Final Form Magian Weapons

[ccl]

Yes, I was naked, my base str was 83.
Here are the number I got in ws gear +65 str and snow gorget

@102tp: 2193
@100tp: 2062
@102tp: 2137
@100tp: 2189
@100tp: 2115

Also I did the WoE path of the gkt, not the masamune, base damage is 103.

I have an ls event going on now, I'll post more test later if they are needed


Edit: If you need more sample or test on different mob, feel free to ask, I can test after my lsevent and post them later tonight or tomorow morning.

[FailureMidgard]

Yes, I was naked, my base str was 83.
Here are the number I got in ws gear +65 str and snow gorget

@102tp: 2193
@100tp: 2062
@102tp: 2137
@100tp: 2189
@100tp: 2115

Also I did the WoE path of the gkt, not the masamune, base damage is 103.

I have an ls event going on now, I'll post more test later if they are needed

So ~1835 average with all points (with gorget) before. If we want to call 2062 an outlier here, 2158 average now. Assuming capped critical pdif @ 3.15:

0 gear: 582.54
str gear: 685.08
Assuming double damage:
0 gear: 291.27
str gear: 342.54
Assuming 1.9 fTP:
0 gear: 153.3
str geaR: 180.28
+65 str is giving +27 damage

Assuming .85ish WSC, we're looking at a ~50% str modifier. No real way to double check this without a more concrete fTP value.

[Nameless]

Hmm... I got more like 75% STR modifier and either 1.9 or 2.0 ftp with double the weapon base damage.

[Nameless]

I did (122+103+(147 x 0.75 x 0.85)) x 2.1 x 3.1

And that gets extremely close to his numbers, tho slightly higher.

[FailureMidgard]

Hmm... I got more like 75% STR modifier and either 1.9 or 2.0 ftp with double the weapon base damage.

Wat? Double the weapon base damage? Why would you double that and not double everything ala an extra overwhelm term? Also, you did remember the WSC alpha right? Also, 1.9 and 2.0 ftp are very different, and there's absolutely no basis to using 2.0 at the moment...

I did (122+103+(147 x 0.75 x 0.85)) x 2.1 x 3.1

And that gets extremely close to his numbers, tho slightly higher.

Well, I forgot gorget for sure, but I still think assuming a 2.0 ftp is just silly, pointless, and guesswork at this point, as 1.9 seems "more accurate". I also have no idea why you doubled the base dmg and not fSTR or WSC. Relic procs definitely double/multiply fSTR at least.

[Nameless]

why would you double everything. I thought relic triple damage only account for the weapon base damage, which is the only thing that is tripled in the calculation. Or I could be wrong on how relic works....

I assume that they have the same mechanic, which could be an extremely bad assumption. So i dunno...

[FailureMidgard]

why would you double everything. I thought relic triple damage only account for the weapon base damage, which is the only thing that is tripled in the calculation. Or I could be wrong on how relic works....

You are wrong on how relics work, but they have to WSC term, so there's absolutely no precedent set for if that would double, but there is 1 for fSTR doubling. And you still have no basis in reality or numbers for a 2.0 ftp. Until we get more numbers, 1.9 or 3.4 are the safer ftp values to use. I'm not saying 2.0 isn't possible, it's within bounds, but the numbers right now do -NOT- point to it, so you shouldn't use it in your math...

[Byrthnoth]

You shouldn't be both taking an average and assuming capped critical pDIF.

(Base damage + fSTR + Mods)*fTP*pDIF*(2?) - No overwhelm on SA WS

83 STR, +0fTP = 1802
83 STR, +0.1fTP = 1862
148 STR, +0.1 fTP = 2193

(103 + 19 + Mods)*fTP*pDIF*(2?)

(122)*fTP*3.15*2 = 1802
fTP's max should be around 2.34, realistically it's probably 2.

If we assume 2:
(122+Mods)*2*3.15*2 = 1802
Mods = 21

If you stick Snow Gorget in at this point:
(122+21)*2.1*3.15*2 = 1891, a little higher than 1862, still within tolerance though

(122+Newmods)*2.1*3.15*2 = 2193
Newmods = 44

Now, even if STR is the only factor here, Mods/Newmods (21/44) should be about the same as STR/WSSTR (83/148), but it's off a bit. It could be that I'm a little off with our estimation of fTP, or it could be that fSTR wasn't capped in the original test. dSTR/4 is supposed to be an approximation of fSTR, but I don't know how far Masa got in confirming that it's correct. 19*4 is 76, which is kinda close to 83 if you squint. We also have reason to believe that he didn't approach the true max for the Snow Gorget test.

Anyway, yeah I get close to a 40-50% STR mod

Edit: If I was in a demandy mood, I'd ask for:
1) Do SA Tachi: Fudos on level 1 bees naked and keep track of your max damage. Do it until you get the same max damage twice. (n=30 minimum)
2) Do SA Tachi: Fudos on level 1 bees naked except Snow Gorget and keep track of your max damage. Do it until you get the same max damage twice. (n=30 minimum)
3) Do SA Tachi: Fudos on level 1 bees with your WS gear on and keep track of your max damage. Do it until you get the same max damage twice. (n=30 minimum)

With that info, we could likely nail down what the exact mods and fTP are.

[Nameless]

Wat? Double the weapon base damage? Why would you double that and not double everything ala an extra overwhelm term? Also, you did remember the WSC alpha right? Also, 1.9 and 2.0 ftp are very different, and there's absolutely no basis to using 2.0 at the moment...



Well, I forgot gorget for sure, but I still think assuming a 2.0 ftp is just silly, pointless, and guesswork at this point, as 1.9 seems "more accurate". I also have no idea why you doubled the base dmg and not fSTR or WSC. Relic procs definitely double/multiply fSTR at least.

Then I am wrong on that one. I thought I read it somewhere, think it was either Kirschy or someone. But take it away, you have relic so you definitely know more about that than me.

If the whole thing is doubled instead of just weapon damage, then it is definitely something around 40-50% str mod.

[FailureMidgard]

Kirschy is the one who told me fSTR doubles.

Anyway, I'm still not sure why you guys are hooked on a 2.0 ftp, no evidence outside of eyeballing for a round number supports it (and hi2u gekko ftp).

[arus2001]

Has the WoE sword popped up anywhere? Curious about CDS.

[Byrthnoth]

Kirschy is the one who told me fSTR doubles.

Anyway, I'm still not sure why you guys are hooked on a 2.0 ftp, no evidence outside of eyeballing for a round number supports it (and hi2u gekko ftp).

If you go lower, it doesn't fit his Snow gorget test well. If you go higher, it doesn't fit the mods well. I don't see a reason to go past one digit with an n of 4.

[FailureMidgard]

If you go lower, it doesn't fit his Snow gorget test well. If you go higher, it doesn't fit the mods well. I don't see a reason to go past one digit with an n of 4.

1.9 is the best fit for his snow gorget test. You're either throwing out the 1 number -you- want to call an outlier, or you're including numbers 10 TP higher on a damage varies with TP ws when we don't know how much by. It's going to be at least a 2% increase in damage (bare minimum imo), more likely closer to 4-5%, implying both of those 115ish TP tests fit better in line with your outlier than the rest of the data.

[FailureMidgard]

Fixing for gorget...

So ~1835 average with all points (with gorget) before. If we want to call 2062 an outlier here, 2158 average now. Assuming capped critical pdif @ 3.15:

0 gear: 582.54
str gear: 685.08
Assuming double damage:
0 gear: 291.27
str gear: 342.54
Assuming 1.9 fTP, 2.0 post gorget:
0 gear: 145.85
str geaR: 171.27
+65 str is giving +25.5 damage

Assuming .85ish WSC, we're looking at a ~45% str modifier. No real way to double check this without a more concrete fTP value.

Still 40-50%, again, 1.9 is a guess at fTP though, though I feel far more valid with current information than 2.0, and WSC at 85 is still unknown as well...40-50 is reasonable enough though, IMO.

[Byrthnoth]

I don't feel it's worth arguing about, because I don't think it is possible to know from the data we were given. The 1600s WS is actually an outlier if you do the significance test, even though you only have 4 data points (3 degrees of freedom in t-test). I don't know why, but <3.0pDIF values are much less common than >3.0 pDIF values. I don't actually know that it's 3.0 as I never thought to check when I was testing, but it's about there.

Also, I think your equations still assume 3.15 pDIF after averaging values (some of which obviously have less than 3.15 pDIF, because if they all had 3.15pDIF they'd be the same number). Ultimately though, we think that fTP is somewhere near 2 at 100TP (could be 1.9, could be 1.875), there is actually a "Double Damage mod", and it's somewhere between 40 and 50% STR, probably with no other stat mods. The only way we'll get much better than that is if he does a lot more testing and nails down max damage values for each condition.

[Motenten]

w/str & gorget:

@102tp: 2193
@100tp: 2062
@102tp: 2137
@100tp: 2189
@100tp: 2115

2062 *has* to be below 3.0, since the ratio between that and the highest result (2193) is greater than 1.05 (it's 1.0635).

The ratio between the second lowest, 2115, and the highest, 2193, is 1.037, within the bounds of 3.0 to 3.15. If 2115 was at exactly 3.0, the max would be 2220. Given the two high results of 2189 and 2193 being fairly close to each other, it's likely that 3.15 is roughly 2200.


Re: fTP:

w/gorget:

1817 @104tp
1834 @100tp (sekkakoki)
1829 @ 104 tp
1862 @ 104 tp

In this case the ratio between lowest and highest is 1.025. That means the highest limit of 3.15 is about 1907, but more probable to be around 1885.

wo/str & gorget:

1787 @116 tp
1669 @104 tp
1766 @104 tp
1756 @114 tp
1802 @104 tp

Using the 104 TP results only, 1669 *has* to be below 3.0, as the ratio between that and 1802 is 1.08. Given that 1802 is higher than both the ~115 TP results, and that there's insufficient additional data (the ratio between the other two points is only 1.02), I'm going to (tentatively) treat it as the 3.15 marker. The max possible value for 3.15 is 1854.


In any case, that means the gorget increased damage by ~82 points at 3.15. That's 4.6% of the non-gorget damage, which would put the base fTP at 2.17. In conjunction with FM's estimates of 1.9 fTP, I think 2.0 is a safe starting point.


More generally, using the w/gorget results, a damage increase of 90-91 would indicate a 2.0 base fTP. A damage increase of 94-95 wouldd indicate a 1.9 base fTP.

If cap is 1907, 1816-1817 without the gorget would be 2.0 fTP, 1812-1813 would be 1.9 fTP.
If cap is 1885, 1794-1795 without the gorget would be 2.0 fTP, 1790-1791 would be 1.9 fTP.

Since the tests without the gorget reached 1802, we know the cap with the gorget has to be higher than 1885 for either fTP value. The lowest the w/gorget result could cap at would be 1892 with 2.0 fTP and 1897 with 1.9 fTP, and we already know it can't be higher than 1907 (which would yield a 1.7 fTP if 1802 was the cap without the gorget).

Further, with 1907 being the cap on the high end, we know the cap without the gorget can't be higher than 1817 (assuming we don't allow for fTP higher than 2.0).

Given that there's at most a 5 point difference in damage between the two fTP values, insisting that there's no possible way for the fTP to be 2.0 seems rather premature at this stage.



If we continue, using the 2.0 fTP and assuming caps of 1810 and 1900 for without and with gorget, and getting 2200 as the max with +65 str, then we can make a rough WSC estimate from that.

+300 damage from 65 str while influenced by 2.1 fTP and 3.15 pDif yields +45 damage. If that is itself also under a 2.0 multiplier (from the weaponskill description), that means about +23 damage. Factor in 0.85 alpha and we get ~+27 from 65 str, or a WSC multiplier of ~40%.


edit:

1.9 is the best fit for his snow gorget test. You're either throwing out the 1 number -you- want to call an outlier, or you're including numbers 10 TP higher on a damage varies with TP ws when we don't know how much by. It's going to be at least a 2% increase in damage (bare minimum imo), more likely closer to 4-5%, implying both of those 115ish TP tests fit better in line with your outlier than the rest of the data.

If the higher TP weaponskills increased damage by up to 2.3%, they'd still fall in line with 1802 as a max, and be within the 3.0 to 3.15 range, not below it with the outlier. If the TP increased damage by more than that, then yes, they'd be in the outlier region as well.

[FailureMidgard]

Given that there's at most a 5 point difference in damage between the two fTP values, insisting that there's no possible way for the fTP to be 2.0 seems rather premature at this stage.

Seeing as I specifically said 2.0 is possible, but there was no specific data pointing it to being a better option than 1.9, I have no earthly idea where you came up with someone insisting there's no possible way for 2.0 to be correct.

If the higher TP weaponskills increased damage by up to 2.3%, they'd still fall in line with 1802 as a max, and be within the 3.0 to 3.15 range, not below it with the outlier. If the TP increased damage by more than that, then yes, they'd be in the outlier region as well.

2% increase (actually a .04 increase in fTP) would be in line with Gekko's progression from 1.56 to 1.87, which is one of the lowest gain WSes out there from 100>200 TP (20% increase, I think only Steel Cyclone is lower). Most the lower ones that only have a .25 increase in pdif start at 1.0, giving them a 25% increase in damage done. Some WSes go as high as +.5 pdif. So I'm not sure why you chose 2.3%. Furthermore, when they are lower than the max @ 100-102 TP, it's pretty obvious that lowering the damage by 2-5% would keep them within your max, my point is they would drop under 3.0 pdif, meaning they should be excluded from any average as sure as the 1669 result. If that was +2.3%, 1756 would drop it 1 damage under 3.0 ftp, and that's assuming 1802 was -exactly- 3.15, which seems like quite the assumption at this point in time, about as much as +2.3% being what's gained with 10 additional TP.

My main point here is people are doing math -trying- to make 2.0 fit, rather than trying to figure what best fits with the current data. Again, 2.0 is definitely possible, but there is no sound mathematical reason to be saying it fits better than something lower. IF you want to exclude the result that is obviously below 3.0 pdif, you cannot include the 114 and 116 numbers, making the ftp significantly higher than you are saying, and really making this all even more inconclusive. I will admit I can personally make this data fit near perfectly with 2.0, or 1.9 or 1.875, which is what is currently occurring in the thread. Either use the data or drop it. 1 point is obviously an outlier, 2 points obviously are off by anywhere from 2-5%, and past that, other points very well could be under 3.0 pdif, you are assuming that the max out of 4 WSes is indeed the max, which is far beyond a bad assumption.

I agree with Byrth that this is pointless without more data, but not once have I said 2.0 isn't possible, I'm saying it's skewing data to try to match what you want it to be.

[Kirschy]

Kirschy is the one who told me fSTR doubles.

Yea I remember the old Apoc numbers by heart since I did so much testing /w it.

103 (DMG) + 19 (fSTR) = 122
122 * 3.0 pdif = 366 capped crit damage
366 * 2 = 732 capped crit + relic proc damage.

One of the relic dagger BG ppl (I forget who) tested SA (which adds your DEX to the base damage), and noticed only the DMG + fSTR parts got doubled, not the entire damage. This makes the SA equation on THF main look like:

{[(DMG + fSTR)*RELIC_PROC]+DEX}*pdif

Weapon Skill Damage calculations place the WSC where the DEX would be. So even though relic procs can't occur during WS, it seems unlikely it would affected the entire base damage. That all said the occasionalli deals double damage is better off tested /w Death Penalty. The AM is stated the same way, and ranged shots cap at 3.0 pdif still, so finding out which parts of the base damage are doubled should be no problem~

EDIT: I'm wrong, and WSC would come before the DEX portion. (WSC is obviously affected by fTP during WS, but the DEX term from SA isn't.) Which makes my argument invalid . Still, someone go test Death Penalty WS /w AM!

[Seol]

Has the WoE sword popped up anywhere? Curious about CDS.

Rethi from Apathy (in Shiva now) has a fully unlocked 85 Almace (With aftermath), but he/she/it is JP so I have no means to communicate.

[Byrthnoth]

Other interesting tests:

1) Its description does not imply it has any TP mods. Do 300TP Tachi: Fudos do the same damage as 100TP Tachi: Fudos? It should be obvious and take about one WS to test.
2) Does it "Ignore Defense" like YGK? I'm not really sure how to test this exactly, but I'm sure someone else here could tell you. Either way, you'll be able to get a pretty good idea by comparing the ratio of Gekko Damage / Fudo Damage on HNMs to Gekko Damage / Fudo Damage on sneak attack'd level 0 monsters. If the Ratio is the same, they ignore comparable amounts of defense.