Sea Gorgets

[Yutuyu]

Explain to me how pDIF was not kept constant. pDIF was *obviously* capped out all times, because this was done on level 1 mobs. Every mob tested on clearly had the same pDIF.

As for the fTP, they're two totally difference tests, essentially. Which is why I listed the 108% WS and the 100% seperately.

[Yutuyu]

Obviously the tests aren't perfect, I'm not saying they prove anything, infact I've been saying all along that they're not conclusive. I'm saying they're enough to show that the fTP +.1 thing is probably not the case.

Since you seem to know a lot Amele, what sort of test would you suggest in order to determine if the fTP is correct with a fair degree of confidence?

[Ferien]

Explain to me how pDIF was not kept constant. pDIF was *obviously* capped out all times, because this was done on level 1 mobs. Every mob tested on clearly had the same pDIF.

As for the fTP, they're two totally difference tests, essentially. Which is why I listed the 108% WS and the 100% seperately.

pDIF is a random number whose constraints are based on the ratio between your attack and a mob's defense. Your ratio may have been capped, but your pDIF is not, unless you're only taking the consistently highest number. Once again, do your research before you try to test something.

[Amele]

basically what you're doing: I'd use the level 1 mobs like you were;

I'd do enough weaponskills to determine either the mode or the maximum, then compare just those values. I'd also use only TP attacks at a given amount (100 should be good) I'd probably use a weapon with very weak stats to max my fstr (like say, a tachi or a lotus katana ).

in fact, I'd probably /rng and use sidewinder so I wouldn't have to worry about TP varying my damage, and can get much more consistent numbers.

sort of like this I guess.

here is a nagamaki test about PDIF effect on gorgets too, btw.

[VZX]

Oh God,
Go back to melee pDIF and learn it properly please before you make any more post here.

People still research about how pDIF average behave in different cRatio value range.

Anything you plunge into base damage equation will give you 1 number
Anything you plunge into pDIF function will give you a number range (i.e. 1.6~2.4, means it can be 1.667, can be 2.349, can be 2.053, can be 1.945 etc etc), so It's not feasible you claim 8% increase average and say those who said "gorgets increase on fTP is +25/256" wrong

melee pDIF is never a single fucking number everytime you hit a monster, it's a range

[archibaldcrane]

Please, just listen to VZX. He's correct. I know the number crunching well enough to understand why, but he explained it better than I ever could. It's fTP+25/256 (0.1 for almost perfect shorthand).

I personally am not discounting a +pDIF also, but I just sincerely doubt it, and I've seen no properly executed evidence to say otherwise. Such evidence could be obtained however, if a +pDIF value is included, it would just be more of a pain to do so. And of course, you can't really prove it's not there, only build overwhelming evidence that it isn't.

[Tankjr]

I'm a little confused after reading some of these posts. Maybe someone can help clear this up.

I read this.

pDIF is a random number whose constraints are based on the ratio between your attack and a mob's defense. Your ratio may have been capped, but your pDIF is not, unless you're only taking the consistently highest number. Once again, do your research before you try to test something.

And this.

Oh God,
Go back to melee pDIF and learn it properly please before you make any more post here.

People still research about how pDIF average behave in different cRatio value range.

Anything you plunge into base damage equation will give you 1 number
Anything you plunge into pDIF function will give you a number range (i.e. 1.6~2.4, means it can be 1.667, can be 2.349, can be 2.053, can be 1.945 etc etc), so It's not feasible you claim 8% increase average and say those who said "gorgets increase on fTP is +25/256" wrong

melee pDIF is never a single fucking number everytime you hit a monster, it's a range

I was under the impression that SA+WS was done to cap your pDIF as a control. I believe he said he was using SA+WS during his tests, which according to VZXs info would push his pDIF to cap. Am I reading the original information wrong, is it obsolete?

The info I'm referring to can be found here:
http://www.freewebs.com/vzx-01/index.htm

[Amele]

I'm a little confused after reading some of these posts. Maybe someone can help clear this up.

I read this.

pDIF is a random number whose constraints are based on the ratio between your attack and a mob's defense. Your ratio may have been capped, but your pDIF is not, unless you're only taking the consistently highest number. Once again, do your research before you try to test something.

And this.

Oh God,
Go back to melee pDIF and learn it properly please before you make any more post here.

People still research about how pDIF average behave in different cRatio value range.

Anything you plunge into base damage equation will give you 1 number
Anything you plunge into pDIF function will give you a number range (i.e. 1.6~2.4, means it can be 1.667, can be 2.349, can be 2.053, can be 1.945 etc etc), so It's not feasible you claim 8% increase average and say those who said "gorgets increase on fTP is +25/256" wrong

melee pDIF is never a single fucking number everytime you hit a monster, it's a range

I was under the impression that SA+WS was done to cap your pDIF as a control. I believe he said he was using SA+WS during his tests, which according to VZXs info would push his pDIF to cap. Am I reading the original information wrong, is it obsolete?

The info I'm referring to can be found here:
http://www.freewebs.com/vzx-01/index.htm

yutuyu wasn't or he wouldn't have speculated (crit?) after several of his attacks.

using SA still doesn't necessarily cap your PDIF, btw.

[VZX]

The rule for melee cirt is pDIF +1.0

So when your cRatio is capped, you can expect pDIF ranging from 1.6~2.4.
But since it crits, you expect the number ranging from 1.6+1.0 to 2.4+1.0, but since melee pDIF crit capped on 3.0, the actual range is 2.6~3.0, it still gives you a number range, unlike ranged pDIF where you only got single number (3.0) when you reach the cRatio cap. You can easily see this by shooting lowest level enemy with your 75 RNG.

[archibaldcrane]

The rule for melee cirt is pDIF +1.0

So when your cRatio is capped, you can expect pDIF ranging from 1.6~2.4.
But since it crits, you expect the number ranging from 1.6+1.0 to 2.4+1.0, but since melee pDIF crit capped on 3.0, the actual range is 2.6~3.0, it still gives you a number range, unlike ranged pDIF where you only got single number (3.0) when you reach the cRatio cap. You can easily see this by shooting lowest level enemy with your 75 RNG.

To sum this up, people force crits with SA for melee testing because it gives a range that is half as large, making it much easier to identify the 3.0 hits (the highest ones) in order to find "fixed" pDIF output.

[Yutuyu]

I gave this some thought and reread the formulas, and came to a couple of conclusions.

The main reason for the confusion was a terminology mistake. Whenever i said my pDIF was capped I meant the cRatio was. The pDIF was of course, still random. Because the cRatio was capped at 2 for all tests, the same randomization was employed.

Also it doesn't matter if the distribution of the pDIF is normal or not. And here's why:

I calculated the average over several weapon skills, since I kept my STR and D constant, and I was using a WS which uses only an STR modifier, WSC, fSTR, and D were constant. I was fighting mobs where fSTR & cRatio were capped as well.

Damage = (D + fSTR + WSC)*(fTP)*(pDIF)

I was averaging the damage between several weapon skills like so.

Average Damage = (Damage1 + Damage2 + Damage3 + Damage4 + ... + DamageN)/N

Substitute the individual damages...

Average Damage = ((D1+fSTR1+WSC1)*(fTP1)*pDIF1 + (D2+fSTR2+WSC2)*(fTP2)*pDIF2 + ... + (DN+fSTRN+WSCN)*(fTPN)*pDIFN)/N

Because the D fSTR WSC and fTP are all constant, they can be factored out.

Average Damage = (D+fSTR+WSC)*(fTP)*(pDIF1 + pDIF2 + pDIF3 + ... + pDIFN)/N

You'll notice that (pDIF1 + pDIF2 + pDIF3 + ... + pDIFN)/N is just the sample mean of the pDIF distribution. Unless the gorget in someway effects pDIF, the mean will (of course) not change. The distribution is the same in both cases.

Average Damage = (D+fSTR+WSC)*(fTP)*u (u instead of the greek letter typically used to symbolize the sample mean, since I can't type it)

Which is quite obviously true as N approaches a sufficiently large number (thus the sample mean approaches the true mean). What is sufficiently large depends entirely on what the distribution of pDIF is, so a confidence interval cannot be developed unless the distribution of pDIF is known.



In short if you're trying to figure out how much damage a gorget raises your weapon skills by, you need only figure out how much damage your average weapon skill is raised by. There's nothing false about this methodology. With a large enough sample size it's a perfectly valid test. I might not've explained it properly originally, but there's nothing wrong with my results anyway, other than that my samples were too small.

The test with sidewinder is rather conclusive, considering the damage was static, but you cannot assume that simply because it is a fTP +.1 mod for a ranged weapon skill which has damage that does not vary with TP, then it's an fTP +.1 mod for a melee weapon skill which does vary with TP.

[Amele]

You'll notice that (pDIF1 + pDIF2 + pDIF3 + ... + pDIFN)/N is just the sample mean of the pDIF distribution. Unless the gorget in someway effects pDIF, the mean will (of course) not change.

except that your sample size is too small, so you can't justify that your difference in 'PDIF mean' isn't the result of error due to small sample size.

you can't even identify the mode damage in your tests, which makes it painfully obvious that there weren't enough done.

The test with sidewinder is rather conclusive, considering the damage was static, but you cannot assume that simply because it is a fTP +.1 mod for a ranged weapon skill which has damage that does not vary with TP, then it's an fTP +.1 mod for a melee weapon skill which does vary with TP.

ockham's razor. Why would the function of fTP modifiers be different for ranged attacks versus melee attacks on the same piece of gear when the application of fTP between the two classes of weaponskills is identical? You're just being stubborn at this point.

[Yutuyu]

3 indepent 5-sample tests disagree with the given formula. That's enough reasonable doubt to question the conclusion, especially when I've seen no data which agrees with it done on melee weapons.

Occam's razor is not applicable here. The only time occam's razor IS applicable is when an additional premise is added without any purpose or reason and thus can be "cut away." There's all the reason in the world to believe that the gorget may work differently for different kinds of weapon skills and different damage types, considering other things DO. Futhermore, I am proposing this to fit my (thus far inconclusive) data.

[Yutuyu]

I did more tests and the simple fact of the matter is, I'm wrong. I tried with Hagun and with Ito, and the +.1 fTP mod is definitely the case. Sorry for wasting everyone's time in a stupid argument (without really knowing what I was doing fully at first especially).

[Max™]

*claps*

Good man, "I screwed up, my bad."

If only we could figure out how to bottle that and spread it around.

[Narse]

Sorry for the 'somewhat' necro-bump, but I don't feel my questions are worth making a new topic for, hopefully someone can help me to understand. I've tried searching forums for the last hour with not much luck. I've noticed in this topic that the modifier from gorgets is 'fTP +.1', which, to my understanding would be the same as TP bonus +10%, no? If so, why are the gorgets listed as "Increases WS accuracy and damage" when TP modifier would generally only increase one or the other? Wouldn't it make more sense to list it as something like "TP bonus" like it is on weapons like Hagun?

I guess my real question would be, What is the effect on Sea Gorgets? Does it give a TP bonus, or is it an actual increase in damage calculation/accuracy calculation?

I'd been told a long time ago that it was a 10% modifier to the damage and accuracy of the WS. After I came across info concerning the effect of the sea gorget not working beyond the first hit of the WS, I started to question what I knew about the Sea gorgets, but there's so much info out there, a lot of which I don't understand, that I am having trouble figuring out exactly what I'm dealing with here. Any help would be appreciated.

Also, if this is covered in some other topic I had a hell of a time finding it using the search feature. I gave up after an hour.

[Whatinthe]

I've noticed in this topic that the modifier from gorgets is 'fTP +.1', which, to my understanding would be the same as TP bonus +10%, no? If so, why are the gorgets listed as "Increases WS accuracy and damage" when TP modifier would generally only increase one or the other? Wouldn't it make more sense to list it as something like "TP bonus" like it is on weapons like Hagun?

I guess my real question would be, What is the effect on Sea Gorgets? Does it give a TP bonus, or is it an actual increase in damage calculation/accuracy calculation?

No, no, no. fTP is component in calculation for WS damage. For some WS whose "Damage varies with TP", fTP has different multiplier at different TP (1.5 100%, 2.0 200%, 2.5 300%, etc.) For other WS, such as Rampage, fTP is constant across all TP because it's "Critical chance varies with TP" - in this case, fTP is 0.5 regardless of your TP.

This fTP is multipled for the first hit only for physical WS (elemental WS have their own little convulted calculation involving INT and enemy INT and resistance and whatnots) by what's commonly referred to as WSD - Weapon Skill Damage. WSD is a combination of base D, your fSTR bonus, and the WSC, the "stats bonus" of a WS (STR10%, etc.) multiplied by a level correction factor.

[Narse]

I see. So the damage increase seems to not be as drastic as I once thought, but still should be the best item for the slot on 1-hit WSs, at least for damage. Any ideas on how they effect accuracy?