Test Server Findings

[Yugl]

Masamune conducted the tests (And he's a very reliable/strict guy), so you'll have to PM him for the experimental data. As for the language on the page (For clarification purposes), a strict reading of the page is as follows:

After cRatio has been calculated, there are three stages to determine the final pDIF
1. Crit or no Crit [Debate over whether this is a straight +1 to pDIF or just a change in pDIF equations; BG-Wiki uses the latter]
If Crit => Use different pDIF formula
If no Crit => Use normal pDIF formula

2. Take the value from (1) and cap that value at 3 if above 3

3. Multiply the final value by 1-1.05

I recall 2 and 3 being known for quite a while. The "change in formulas" for critical hits is new to me, but that doesn't apply here.

[dasva]

Yeah they've changed the formulas a few times. Like before non crit used to actually have a cap.

Also side note I think somewhere on his site Robonosto tries to determine average pdifs for given Cratios based on distributions they end up falling under which is worth looking at. Though reading some of his stuff can make your eyes and/or brain hurt sometimes lol

[Byrthnoth]

Haha, check it out!

[Colenzo]

lmfao

[Prothescar]

is confusing too...
109*0.85=~92
(92+22)*5.375*2.92(if that really is max pdif)=1789 ...
Where is 2830 coming from?

You're neglecting to add in Mighty Strikes and WAR's Critical Attack Bonus, and fSTR.

[Atoreis]

You're neglecting to add in Mighty Strikes and WAR's Critical Attack Bonus, and fSTR.

Ah sorry tho thats w/o mighty strikes.

Anyway I have some data for you guys:
DRK/SAM
119STR
fSTR 10
D22 (claymore)
No DA gear obviously

highest normal hit observed was 87 with claymore and thats 87/32=2.71 (I really have doubts for 2.92 max pDIF because it was easily like 300 hits)

1. 736 ATT
highest Resolution 998
pDIF: 1.5

2. 901 ATT
highest Resolution 1268
pDIF: 1.9

3. 1015 ATT (this one is very low sample because I died)
highest Resolution 1475
pDIF: 2.21


Feel free to try figure out some formula for this because I'm too sad (seeing those results) to even try atm :/

[Byrthnoth]

Against Colibri? Need to know so we can figure out your fSTR.

[Atoreis]

Against Colibri? Need to know so we can figure out your fSTR.

Yeah yeah Im a puppet so I went to colibri as told ^^b

EDIT: Now I think I should have tried full attack and check if I can actually cap pDIF with like 1300+ att.

EDIT: It looks like around -30%att and in this case it still might be better for Mighty strike zerg than Ukko's but much closer than I think it was 2 days ago :/ and probably on par with Upheaval now.

[Byrthnoth]

Hmm... since the "highest damage" is inherently an average of 5 pDIFs, it would have been a little better to give the average 5-hit Resolution.

Still, we can guess that your Ratio was approximately 1.4~1.5, or about 2/3 what it should have been. That holds up for the other attack values as well. It looks like the penalty is applied before the cap, at least.

[Atoreis]

Hmm... since the "highest damage" is inherently an average of 5 pDIFs, it would have been a little better to give the average 5-hit Resolution.

Still, we can guess that your Ratio was approximately 1.4~1.5, or about 2/3 what it should have been. That holds up for the other attack values as well. It looks like the penalty is applied before the cap, at least.

Well last sample with 1015 att and highest 1475 suggest that if it's really -%att than its not more tahn -28.5% since 1475 is ~2.218 and 1015att is ~3.1
2.218/3.1=~71.5%

[seiji]

Entropy, Resolution, and Stardiver are the only 3 melee WS that Gobli said they suspected where 1st hit fTP goes to subsequent hits.

anybody tested with high multi attack rate to see if initial ftp also goes on extra hit from multiattack or only on natural #WS hit?(only hits from base WS get "nerfed")

[VZX01]

Ah sorry tho thats w/o mighty strikes.

Anyway I have some data for you guys:
DRK/SAM
119STR
fSTR 10
D22 (claymore)
No DA gear obviously

highest normal hit observed was 87 with claymore and thats 87/32=2.71 (I really have doubts for 2.92 max pDIF because it was easily like 300 hits)

1. 736 ATT
highest Resolution 998
pDIF: 1.5

2. 901 ATT
highest Resolution 1268
pDIF: 1.9

3. 1015 ATT (this one is very low sample because I died)
highest Resolution 1475
pDIF: 2.21


Feel free to try figure out some formula for this because I'm too sad (seeing those results) to even try atm :/

I'm skeptical, if it's in-any-way-accurate, we are able to figure out the attack penalty this way, because if those are 5*pDIF max and assuming pDIF max happening 50% of the time, then the expected outcome of 5 pDIF max occurs at once is at 3.125%.

Now if the chance you see max pDIF is 10% on distribution (which I think, still a bit higher to be true), you need 100,000 WS to see this case, on average.

ADD: I think we should target mob with lower HP with known or easy-to-figure defense value, so that we can just rely on 1 hit pDIF instead of dealing with randomness of 5 hits
Chigoe? Detector in sky?

[Byrthnoth]

Calculating it exactly using only his highest damage Resolutions gives you a lower bound for the penalty (minimum penalty) that comes out around 30%. Realistically (because we were using his highest damage resolutions instead of average of 5-hit resolutions), the actual penalty is going to be larger than that. So it would be more accurate to say it is "at least" a 30% penalty. On BGwiki I have it as about a 1/3 penalty at the moment, but it really could be as high as 50%.

[VZX01]

Translating what we know atm, it effectively changes a 5-hit WS into a 2.5~3.3-hit WS.
Really ewwww..

[Raelia]

Ehhhh 2.772 is with the 1.05 already thrown on. 2.92 you've thrown the secondary randomizer twice for some reason.

2.2 capped Ratio (apply LC here) -> 1.2 max pDif multiplier -> 1.05 max secondary randomizer = 2.772

Don't know why people continue to screw this up. The only time it gets poked at anymore is with crits and the whole 1.0 vs different formula debate.

[Byrthnoth]

Is this gobli link from a few pages back claiming that SE changed a bunch of the 300-skill WS mods?

[VZX01]

We need the pDIF distribution... If anyone can attach it to forum, that should be helpful for figuring out the small detail.

Is this gobli link from a few pages back claiming that SE changed a bunch of the 300-skill WS mods?

Maybe? It's titled "9 September 2010 patch"
EDIT: very likely

[Raelia]

Is this gobli link from a few pages back claiming that SE changed a bunch of the 300-skill WS mods?

Yeah. Looks like just the buffs to the elemental WSes. Herculean Slash was particularly crappy and thats about how much better it got. They were doing ~300 damage when first added and were buffed to 500-550. That's when I started swinging OA2-3/2-4 Scythe with Cosmos and MAB atma for spamming 1.2-1.4k Infernal Scythes with 86 damage Endarks, someone else made an OA2-4 GS and swung Herculean Slash with Beyond and MAB and was doing solid 1.7-2k (he had Ultimate and something else while I didn't). There was still that short period when they were first added that they just downright sucked.

Gave crit jobs a run for their money until Empy WSes came out. Fell Cleave probably didn't see any use until after that patch either.

[urat]

http://www.ffxiah.com/item-sets/240079

How does this look for a quickly tossed together Externator build? Im not sure on some slots as to what would be better, IE Austur's Pearl vs Aesir Ear Pendant.

Note: this would mostr likely be on dnc, outside of abyssea and sometimes inside abyssea. Specifically inside abyssea because it seems Externator -> Evisceration is darkness, which would probably do significantly more damage than a normal Evi->Evi with no SC, factoring in Dnc's dumb skillchain bonus traits.

Just wanna know what kind of gear I should collect together before the update comes out so I'm ready to zerg to my sexy new WS and start wrecking shit with it.

[Atoreis]

Calculating it exactly using only his highest damage Resolutions gives you a lower bound for the penalty (minimum penalty) that comes out around 30%. Realistically (because we were using his highest damage resolutions instead of average of 5-hit resolutions), the actual penalty is going to be larger than that. So it would be more accurate to say it is "at least" a 30% penalty. On BGwiki I have it as about a 1/3 penalty at the moment, but it really could be as high as 50%.

With 1015 attack and mob's 327 defense I'm at ~3.1 cRatio and I did WS with pDIF around ~2.218

2.218*(20/21)/1.2=~1.76
1.76/3.1=0.567

Its only a proof that penalty cant be higher than 44% but thats still a lot.

@Raelia That is exactly how I understand it too and thats why I was confused by what Yugl is writing. I expected more from a math person like you Yugl!