Resist Testing

Yeah, I'm aware of the different resists. Now, I understand you are just trying to get Resist! rates (I thought you were finding differences between both meva and resist specific gear for silence. Which, now that I realize you are capped no matter what you wear, it doesn't make sense for you to be testing meva).

So, now that I have that context. I understand now what you are saying about a cap of less than 90% seeing as how meva is capping at 86% when you are capped on meva. Got it.

I guess the only way to test meva+, resist status effect, and resist specific is to test on a higher level NM.

This seems to conflict with Sagasinger testing.

https://www.bluegartr.com/threads/108199-Random-Facts-Thread-Other?p=5345236&viewfull=1#post5345236

Sagasinger is also worded differently than any other items. Of bigger concern is that thus far, Etiolation and Shrieker's do not appear to have the same effect proportional to their listed quantities despite being worded identically.

It's been theorized by Pergatory and others that some items with 'Resist Status + X' provide X m.eva toward that status, while others provide a X% chance to resist that status outright. This is what I'm trying to nail down, and I will eventually also try to nail down values of items like Hearty Earring.

I've been parsing Founder's Greaves on Azi Dahaka for the past couple months, and it's damn close to +50% as well. Before I stopped recording, in 439 Absolute Terrors, I got hit 236 times (with a very small percentage of those times being in WS gear that doesn't have the greaves).

Hearty Earring is actually what I cared about the most so it's what I am testing atm in PvP but I plan to do it on your targets (MvP) to to see if it changes or stays about the same.

PvP Test:

^Seems mostly useless unless we figure out how it scales the gear. Now that we confirmed it gives it a huge bonus.

Gargoyles (NM) - Hearty Earring:
(Resisted!): 38 Regular Resist: 802 Landed: 129 Total: 969 Percentage: 3.921569



I wrote a little addon to block No Effects from counting and a bunch of other checks. I'll add it later once I change it from using logger since that seems to be giving a stutter now in Windower.
Edit: PvP Data gives a huge bonus to the gear. Might still prove useful if we figure out a scale.

Did a quick sample in brenner using only etiolation earring and the same set I used yesterday.

259 casts
28 silenced
68 resist
163 normal resist

26.2% resist! proc - It seems resist! is supercharged here for some reason, as suspected
85.3% normal resist rate(so cap seems similar, if not the same)

will leave the test a while longer and check with shriekers cuffs as well, since it's such quick and easy data collection

Finished up 20min~ run of Etiolation earring only brenner test, results:

768 casts
90 silenced
214 resist!
464 resists

27.8% resist! rate
83.7% resist rate(whm in inyanga+2 vs 5 blms in fast cast set.. but with the ballista resistance it's fair to assume that's also capped m.eva)

Adding this to second post, and doing a 20 min sample of shrieker's now.

Thanks for doing these tests.

Shrieker's + Etio (Brenner):
1000 casts(not intentional just happened that way i guess)
27 silences
782 resist!
191 resists

78.2% resist!
87.6% plain resists

So, a couple thoughts here..

We look at the ratio of resist! procs to resist silence+ listed on the gear:
78.2 / 27.8 = 2.813
40 / 15 = 2.667

--The ratio is relatively close to what we would expect. There's a strong possibility the pieces add linearly.

--Ballista testing is definitely not able to provide an accurate procrate. However, if it works linearly and we can figure out how to scale it, it may still be useful for measuring specific gear.

--My results to date on Gargoyles are very contrary to these results as far as Etio and Shrieker's are concerned. Perhaps it's an issue of small sample size, or there's another factor we have yet to comprehend at play here.

I've switched to your test since I got 2k samples in PvP but based on your test that data is useless anyway but I rather have had some hard #s then the eye balling over the years even if it was a lot of people saying the same thing.

I'm not quite done with ballista, since it offers so many samples in so little time. I got that 1000 cast sample in a single 20 minute round using my 5 blms with a brd to march/ballad them. Going to at least try out staunch on it's own, staunch + hearty, and staunch + etio to observe interactions. Right now, it seems to be approximately 2.75x more potent than face value, but I'd like to see if that holds true for other gear.

If you could post your full 2000 sample, it would be appreciated.

Quick question: Do the Arete del Sol/+1 earrings stack resists with each other?

That is the type of question this thread aims to answer, but currently no monster has been proposed for any debuff besides silence. I would suspect they do stack, but data collection from something like this is a very long process. The more people get involved, the sooner we'll be able to give sure answers to this type of thing.

I proposed pulling a few dozen Abyssea Mandragoras and letting them attack you until you time out for Sleep.

Carmine dobsonflies may be a good option for slow. Not sure how their mp pool is or how their cast priorities go, but worth a look. I'll run a character out with dagon and see how it goes once I'm done with ballista.

If you can find dynamis zones empty enough(maybe not so lucky for asura players), you could probabaly get some nice samples by training the zone and clearing everything except BSTs and pets. Slimes in jeuno cast para every 30 seconds, scorpions in bastok cast breakga every 30 seconds, ravens in windurst cast silencega every 30 seconds, and hecteyes in sandy cast dispelga every 30 seconds(maybe not quite as useful as the other 3, but dispel has some interesting resist properties too).

Staunch Tathlum +1 (Brenner)
895 casts
91 silences
210 resist!
594 resists

23.4% resist!
86.7% resists

Everything does seem to be roughly in proportion, suspect with more samples it would even out more, but ballista's inability to provide real #s is a limiting factor. To model ballista results to actual results we'll need a lot more samples from a PvE situataion. I'm going to see how practical dynamis is, and how much MP the BST pets actually have. If I can get TEs and pull them in 20 minutes, and their mp lasts the remaining 100 min, could potentially get a sample of 1500+ on real monsters in a run.

I think I cleaned it up good enough. Sorry no timestamps IDK wtf windower isn't logging timestamps now. I'll look after this but really the console log is really not needed with the addon I made. I know you probably don't trust it since you haven't seen the code so here is the version I used when doing those. I know you're not into windower but I'm sure you can follow it through NP.

data: https://pastebin.com/gXd5gDpW
addon: https://pastebin.com/Xwyf9esu

It's not that I don't trust an addon, it's simple enough to pull action message packet and decode it, it's that having a full console log gives a bit more data to later decode. For example, I had initially factored 'no effect' as a cast landing, but having the actual log instead of just raw #s let me later correct it when you pointed out that a same potency silence will default to 'no effect'.

Not sure there are any more cases like that, but it's just more data. Very easy to run a quick ctrl-f and count in notepad, or GREP the files to get numbers from a log file. Not so easy to get a log file out of numbers.

And the other problem with going direct from packets is that XI has a very strong tendency to send the same packet multiple times. To the best of my knowledge, windower doesn't correct that. FFXI automatically discards it, so only one instance makes it to log.

Windower removes incoming chunks from event handlers with identical payloads if their sequence # is less than 4 apart for the following IDs:

0x009
0x00A
0x00B
0x017
0x028
0x029
0x02D
0x030
0x06F
0x0DC

Dynamis is definitely viable. I was able to gather the BSTs, and only the BSTs, by going WHM/RDM and putting my mule on BRD/WHM outside party. If there were no BSTs in an area, I trained it on BRD and did Horde II. If there were bsts, I trained it on WHM and slept the non-BSTs using single target to avoid taking BSTs from my WHM. Looped from the first bst before bridges northwest around the bridges, and finally past AH to megaboss spawn, where my whm could safely hold them. Ended up with 10 BSTs and 10 crows, which matches the DAT, so safe to say I got them all. There were none near heavens tower, on top of the buildings, or on the bridge leading to windurst waters. Using the following for gear:
Terras/Mensch+1
Inyanga+2/Loricate+1/Etio
Inyanga+2/Shriekers/Defending/Shneddick
Moonlight/Fucho-no-Obi/Inyanga+2/Inyanga+2

I am able to stay alive using this set and no swaps with just phalanx and an occasional cure4 I built into a simple script alongside echo drops. 400 casts deep in a bit under 30 minutes and none seem to have ran out of MP. Currently seeing 42.6% for Shriekers+Etio here, the stark contrast to Gargoyles is questionable. I wonder if performance is reduced on NMs(dyna mobs are NOT itg), or it's just an issue of sample size.

Totally could be sample size but man this is smelling way to much like BST trait stuff now that you mention NMs...

P.S. I was on naked(except earring) 99 COR/nin in brenner vs my rdm(mastered) in a decently high macc set still from the refresh/fc pieces I choose. just for an idea on meva workings in there.

Well, that's a horrible thought... Resist+ potentially being nerfed vs NMs I mean. That should never be a mechanic, and SE shouldn't have done it to killer effects either.

Was curious if it would be viable to have a second character tag my birds when I time out so I can continue to use them, or mp would run out and it'd be a wasted effort. I went into dynamis bastok on other characters(didn't want to interfere with my own sample) and aspired a lv75 vanguard's scorpion dry. It had 12,730 MP.

Couple assumptions here, that crows have similar mp and that mobs are paying the 12 mp(paralyga), 32 mp(silencega), and 78 mp(breakga) that dats show. If those hold true, scorpions would be good for a bit under 2 hours, crows for over 4, and slimes for over 10 hours. Think this is as good as it gets for data collection.

Would love to get in on that testing but I know with all the dyna bots on Asura. I'd be cursing more then anything.

Hopefully can get some more people on lower pop servers to help out.

Still will be good to have more data from the Gargoyles if you're working on that. I will do a few different params on birds before going back there, but if NMs are different we definitely need a lot more NM samples to be sure.

You know. Considering that the dynamis mob are casting AoE's, couldn't you drastically increase the sample gathering speed simply by bringing additional characters that have the same resist+ gear in?

Could have every ga hitting 6 or 12 or more characters. The catch, I suppose, would be making sure they have the exact same resist+. Or, you could have different characters doing different tests at the same time. You'd just have to filter per character. But that could get a bit confusing.

Yea, I was planning on doing that for next round. I have 4 characters that can be made into identical WHMs. Got another 10 minutes before I time out (etio + shriekers sample atm), going to try to save my birds because pulling them was quite a hassle.

Nice. Sample sizes are gonna be nuts.

Etio + Shrieker's Cuffs (Dyna-Windy):

1599 casts
132 silence lands
610 resist!
857 resists

38.1% resist! rate
86.6% resist rate


Currently running:
Etio by itself
Shrieker's by themselves
Hearty by itself
Staunch by itself

all with same setup in other slots:
WHM/RDM (3 taru, 1 mithra, but all same MND)

Terra's Staff/Mensch Strap +1/x/x
Inyanga+2/Loricate Torque +1/x/x
Inyanga+2/Inyanga+2/Defending/Shneddick
Moonlight/Fucho-No-Obi/Inyanga+2/Inyanga+2