[milich]

hello. i recently made this post on alla: http://ffxi.allakhaz...269997124935765

it was in response to a view i've been hearing that sea gorgets add +10ACC to the first hit of a WS only (it's always been known that the .1 fTP bonus only applies to the first hit, since fTP only ever applies to first hits anyway). the idea that my wearing shadow gorget for asuran fists was actually dropping ACC struck me as highly dubious, considering my WS numbers and average number of landed hits seemed to go up considerably after i got shadow gorget. for those of you averse to following links, here's the relevant information from my post, where i listed numbers from a recent parse that--i think at least--contradict the idea that shadow gorget a) only adds 10ACC (i always thought it was 10%) and that the ACC only applies to the first hit:

here's numbers from my last meripo pt (nyzul birds, 2BRDs, ate about 36 mithkabobs):

202 fights
ACC:87.02%
dmg/hit: 75.48
WS average: 758.23

758.23/75.48 = 10.05. just eyeballing that, those don't look like 87%ACC WS numbers, but i'd need to know the VIT, DEF, and level of greater colibri to estimate more precisely. oh wait...

this says: Greater Colibris Level:81-82 Def:322-336 VIT:67

well, i was TPing in about 100STR (with about 80VIT), so my fSTR was capped. i had about 500ATT, so my expected melee average would be about:

(53 + 10) * (500/330-.05*6) = 76.55

well, that looks right.

my first hit in AF should then (with shadow gorget, which i used exclusively) be:

(53 + 10 + 14) * 1.1 * (1.21515) = 102.92

and every next hit would be: 93.56

so, my average 8hit WS should be: 753

---

if 2ACC = 1%ACC, and all shadow gorget does is add 10ACC to the first hit, my swapping it in for c.chain gives me 89.5%ACC on the first hit, and 84.5%ACC for the later, unaffected hit.

the fact that my average WS dmg was a little above what i could expect from landing 8 hits with average pDIF, i'm VERY skeptical that i was dropping ACC for those last 7 hits. 85%ACC gives me a 27% chance of landing all 8 hits. it gives me a 32% chance of landing the first 7. i'm pretty convinced that a) shadow gorget adds MORE than 10ACC, and the ACC bonus affects every hit.

if someone with a little more mathematical acumen (or at least time, i have to go) can project my WS ACC from my average dmg, feel free. it looks to me like 95%, or at least higher than my melee ACC.

i attribute my ave WS being higher than that predicted by mob stats vs my stats to the occasional boost and occasionally having warcry up for WS from the WAR or NIN. even if such things at all taint this information, it seems really hard for me to look at those numbers and believe that i had less ACC on WS than for melee, because i can't see how that number says "<87%ACC".

still, it could be that i'm not interpreting this data as effectively as others could, so i'm posting it here to see if anyone else has opinions on it (or has more testing/data relevant to the topic).

-----
edit: i got the greater colibri stats from here: http://ffxi.allakhaz...178779245579474 . the only reason i needed really them (as opposed to just projecting WS DMG from my own recorded average dmg/hit) was to find my fSTR and therefore to see what effect the lolmods had on AF.

if anyone wants to speak to any of this but is unfamiliar with the dmg equations, you can read all about them...
here: http://wiki.ffxiclop...Physical_Damage + http://wiki.ffxiclop...on_Skill_Damage
and here: http://www.freewebs.com/vzx-01/

[Gengu]

idk, someone get aurik on this shit

[Cream Soda]

Gengu :
idk, someone get aurik on this shit

[Tyche]

milich :
hello. i recently made this post on alla: http://ffxi.allakhaz...269997124935765

it was in response to a view i've been hearing that sea gorgets add +10ACC to the first hit of a WS only (it's always been known that the .1 fTP bonus only applies to the first hit, since fTP only ever applies to first hits anyway). the idea that my wearing shadow gorget for asuran fists was actually dropping ACC struck me as highly dubious, considering my WS numbers and average number of landed hits seemed to go up considerably after i got shadow gorget. for those of you averse to following links, here's the relevant information from my post, where i listed numbers from a recent parse that--i think at least--contradict the idea that shadow gorget a) only adds 10ACC (i always thought it was 10%) and that the ACC only applies to the first hit:

here's numbers from my last meripo pt (nyzul birds, 2BRDs, ate about 36 mithkabobs):

202 fights
ACC:87.02%
dmg/hit: 75.48
WS average: 758.23

758.23/75.48 = 10.05. just eyeballing that, those don't look like 87%ACC WS numbers, but i'd need to know the VIT, DEF, and level of greater colibri to estimate more precisely. oh wait...

this says: Greater Colibris Level:81-82 Def:322-336 VIT:67

well, i was TPing in about 100STR (with about 80VIT), so my fSTR was capped. i had about 500ATT, so my expected melee average would be about:

(53 + 10) * (500/330-.05*6) = 76.55

well, that looks right.

my first hit in AF should then (with shadow gorget, which i used exclusively) be:

(53 + 10 + 14) * 1.1 * (1.21515) = 102.92

and every next hit would be: 93.56

so, my average 8hit WS should be: 753

---

if 2ACC = 1%ACC, and all shadow gorget does is add 10ACC to the first hit, my swapping it in for c.chain gives me 89.5%ACC on the first hit, and 84.5%ACC for the later, unaffected hit.

the fact that my average WS dmg was a little above what i could expect from landing 8 hits with average pDIF, i'm VERY skeptical that i was dropping ACC for those last 7 hits. 85%ACC gives me a 27% chance of landing all 8 hits. it gives me a 32% chance of landing the first 7. i'm pretty convinced that a) shadow gorget adds MORE than 10ACC, and the ACC bonus affects every hit.

if someone with a little more mathematical acumen (or at least time, i have to go) can project my WS ACC from my average dmg, feel free. it looks to me like 95%, or at least higher than my melee ACC.

i attribute my ave WS being higher than that predicted by mob stats vs my stats to the occasional boost and occasionally having warcry up for WS from the WAR or NIN. even if such things at all taint this information, it seems really hard for me to look at those numbers and believe that i had less ACC on WS than for melee, because i can't see how that number says "<87%ACC".

still, it could be that i'm not interpreting this data as effectively as others could, so i'm posting it here to see if anyone else has opinions on it (or has more testing/data relevant to the topic).

-----
edit: i got the greater colibri stats from here: http://ffxi.allakhaz...178779245579474 . the only reason i needed really them (as opposed to just projecting WS DMG from my own recorded average dmg/hit) was to find my fSTR and therefore to see what effect the lolmods had on AF.

if anyone wants to speak to any of this but is unfamiliar with the dmg equations, you can read all about them...
here: http://wiki.ffxiclop...Physical_Damage + http://wiki.ffxiclop...on_Skill_Damage
and here: http://www.freewebs.com/vzx-01/

The whole "27% chance to land all 8 hits" (.85^8) is incomplete because of the ACC bonus you will get from TP. Asurans' ACC varies by TP. They ran extensive testing with ragining fists on a BG thread that i linked in the "OK, WTF" thread. Trying to calculate it with Asurans is going to skew results depending on how much TP you had when the WS went off. Too many variables and a somewhat small sample size to develope any type of conlcusion from the data you showed. Was focus ever on? Was madrigral on all the time (You said it was a 2BRD party), etc?

[milich]

focus was used every 5 min, madrigal was on full-time (marchx2, minuet, madrigal) and i never store TP over 150% (generally never more than a punch over 100% though i don't like to waste it on overkill). 202 fights isn't exactly a small sample size either... you actually think that the ACC from TP raises people's AF to sushi-like accuracy when they have around 85% ACC? focus and madrigal aren't a factor, considering they affect melee accuracy just as much as WS.

if it helps, i did 150 WS for a total of 113,734 dmg. my average dmg/punch in AF should have been about 94 (this is backed up by what my average WS dmg actually was). 113,734/94 = 1209.94. 1209.94/150 = 8.06.

the ACC modifier from TP was never that powerful before i had shadow gorget...

---

i have another parse vs mamool at nyzul where i had red curry up from beforehand and ended up with 69.29%ACC. numbers from this one:

ACC: 69.29%
dmg/hit: 80.57
WS average: 684.33

i can tell you, i don't have enough STR and VIT to make that WS average not pretty damn high for someone with 70%ACC (though i can't elaborate with the same precision as talking about birds because i don't know the stats of mamool). if it matters, that party spanned about 231 battles (parser includes pets though) and in it i did 101 WS, for 69117 total dmg. i landed 2085 melee hits for 175,432 dmg.

if people are convinced by that raging fists test, please consult parses with and without sea gorget for average WS dmg. i personally don't swap on any extra ACC besides shadow gorget either... these two parses i'm referring to really don't make sense if gorgets drop ACC in relation to c.chain/PCC... unless the ACC up from asuran fists is a lot more powerful than i think it is... and more powerful than it was before i got shadow gorget.

[Tyche]

Just too many variables...You are saying...sometimes i went at 100% TP, sometimes 150% TP, sometimes focus was up for WS, sometimes it wasn't, all this on a 8 hit WS with a ACC modifier for TP. Just not controlled. You want answers, set up a more controlled test similiar to how naga set it up.

[milich]

tyche, listen to what you're saying. if that's too many variables, how do parses mean anything at all? or are you one of those people who don't believe parses mean anything at all-_-;.

focus is not going to bleed over and allow WS dmg that implies i'm hitting 7-8 AF more often than not. it's not like i'm landing 12 hits on my focus WSs.

edit: what WOULD invalidate what i'm saying completely, is if someone can point out some way in which i'm making a big error in interpretation regarding what my WS average implies about my WS ACC. further edit: this isn't a challenge, but a request... coming from the fact that i know i'm not as good at statistics as some other monks out there (cough, aurik).

gyah, one more edit. a tyche style "you're not considering how X invalidates your results completely" could sway me from at least believing that these parses prove anything (though i'm still interested in constructing a viable controlled test for AF ACC that won't take 10 hours). it's just that "ACC modifier on AF is giving you sushi WS ACC" doesn't really fly... we've been eating meat for a long time, and you don't hear people raving about the AF ACC mod. if anything, people have always scoffed at those who wouldn't eat meat *because the ACC hit their WS took*.

[pathwriter]

Parses really don't mean much.

[milich]

Parses really don't mean much.

eh, they don't mean much when they're used to prove "check it out, DRK/SAM > ridill WAR lololol" but they're useful for things like checking crit%, gauging how much ACC is necessary in certain camps, etc. it seems to me that they're also useful for gauging potential disparity between normal ACC and WS ACC, unless there's a real factor i'm missing that would invalidate such observations.

[Tyche]

No need to get edgy dude...you asked for input on your tests, don't get offended when someone points out the flaws. Seriously, don't ask for inoput if you are afraid of criticism. It's too many varbiables to test anything conclusive regarding a gorgets effect on WS re: the ACC. You sound intelligent enough to set up a controlled test for more conclusive results, don't insult your intelligence by trying to go off these numbers.

Edit: Sorry, missed his above edits. Seriously man, throwing out insults and putting words into my mouth that i never said is absolutely egregious. I never said the AF modifier would give you sushi style results, but add focus to a WS some of the time and varying ACC modifiers from TP, that makes your tests a little inconsistant. AF is a really bad WS to test for ACC is all I was saying. I'm trying to help you setup a test to obtain the answers you want, and you act childish. C'mon man, I'm sure you are better than that.

[pathwriter]

milich :
it seems to me that they're also useful for gauging potential disparity between normal ACC and WS ACC, unless there's a real factor i'm missing that would invalidate such observations.

Asuran Fists: Accuracy varies with TP

[milich]

you're misinterpreting my tone. i don't mean to insult you, and i'm not putting words in your mouth.

the fact is, those parses indicate sushi-style ACC on my WS (words came from my mouth, not yours). focus can't do that, because it's only up for 2min, and the AF TP mod can't do it because it isn't strong enough (not to mention that at least 80% of my WS were below 110% TP). if AF could land more than 8 hits, my results would be rendered suspect by things like focus, since the WS average seems to imply 7-8 hits landing consistently, and focused WSs landing like 12 hits would be the obvious reason why i get such an average even with bad ACC. however, i'm getting what looks like an average of 7-8 hits on an 8-hit capped WS... 2min of all-8 hits and 3min of 5-6 wouldn't yield that.

when i edited in that "tyche-style objections" or whatever would invalidate the results, what i meant was that i welcome considerations like yours to throw off my confidence in what those parses show. i just don't think the invalidating variables are there... madrigal affects all ACC equally, as does focus when you come down to it, so my melee ACC should be reflected in my WS ACC. if anything, i might be drawing invalid inferences from valid numbers (from false assumptions about what math to do), but i still can't see what's wrong with the "test" itself.

you're right that i should construct a test (which could more or less be done vs colibri or crabs, though it would take forever to get a sample size, and the amount of TP at which i WS would have to be controlled, which would also be a pain-_-;;). i don't have the time presently, and without any reason to be suspicious of my parses (like, say, someone who doesn't swap ACC on for WS who can show parses where their WS averages "seem to imply 7-8 hits despite 85~% melee ACC), i'm using them as data.

looking over my posts i don't see anything insulting. i'm not coming here to do the "hey guys i need advice (gets advice) you guys don't know anything anyway!" thing. thanks for the comments on the test; my replies were just my thoughts on those comments. i'm giving you more respect than the average "lolno" response;;.

edit: path, you really think that someone with 70%ACC will get what looks like sushi ACC WS average because oops they waited until 150% TP to WS that mamool? because that's what raising that objection amounts to saying... unless i'm wrong about what the WS average implies.

[pathwriter]

The problem is that you're involving too many variables. If you're going to test for Accuracy, you need to minimize all variables that affect Accuracy. Therefore, no Madrigal, no Focus, no Accuracy varies with TP weaponskills, and consistent gear across the full test.

You're speaking about elemental gorgets - you just need data from weaponskills. Normal melee accuracy is a red herring.

[Tyche]

You implied that i completely dismiss parsers altogether, which is quite insulting, then you throw a "style" to my critique. Asuran fists variable TP invalidates your results. The fact you are comparing melee ACC to WS acc is invalidating your results. That is TWO variables. The WS has an ACC modifier, whether it be minimal modifier or not, it's a variable, which is something you do not want in testing. This is really really easy to test, and you don't even need a parser. Go out and find a level 80ish mob, or at least something you aren't goig to hit the ACC cap on. Crabs would be bad obviously. Use raging fists (because its a damage modifier), and count your TP return to see how many hits landed. Do this several hundred times without a neck piece, with a PCC, and with a gorget. Compare results. Don't use a BRD, Focus, whatever. Make the only variable in the entire test be the neck peice. Voila, answers.

[milich]

but there is a style to your critique:). it's attacking the data instead of the results. it's not any less valid than any other style, and therefore not insulting. edit: about dismissing parses altogether, the variables both of you are bringing up would have to be having really phantom effects on my results. focus and madrigal affect all ACC equally (especially over 202 fights vs the same mob), and asuran fists' ACC mod is not enough to make swapping off ACC gear look like swapping on a bunch of ACC gear. if you're looking at the results i'm posting, the only way i can really take these criticisms is "well, lots of weird stuff happens so parses don't work".

i see where both of you are coming from, but the presence of variables doesn't make those variables relevant. also, if you've ever done any testing like that tyche is describing, you know doing over 100 WS each with 3 different pieces of gear would take forever (it's moot for me, since if i agreed with you both, AF would be invalid to test anyway, and i don't have the RF/DK gorget yet). i don't understand why you think AF's ACC mod is so powerful all of the sudden...

normal melee ACC is not a red herring-_-. i'm a bit baffled as to why you would think general ACC% is irrelevant to WS ACC.

[pathwriter]

Because... we're not testing melee accuracy.

Tell me, were you one of the ones who convinced the Bush administration that Al Qaeda, a Saudi organization with no ties to Iraq or the Shi'a, were directly connected to Saddam Hussein?

[milich]

melee accuracy is determined by the exact same thing as WS accuracy, barring TP mods and gorgets. either gorget gave me a bunch more WS ACC, or TP mods did. AF TP mod didn't give me all that ACC before i got gorget. hence, my conclusions. i know you're smart enough to understand this.

[pathwriter]

If I'm smart enough to understand that, why are you not bright enough to pick up on the very basic flaws in your testing premises that Tyche and I have pointed out? You want to know about the efficacy of elemental gorgets, go test elemental gorgets. You're making the parameters too complex and riddled with inaccuracy due to the number of variables. Then you complain that people are giving you direct feedback on your data and method because, what, it's not a pat on the back?

[milich]

i don't want a pat on the back, just a relevant critique. look at the results: they seem to indicate capped ACC for WS. magic variables don't just produce that... i mean, do you see how huge this would make asuran fists' TP mod? it would mean that everyone who doesn't swap on STR rings and such for asuran fists is a fool... but people who swap off ACC for asuran fists are fools. why have people like you and me been telling everyone that for years when asuran fists has that big ass ACC mod?

edit: (yes i edit a lot) i'm not married to my analysis being accurate. i just fail to see how focus, madrigal (both affect melee and ws acc equally) and AF TP mod are valid results-skewing variables.

[Tyche]

milich :
i don't want a pat on the back, just a relevant critique. look at the results: they seem to indicate capped ACC for WS. magic variables don't just produce that... i mean, do you see how huge this would make asuran fists' TP mod? it would mean that everyone who doesn't swap on STR rings and such for asuran fists is a fool... but people who swap off ACC for asuran fists are fools. why have people like you and me been telling everyone that for years when asuran fists has that big ass ACC mod?

edit: (yes i edit a lot) i'm not married to my analysis being accurate. i just fail to see how focus, madrigal (both affect melee and ws acc equally) and AF TP mod are valid results-skewing variables.

Focus skews it because you have it on for some of the melee, and some you don't. You have it on for some of the WSs, and some you don't. The up/down ratio for focus for melee hits is the exact same as the up/down ratio for the amount of WSs you did? You have no idea, you can't parse that. You didn't count any of your TP returns, you just went strictly off averages, which may be very close, is not an absolute. Forgive us for questioning a shakey, variable filled test that disputes a well-thought-out, controlled test. Like we said, there's a way to do this, and if you don't want to put the effort into doing the extensive testing corectly, then stop complaining about us questioning your half-ass test.