Prime WS Information, Testing, Discussion

Sorry to post on myself so much, but at some point do you just stop noting results unless it's a new minimum or maximum?

really min is the only number you care about.

Right, but I note new maximums because it informs on what the potential minimum value is.

Sarv Stage 4 Prime Bow:

ML40 Hume RNG/SAM
1640 ranged attack
140+240 STR
148+260 AGI
64% WSD

Tiny Mandragoras, West Sarutabaruta.

No Velocity Shot, No Hover Shot.

All WSs performed from 20.4' to eliminate any True Shot bonuses.

No TP Bonus Gear in sets. In "no fotia" testing, neck slot empty.


2k WS values vary most I believe as it was next to impossible to hit 2k exactly for me solo. Most performed between 2000-2050TP.
bah- final entry for 3k+fotia was a typo moment it matched the prior entries.

I'm guessing some of those aren't all the same tp? Should be no difference in this case. I'll do the numbers in a little bit

Some of the 1k values were early non-Sekka tries around 1030TP. the 22668 non-fotia 1ks were all Sekka results.

can you shoot rabbits or something that don't have a pierce weakness

This, dmg resistance/weakness makes it annoying. Could you also do it in a totally neutral gearset to remove any potential outliers that you may have forgotten to document?

can do the non-piercing np. please let me know what "potential outliers" you are referring to that i haven't removed already that need removed.

Since we don't know the specifics of the WS it could be any number of things. There are some oddities coming up from your numbers and we want to remove as many variables as possible.

I suggest Tunnel Worms for the mob; they can't move before you fire and are piercing neutral. You can also WS naked, you don't need any stats from gear here.

i would use enough str to cap fstr, though, since it seems likely it will have a str mod

No, I'd like specifics on what you think is causing outliers for my own personal knowledge bank, if possible.

I removed all TP bonus, got outside of a range where True Shot could possibly proc, no trusts, no /ja buffs. No crit gear in WS set. If there's something else I'm unaware of, I'd rather be told what that was instead of "do it naked".

i think something is off with pdl, possibly the trait so I am asking proth to test to verify. everything comes up off by about ~50 damage

but the worm idea is best, just wear enough str to cap fstr and avoid pdl and wsd if possible

if you could retest on worms (since they won't move), without any pdl or ws damage gear, i can get the right numbers later. base stats don't matter as long as you're capping fstr and can leave slots open to allow other increase in base stats for mods after

So the solution to the problem was simple, the original testing for PDL traits was sloppy and the conclusion was made incorrectly. For all but tier 5, the number is different than the assumed 0.1, 0.2, 0.3, etc. because the PDL+ trait is x/256, meaning there are small fractional differences if you compare expected, calculated ranged damage (any damage, but it's easiest to see on ranged) with in-game damage numbers i.e. 0.1 =/= 0.1015625. This won't make a big difference in practice (i.e., your gameplay isn't going to change) but it does explain the headaches with testing since PDL traits became a thing. Numbers not lining up despite there being no randomizer meant something was fundamentally wrong with the calculation, and this was the easiest answer. Luckily it was a good guess.

Anyway, I went and tested them all besides PDL+ IV because I don't care and neither does anyone else.

PDL1 - 26/256
PDL2 - 51/256
PDL3 - 77/256
PDL4 - lol idk
PDL5 - 128/256 (still .5)


Numbers should line up better w/ these values. Gear PDL still seems to work the same way as expected, i.e. just a straight +1.xx multiplier onto your final, post-trait pDIF cap.

When people mentioned small deviations from expected values, I assumed it likely had something to do with how FFXI use's /1024 or a similar divider for everything percentage. Like how people treat fotia as +0.1 when it's really +100/1024 (+0.09765625), or 100% DA is really 1000/1024 DA (97.65625%). Small enough that it won't really effect gameplay, but will frustrate testing.

Also, if you have access to a geo that can help you for a few minutes, geo bubbles make adjusting stats for finding potential wsc parameters very easy

Ah so my laziness in not removing some DA from my set and leaving it at 103 is actually paying off!

My plan for mods has been to do whm, geo, brd for boost, Indi-stat and etudes

Haha, that is how me and a buddy tripped over it. We were messing around with sets that had exactly 100 DA, and would occasionally see a single attack round. Both of us had the same parse data showing single attacks (flippants records this). We're talking a ~2% single attack rate, which lined up perfectly with 1000/1024. Added 3 DA and zero single attacks recorded after.

Someone in our group got stage 4 GK today, posting some limited testing here, sam/nin Mlvl50 on west ronfaure wild rabbits from behind to avoid overwhelm.

Will update this post with more data when we find the time for it, just some initial stuff for now

STR 180
DEX 188
VIT 153
AGL 158
INT 156
MND 140
CHR 148

3k tp naked
24191

3k tp fotiax2


3k tp naked with stat +44

STR 25908
DEX 24965, 25489
VIT 24029
AGL 24645
INT 24501
MND 24992
CHR 23818

On another note, are the values for etudes wrong on bgwiki? I was getting 19 str from the lower tier and 25 from the higher tier stat songs but wiki has them listed as 18 and 24. (Not wearing any emp for the stat bonus).
Edit: Nevermind I was in 2 pieces of empy.

Can't get much of anything from that yet but it points toward the mods probably being STR/DEX as expected. Every other weapon has followed the stats on it besides maybe dagger.

If you want to confirm without any reasonable level of doubt, add a GEO bubble in there for each stat. That'll separate the numbers nicely.

I did some testing yesterday with my GK. I haven't done enough to see the 5% WSD spread but I've got some rough numbers. All done naked with geo bubbles and some accessories to give stats. This is my first time doing this, so I welcome feedback.

Base stats: STR:180, DEX:190, VIT:149, AGI:154, INT:149 MND:154, CHR:144
Weapon damage 323


Base ---- Fotia --- 225 STR --- +49 STR -- +45 DEX

29100 --- 29206 --- 29887 --- 31645 --- 31117
29037 --- 29671 --- 30113 --- 31598 --- 31804
29318 --- 29478 --- 30509 --- 32370 --- 31657
29647 --- 29287 --- 29945 --- 32362 --- 30822
29031 --- 28727 --- 30730 --- 31126 --- 31584
29236 --- 29656 --- 31246 --- 32256
29091 --- 29757 --- 29836
29666 --- 29286
29332 --- 28781
28492 --- 28706
28407 --- 29739
29256 --- 29550
28506 --- 29021
28956

4.43% --- 3.66% --- 4.73% --- 4.00% --- 4.65% <-- dmg spread

The numbers I arrived at after accounting for overwhelm were 9.5 ftp, 58% STR, 47% DEX

I know I probably shouldn't make assumptions, but are there at least a few that we should make at this point about ws mods on these WS?

Assumption 1: All the mods are evenly weighted?
Assumption 2: multiple of 5?

I've been unable to get any new lows/highs on my dagger testing, but I know what the possibilities are. One of the "potential" results would point to pretty much exactly a 20/20 DEX/VIT.

More Ruthless Stroke stuff

I have confirmed mods are DEX/VIT using Brd, Whm, Geo to add single stat values one at a time. STR, AGI, MND, INT, CHA damage all fell within the normal range at 1k TP. DEX and VIT both provided damage well outside range.

	===	1000 TP	===	1000 TP Fotia	===	2000 TP	===	3000 TP
Min		4271		4346		11101		18270
Max		4482		4561		11603		19171
Possible Min?		4269		4344		11051		18259
Variance		1.04940		1.04947		1.04522		1.04932
Samples Recorded		184		115		18		66

I'm working on getting more samples at 2k, but I've basically stopped recording data unless it's a new high or low value.

I'm not sure how to lock it down, I've done tons of WS at 1k both with and without Fotia, but I haven't been successful at getting either a new high or new low value, even though potentially, both could go as much as 2 higher or lower.

Interesting enough, that 2 damage is significant in terms of potential WSC value, ranging from 62-74 based on whether I haven't currently hit minimums.

Based on Currently observed minimums:

	===	WSC	===	ftp 1000	===	ftp 2000	===	ftp 3000	===	WSC Multi
		68		5.57223		14.07702		22.53608		0.222222

Edit For Reference: base DEX 177 VIT 129

Those numbers are enough. I haven't run them myself but your results look fine, except WSC is almost certainly a total 70% for 35DEX 35VIT (possible it's lower but your numbers are tight enough that that isn't likely. Again though, haven't run anything myself yet). Those min/maxes are more than enough as it is, 2000TP could use some more but it can be worked with @ 4.5%. My guess would be something like

35% DEX
35% VIT
5.535~ @ 1000
14 @ 2000
22.5~ @ 3000

This is, again, without checking verifying or running any numbers myself.

I will now include my disappointment in how much worse the katana WS is compared to this.

How do you get 35% DEX and VIT from that out of curiosity?

Edit: I was putting it at somewhere between 20-24% DEX/VIT based on the potential minimums at 1k TP

Remember when Rudra's was nerfed because single hit 19.5 at 3k TP was too strong? We've finally become more broken than that again.

Because 68% @ 34/34 (or some other split of it, though these Primes have been even splits so far) makes no sense and doesn't follow any sort of paradigm that we've ever had for WSC. The only WSs that have had weird WSCs like those were the merit ones and some of the ones that only Trusts can use. 70% would be in line. I can't really check using your numbers since I don't know your job, DEX, or VIT though.

Oh, I see, 68 was an absolute value, not a percentage.

DEX 177
VIT 129

WSC 68

So I just did 68 / (177 + 129) = 0.22222