Gorgets and multi-hit accuracy

[Valyana]

I just went to dig up the gorget accuracy testing data for an LS conversation.

viewtopic.php?p=579313#p579313

Repeating one more time, for an acc test this sample size is way too small, need at least a few hundred hits per set and control tests with Peacock Charm and no neckpiece at all before taking any conclusions.

Anyone done more testing since then?

[aurik]

no.

[Chupunpupun]

http://releenaseraph.livejournal.com/11 ... 65#t223765

Mystery solved.

edit: adds ~10ACC during all hits of ws.

[Kriz]

So back to multihit with gorgets good?

[Max™]

That is interesting, cause I've been using Snow with Raging Rush to see if I could notice the fTP on a crit based WS, and haven't had any acc issues in merits, or the few times I've RRed Kirin with it.

Been trying to think of how I could test it better, but without a parser it would be really annoying.

[aurik]

So back to multihit with gorgets good?

Sample size still too small, so we're actually still at "use whatever voodoo you like"

[Sanjiyan]

Wouldn't it be an easy test if we use asuran fists on something evasive? Let's say something like Mamool Ja Lurkers.

I was thinking about it and assuming that a gorget does provide accuracy bonus on all hits of the WS (no reason why it'll only apply to just a few hits if it exists), reading TP returns on asuran without a neck piece, with PCC, and with gorget should give a pretty good indication after a good number of WSs, since each one is 8 hits and basically provide 7 samples (assuming the first hit always has a acc bonus from gorget) for accuracy. Asuran fists is also nice because it's 8 hits WS, so we don't have to worry about double attack and/or offhand hits (like /NIN) which could or could not be counted as part of the WS.

Maybe I'm oversimplifying it. Anyone who's good at statistics? How many samples will be required in order to say for sure if a gorget offers acc bonus or not on all subsequent hits for something like a 5% difference?

[Howard Roark]

Wouldn't it be an easy test if we use asuran fists on something evasive? Let's say something like Mamool Ja Lurkers.

I was thinking about it and assuming that a gorget does provide accuracy bonus on all hits of the WS (no reason why it'll only apply to just a few hits if it exists), reading TP returns on asuran without a neck piece, with PCC, and with gorget should give a pretty good indication after a good number of WSs, since each one is 8 hits and basically provide 7 samples (assuming the first hit always has a acc bonus from gorget) for accuracy. Asuran fists is also nice because it's 8 hits WS, so we don't have to worry about double attack and/or offhand hits (like /NIN) which could or could not be counted as part of the WS.

Maybe I'm oversimplifying it. Anyone who's good at statistics? How many samples will be required in order to say for sure if a gorget offers acc bonus or not on all subsequent hits for something like a 5% difference?

the dude that replied twice in this thread already knows a bit about math and asuran fists, i'd trust him for the time being until you find a better way to test it

[Sanjiyan]

Well it seems like it's not hard to test, just time consuming then. I'd do it myself but I don't have a PCC nor the time (school, etc.)

So I guess everyone's lazy? I just think it's a question that would be great if it's answered definitively. I mean, I never thought we could figure out emnity, but somebody did! I was amazed at that.

[Howard Roark]

Well it seems like it's not hard to test, just time consuming then. I'd do it myself but I don't have a PCC nor the time (school, etc.)

So I guess everyone's lazy? I just think it's a question that would be great if it's answered definitively. I mean, I never thought we could figure out emnity, but somebody did! I was amazed at that.

a pcc is holding back your testing lol...

[Seditedi]

Well it seems like it's not hard to test, just time consuming then. I'd do it myself but I don't have a PCC nor the time (school, etc.)

So I guess everyone's lazy? I just think it's a question that would be great if it's answered definitively. I mean, I never thought we could figure out emnity, but somebody did! I was amazed at that.

a pcc is holding back your testing lol...

First of all a PCC is Accuracy +10 while a Gorget is widely believed Accuracy +10%.
If you were trying to test WSing with PCC versus WSing with Gorget for a multi hit WS you wouldn't be getting accurate results to begin with because you'd be parsing the TP returns on totally different accuracy bonuses, whether or not the Gorget accuracy bonus takes place on all hits or not. You'd be comparing a rather large accuracy bonus to a minimal accuracy bonus of +10.

[Ferien]

It would probably be easiest to do using a 2-hit WS (Fast Blade, Enpi, Retsu, and a couple other lowbie WSes come to mind). Do a whole bunch using no neck, and then compare using Gorget. You'll have evenly-sized sets for 1st hit and 2nd hit, and could easily compare the two.

Then again, using Asuran, your pool of "subsequent hits" would be 7x the size of the first-hit pool, leading to a much, much higher rate of accuracy.

Alternatively, since it can be assumed that your ws hit rate will be the same as your melee acc, compare the pool of extra hits with that. Most of us have parsers, that's easy enough. You just can't switch gear between the two or you'll skew the results.

Besides controlling gear and acc, controls need to be in place to stop Double or Triple Attack, and any ws that kills a mob needs to be thrown out. Besides that, it should be a pretty simple (just time-consuming) test.

[Sanjiyan]

First of all a PCC is Accuracy +10 while a Gorget is widely believed Accuracy +10%.
If you were trying to test WSing with PCC versus WSing with Gorget for a multi hit WS you wouldn't be getting accurate results to begin with because you'd be parsing the TP returns on totally different accuracy bonuses, whether or not the Gorget accuracy bonus takes place on all hits or not. You'd be comparing a rather large accuracy bonus to a minimal accuracy bonus of +10.

If so, then gorget should have a huge, sushi like effect, on asuran accuracy and that would be easily seen. In fact I don't even care about the exact percent boost (or number boost), it still needs to be determined if gorget is actually viable at all.

Alternatively, since it can be assumed that your ws hit rate will be the same as your melee acc, compare the pool of extra hits with that. Most of us have parsers, that's easy enough. You just can't switch gear between the two or you'll skew the results.

That's a good point, so what we need is a MNK to go exp with their normal gear on a relatively homogenous set of mobs, preferably without capped accuracy (like 80-85%), and then asuran in the same gear, recording all return TP gain. Then another outing where only the neck is replaced by a gorget when WSing.

Seems simple enough. Maybe I'll do it during the summer if nobody still got around to it.

[Shuemue]

I suspect you could persuade a coder to add a function to calculate your average tp return, no? I assume the data is accessible seeing as TParty can display it. Then it'd be easy to have a whole bunch of people collect samples according to the pcc/gorget/nothing deal.

Or Aurik could add it to his private parser and go punch a few thousand mobs >_>

[Awoir]

The problem with Asuran like that is if the first hit connects and the rest miss your return is 5TP, if your first two hits miss and another 5 extra hits connect, your TP will also be 5, but you should be able to tell a difference in damage. This is something a Parser won't differentiate between.

[Shuemue]

Good call, would have to limit it to a function that spews out a list of each WS you do and its DMG and TP returns and then manually draw conclusions.

[Ferien]

Easy way around using a parser for it would be to call an /echo <tp> after every ws, with some signifier at the beginning of each line to trip a script or general log parser. Keep logmon or something similar going while you do the tests, run the script on it, and bam, you've got the TP return of every WS in the session in one spot. I've written scripts to do similar things before, and if I go out to test it myself that's how I'll be doing it.

The problem mentioned with Asuran is a pretty big one, though, and is one of the reasons I'd use something smaller for this test. Vorpal Blade, Rampage, and Evisceration (so long as you're not WAR, /WAR, THF, or dual-wielding) are probably the best ws to do it with, as they've got quite a few hits in them, don't have a TP-Acc mod (ruling Penta out, which would have been the ideal candidate), and don't have enough hits to match the first swing's TP, giving you an easy pool of data to sift through.

[Seditedi]

First of all a PCC is Accuracy +10 while a Gorget is widely believed Accuracy +10%.
If you were trying to test WSing with PCC versus WSing with Gorget for a multi hit WS you wouldn't be getting accurate results to begin with because you'd be parsing the TP returns on totally different accuracy bonuses, whether or not the Gorget accuracy bonus takes place on all hits or not. You'd be comparing a rather large accuracy bonus to a minimal accuracy bonus of +10.

If so, then gorget should have a huge, sushi like effect, on asuran accuracy and that would be easily seen. In fact I don't even care about the exact percent boost (or number boost), it still needs to be determined if gorget is actually viable at all.

Alternatively, since it can be assumed that your ws hit rate will be the same as your melee acc, compare the pool of extra hits with that. Most of us have parsers, that's easy enough. You just can't switch gear between the two or you'll skew the results.

That's a good point, so what we need is a MNK to go exp with their normal gear on a relatively homogenous set of mobs, preferably without capped accuracy (like 80-85%), and then asuran in the same gear, recording all return TP gain. Then another outing where only the neck is replaced by a gorget when WSing.

Seems simple enough. Maybe I'll do it during the summer if nobody still got around to it.

It wouldn't be easily seen if the accuracy bonus only applies to the first hit of the WS, which from data I've seen, firmly believe.

[Skeetaru]

I would like to thank this thread for officially ruining my world.

Someone test this and come back with definitive results - thats an order.

**scampers back to the relative safety of his PCC and weeps uncontrollably**

Also, has anyone done any research, definitive or not, on sea torque vs ws gorgets?

[Ferien]

Also, has anyone done any research, definitive or not, on sea torque vs ws gorgets?

Refer to Aurik's previous statements.