Light Magic Accuracy. This seems probable really.
3) Gauge tells the odds of being able to charm something, which potentially depends on MAcc vs. MEva from the assumptions above
4) If we figure out how and why the Gauge messages change (and what they mean), we can measure MAcc using Gauge instead of /check like how we measure regular Acc.
Does anyone know if this has been tried before, or if there is any testing related to Gauge? It seems like it would be work-intensive on the front end (verifying that the technique works) but then really easy after that. Downside is, maybe can't be used to test any MAcc Atma because nothing in Abyssea is charmable.
Where be these 75 vit Ul'Hpemdes? Ran around for 2 hours with 1 dmg weapons and 71 str, and couldn't find a single sucker that took 0 dmg. None of the Ul' or Om's in regular sea or the limbus areas seemed to work.
You found a Hpemde you hit for zero at what strength?
Tried to figure out what I did wrong, and I think I got it.
Either I suck at reading comprehension and am missing something obvious or the method for testing vit in this thread is incorrect.
Bolded certain parts for emphasis. That implies to me, that if you have a known value of vit, you can take that value, subtract 4 for a specfic strength value that should cause a 1dmg weapon to hit for zero.Determining VIT of a monster, Method 1:
Ingredients:
* A 1 base damage weapon of your choice
* Lots of different types of STR gear ranging from negative gear to positive gear.
* If you are unable to make yourself crit for 0, you may need more negative VIT gear or to change to a lower STR job.
* If you are determining the VIT of something with an ambiguous check message and level range whose level you can calculate from the xp it gives, you will need to be able to kill it (obviously).
Instructions:
1) Find the monster you want to test the VIT of, start beating it with your D1 weapon, and adjust your STR until you hit for 0 on Crits.
2) Take this amount of STR and add 4 to it to find the monster's VIT.
Theory:
By using a D1 (rank 0) weapon, your minimum fSTR is -1, which means you can swing with a weapon that has an effective base damage of 0. Critical Hits give you a minimum pDIF of 1, so if you think of damage like (Base Damage + fSTR)*pDIF, then the only way to get a 0 damage crit is to have a (Base Damage + fSTR) of 0. fSTR of -1 occurs when your STR is 4 points lower than your target's VIT.
http://wiki.ffxiclopedia.org/wiki/ST...rence_Function
I just went out and retested it, found an Om'Hpemde outside of Temenos that conned Tough to a lvl.75 (meaning given Om'Hpemdes are listed at lvl.73-77, this Hpemde had to be either lvl 76 or 77. According to:
http://www.bluegartr.com/threads/687...=1#post3204151
Om'Hpemdes at lvl76 or 77 both have 80 vit. At 66 str, my Lady Bell hit for 0, including crits. At 67 str, it hit for 1.
Meaning you need -14 dstr, not -4.
After digging through crap, I found this that seems to confirm:
http://www.bluegartr.com/threads/687...=1#post3193007
Ah sorry, by the time I realized my mistaken I'd lost mod status and couldn't edit the OP. Forgot about it.
Based off the formula it should be -11. Either the fstr function does not floor at the end step or his VIT numbers are wrong. -11dstr comes out to -.25fstr which, if floored, should be -1. Unless of course they changed how fstr works since I tested homam corazza on the little fishes back in the day.Meaning you need -14 dstr, not -4.
Maybe if it truncates however then -14 is definitely the number you're aiming for. Nobody would notice the difference between ffxi using flooring and truncation unless they were testing with negative values so it's very possible it's truncating.
Has anyone tried to code a simple app that loop through possible values while checking at each step if resulting formula yield a mismatching stat (STR DEX... CHR) for all possible combos (Level/Race/MainJob/SubJob) ?
Asking because the Player_Stats tables im hosting @ googlesheets got a few more updates. Maybe the time is ok to try now?
A little recall of the context: Since that update when SE "buffed" base stats, it appeared quite some disparities from currently known formulas, as can be seen in Motenten stats page. He tried to fix but there are still some...
I don't know which method he uses, so i'm asking here.
Is there a reliable way to determine if a dmg calculation is based on ranged calculations or standard melee calculations ?
Asking that in regard of Benthic typhoon, is its dmg calculated from ranged formulas (and the distance formulas) or from standard melee calcs?
EDIT: maybe a mob immune to ranged attacks will make Benthic (and by extension any ranged-based calculations) dealing 0 dmg? i don't know which mob has such immunity.
Okeanos in Apollyon - NE had ranged immunity. I think it's more like -87.5% ranged damage now, but you should still be able to tell.
How to test accurately Attack modded WSs, like IronTempest Sturmwind BladeShun Requiescat varying with TP, or Mandalic TrueStrike MysticBoon TachiYGKS Static cRatio boosts ?
Some of you would say some of those WSs are already tested but.... i didnot find at all any test.
By accurately i mean bypassing the randomness of PDIF.
EDIT: if find a mob's defense and testing the WS at various TPs but always same attack=mob's def, would that work ?
I am also interested in this, and would like to add Hexa Strike to the list as having a potential penalty.
So far my method has been:
* Melee enemies with relatively constant defense (if you keep monster families EP in Abyssea, they a 2-level range).
* Compare melee crits range (divide by base damage) to WS crits (divided by base damage).
What would be the best ratio range to do this for? Should we be using non-crits instead?
Anecdotal evidence for Hexa's penalty:
* Club base damage is generally equal to or higher than Axes.
* Rampage is 5.5 fTP and 30% STR in practice.
* Hexa Strike is 7 fTP and 20% STR/20% MND.
* They are either equal in base damage or Hexa is ahead, even considering WAR's equipment advantages.
* Hexa has more fTP.
* Why does Hexa do less damage than Rampage (even without Berserk) on most targets?
For weak numerical evidence, check the talk page on bgwiki.
Using pDif range seems the easiest/most accurate way of gauging attack bonuses/penalties. Works best when you can get a single hit of the weaponskill to land, so using the weaponskill at ~5% mob HP. Takes extra work to use with "damage varies with TP" weaponskills, though, since you spend extra time figuring exact base damage based on the TP at the time of the weaponskill.
Example methodology based on my suggestion for testing Shijin Spiral:
If the weaponskill can crit, obviously include crit damage range as well. While Greater Colibri were used in the example, mnk can more easily get away without worrying about TP moves such as Feather Tickle; other targets are preferable most of the time. You just need to be able to exactly specify the defense for each level of the mob, and be able to determine the mob level for each kill (generally by XP given).Originally Posted by Motenten
If the bonus can vary by TP, it's better to have a second person assisting by taking mobs down to low HP so that the weaponskill user can kill them at various TP points. A blm using specific nukes for exact damage totals could likely work well.
I think you are making a strong mistake adding each hits fTP for a total fTP saying "Hexa has 7fTP".
Each hit being calculated independantly, correct saying should be "Hexa has 1fTP".
Also if i'm asking how to test attack modded WSs, it's because i think ppl tends to think a fTP<1 = "Attack Penalty OMFG FU SE !"
...which isnot a tester/developper would not understand like this, but more like "Huh? cRatio (or attack) isnot modified there s no penalty or bonus..."
Semantics...
but in term of testing, such confusion changes a lot: someone thinking "attack penalty" will try to explain discrepancy between formula's basic results and parsed dmg, with a cRatio penalty...
...while other tester would keep cRatio mod=1 and try different combos {fTP, STAT1%, STAT2%, dSTATmultiplier}.
The other difficulty for proving an attack penalty is the randomness of PDIF which can easily lead to error...
Consequence: i don't buy at all those attack penalties rumors i see bit everywhere until someone show me how they proved this, i wouldnot be surprised to find an error like explained above.
On another topic, i re-tested Primal Rend and got a discrepancy right at 1st WS :
882dmg on a lvl65 Lesser Colibri(34exp without sanction as BST95/sam) with DEX112 CHR145 INT94 MaB11%.
lvl65 Lcoli stats: INT80 CHR75 MdB12%
Calculations were:
DMG = Floor( Floor( (Lvl+2 + Floor(Floor(DEX30%+CHR40%)*WSalpha))*fTP + dSTATmod*(playerCHR-mobINT)) * ((1+MaB)/(1+MdB)) )
DMG = Floor( Floor( (97 + Floor(Floor(112*30%+145*40%)*0.85))*4.5625 + 1.5*(145-80)) * ((1+0.11)/(1+0.12)) )
DMG = 883
I'll continue searching but imho i don't buy the dSTAT playerCHR vs mobINT, but at least i managed to prove there is indeed "something" from the mob impacting dmg :
with same WSgears i dealt different dmg on a lvl63 Lcoli than on a lvl65 Lcoli.
All tests were done @ 300TP btw, to bypass the "Damage varies with TP".