[mikesjustice]

Test 1: Iron Sheets
Direction NW
Incremental synth pattern at 0:00, 0:20, 0:40 times.
72 batch each elemental day covering 3 different elemental days.

Results: 46%-50% HQ average achieved. Multiple cross tests done covering 4 rl months, synthing weeks at one set time/moon/day, attempting to control all factors as best as human error will allow.


Test 2: Iron sheets
Direction N
Continuous synth pattern
Batches of approx 100 synths per elemental day nonstop covering 3 elemental days. Tests were done for 1 rl weeks time.

Results: 50% HQ average met approx. I don't remember the exact percentage, I think 46%-48% it was, but nothing noticable to throw off the numbers.


Opinion: HQ average > all. I've done other tests as well, but unfortunately all it takes is one "other" direction scenerio to support the direction theory doesn't hold water. If we want to go even further, we can look at the fact that NW is a Light based element, and N represents a Dark element, yet we see similar results.

[ikkleste]

Originally Posted by Nakti

Directional synthing has always been of interest to me.

I've devised a directional synthing (not desynthing like you're doing) to do with my husband. Conducting the experiment with someone else at the same time eliminates most of the time- and weather- related variables.

Experiment synthing Antidotes

Control Test (just to verify no difference based on race/gender/equip):

Me: facing Light direction (NE)
Hubby: facing Light direction (NE)

If our set of NQ, HQ, and break % aren't the same (within error margins), there must be a difference due to race/gender/equip. Both of us will have Alchemy at the same exact skill level.

Experimental Data

Me: facing Light direction (NE) always
Hubby: facing each of the other directions in turn (ie 100 synths facing East, 100 synths facing SE, etc)

The experimental NQ, HQ, and break % are normalized against the percentages from facing the Light direction. This eliminates the need to synth always on a certain moon phase or day.

I figured the cost would be a minimum of 500k for materials and a minimum of 8 hours testing. I'm working on some other tests right now so if someone else beats me to it, please do!

it occurred to me to run the same experments but i don't have the facilities to do so (only one computer on the game at the moment.) the other way you could do it would be to get ten people (from a crafting LS for example) to conduct a similar expt.

If our set of NQ, HQ, and break % aren't the same (within error margins), there must be a difference due to race/gender/equip. Both of us will have Alchemy at the same exact skill level.

alternatively there is a true randomisation factor (not based on time/date/moon, but a random generated number each time a synth is done, instead of all conditions coming together to say if you succeed or fail, they come together to give your chance of succeed or fail.).

You could further check this by starting mules and conducting the same expt... then you could set race/gender/equip to be identical, eliminating other factors. With this setup you could actually check a hell of a lot (whether direction, gears/resistances, etc), or finally prove whether there is a true random factor or not.

It also might be worth to be in the same place, ther may be other local conditions that aren't aware of (say degrees of weather rather than just fire or double fire it might be possible to be 'a bit fire but not quite a whole one') being in the same place would eliminate this (admittedly unlikely) idea.


If it does prove to be that there is a random factor then we at least know for sure the best we can hope for, in any investigation, is a statistcal analysis.

[Nakti]

mikesjustice, what was your skill level compared to the cap of the iron sheets? It looks like you were in the +31 tier. I remember seeing some tests suggesting (but not proving) that conditions affect the tiers differently. ie element day > skill for +11 tier, and skill > element day for +31 and up tiers.

ikkleste, I would love to have 9 or 10 people testing at one time. ^^ But that's not going to happen at my house, so I'll just have to do what I can when I can.

sevenpointflaw, thank you for all this testing!

[ikkleste]

Originally Posted by Nakti

ikkleste, I would love to have 9 or 10 people testing at one time. ^^ But that's not going to happen at my house, so I'll just have to do what I can when I can.

hehe, i'm sure they'd make the trip to where ever you live ^^

nah, seriously, once you get the initial testing done with the two of you, you could try doing something with extra subjects. Rogue results will be easier to spot, and while the error margins may increase, even running from different places you might be able to get some meaningful results.

[Aeni]

Originally Posted by Nakti

mikesjustice, what was your skill level compared to the cap of the iron sheets? It looks like you were in the +31 tier. I remember seeing some tests suggesting (but not proving) that conditions affect the tiers differently. ie element day > skill for +11 tier, and skill > element day for +31 and up tiers.

Not just that, but what are the elemental days even? mikesjustice was not specific in this regard.

Light - Light, Fire, Wind, Lightning
Dark - Dark, Ice, Water, Earth

Any crafter worth half their salt or 60 or more levels into their main job should know this by now (If not, stop playing the game and cancel your account)

The tests done could've been made on days where it would be strongly adverse to one direction and the next set of testing done on another set of days which skews the data and sugarcoats the results. Also 72 batches and then 100 later isn't exactly the same thing.

I want real honest to goodness testing with exacting and specific process.

This is called the scientific method, which most of the testing I've seen have failed in. Until then, I'll continue to use the crafting timer, develeped over many years by pooling results from various crafters and their experiences as well as using mathematical formulae that actually makes sense.

[sevenpointflaw]

Originally Posted by Aeni

Not just that, but what are the elemental days even? mikesjustice was not specific in this regard.

Light - Light, Fire, Wind, Lightning
Dark - Dark, Ice, Water, Earth

Any crafter worth half their salt or 60 or more levels into their main job should know this by now (If not, stop playing the game and cancel your account)

The tests done could've been made on days where it would be strongly adverse to one direction and the next set of testing done on another set of days which skews the data and sugarcoats the results. Also 72 batches and then 100 later isn't exactly the same thing.

I want real honest to goodness testing with exacting and specific process.

This is called the scientific method, which most of the testing I've seen have failed in. Until then, I'll continue to use the crafting timer, develeped over many years by pooling results from various crafters and their experiences as well as using mathematical formulae that actually makes sense.

While I can't speak for other people, the gripe I have with the crafting timer is that it isn't based on any actual formula. It took me for ever to track the development thread down on it, but it is based highly on assumptions, and basicly just calculates assumed ralative difficulties.

And while it seems to be highly accurate in that regards, it is hardly scientific, based instead off non-tested assumptions about how various directions affect synthing, as well as the assumption of how much each direction affects synthing.

I can't deny that the crafting time seems more acurate than randomly spinning around to HQ an Aristocrats Coat, but show me the hard, concrete data used to put the crafting timers forumla together. You won't find it. Even the original thread where the author of the timer talks about it states that they are assumptions.

Hense, the long ardorous task of the Cotton Kayhans and Shinobi Tabi that I'm doing. Hopefully it will confirm some of the theories and debunk some of them.

[Hamlet]

Also, make sure to note what elemental resistances you have equipped, because you may resist your synth and scew the results.

j/k I'm looking forward to more results. Anytime I've tried to do directional synthing, the results seemed completely random (whether it be facing in the direction of the crystal, facing the opposite of the direction, or facing the direction of the element the crystal is weak to- I seem to get the same results overall. Every once in awhile I'll get an hq on the first one and think that's the right direction, but then like 6 nqs in a row on something I should be able to hq 25% of the time seems to disprove that).

[sevenpointflaw]

No new data added, but I fiddled with the way it looks some, and added % results to what I have so far.

[mikesjustice]

Quote:
Originally Posted by Nakti
mikesjustice, what was your skill level compared to the cap of the iron sheets? It looks like you were in the +31 tier. I remember seeing some tests suggesting (but not proving) that conditions affect the tiers differently. ie element day > skill for +11 tier, and skill > element day for +31 and up tiers.



Not just that, but what are the elemental days even? mikesjustice was not specific in this regard.

Light - Light, Fire, Wind, Lightning
Dark - Dark, Ice, Water, Earth

Any crafter worth half their salt or 60 or more levels into their main job should know this by now (If not, stop playing the game and cancel your account)

The tests done could've been made on days where it would be strongly adverse to one direction and the next set of testing done on another set of days which skews the data and sugarcoats the results. Also 72 batches and then 100 later isn't exactly the same thing.

I want real honest to goodness testing with exacting and specific process.

This is called the scientific method, which most of the testing I've seen have failed in. Until then, I'll continue to use the crafting timer, develeped over many years by pooling results from various crafters and their experiences as well as using mathematical formulae that actually makes sense.

If you had read my statement then you would have your answer. Any crafter with half of mind would comprehend what was mentioned.

72 batch each elemental day covering 3 different elemental days.

Batches of approx 100 synths per elemental day nonstop covering 3 elemental days. Tests were done for 1 rl weeks time.

Originally Posted by Nakti
mikesjustice, what was your skill level compared to the cap of the iron sheets? It looks like you were in the +31 tier. I remember seeing some tests suggesting (but not proving) that conditions affect the tiers differently. ie element day > skill for +11 tier, and skill > element day for +31 and up tiers.

If I was posting results with nearly a nearly 50% HQ rate across thousands of test results, would I be in tier 31? Come now.

Now tell me the difference between specifying particular days or picking 3 consectutive elemental days, and controlling your tests to repeat the test pattern again and again?

All it takes is one elemental day to see a variance between another elemental day. Not only did I use 3 elemental days, I followed it up by repeating the tests again covering those same 3 elemental days over a duration of time.

Also, using the controlled method, taking in account travel time, drop or npc time, etc, you can only synth 72-120 ish synths in any one given elemental day, as you are bound by time. Of course it's not a large sample, hense the reasoning for continuing the tests over a large period of time.

Until then, I'll continue to use the crafting timer, develeped over many years by pooling results from various crafters and their experiences as well as using mathematical formulae that actually makes sense.

Well, if you want to continue on with the crafting timer, by all means. Understand your typing along side of one of the members that helped theorize the concept. Divisorytheory, Lok, PurpleNV, myself, we all contributed our knowledge whereas NV developed the mathematical formula and lok put it to table form. We are some of the craft veterans of the Allakhazam boards, where divisor and I nearly tested for every theory that came across the site, including cross server sync tests.

Also know that the crafting timer newbie crafters worship is an excellent starter tool, but it's simply that, a starter tool and is rather outdated at this point in time. The only thing truely accurate about it are the directions in relation to elements. Any experienced crafter, and I mean experienced, knows that regardless of skill from cap, 1 or 5 out from item cap, you will skill up just the same. That's fact. The difference is that the further you are from cap, the larger the increase for fail - duh. Also know that should you use support or gear to "modify" your gap, you can still skill up consistantly at item cap, or 1 point above. Support will never hinder you, however it puts you into the positive for two specific reasons.

1.) Once you break item cap through support, modifying your gap, you break into the first tier for HQ, which is greater than the HQ odd when you are below item cap.

2.) Your fail rate decreases slightly. Once you break ground zero, or item cap, your fail rate will decrease. This is easily tested for, I encourage you to go do some testing.


Now, I'd rather not be so blunt, but understand that some of us are not on level playing fields as far as knowledge of crafting goes. Some of us have walked the path of a thousand miles and all we do is share our finds.

What you believe to be relevant now, is exactly where we were 2+ years ago. We've tested and made our finds. After much experience and wasted time/gil, the answers come to you.

Ultimately it's just as foolish to suggest "individual factors" play any part of the HQ formula as it was it to suggest "Which elemental day" did I synth during my tests. Without understanding what I was doing specific for those tests, it's clear your not ready to move towards an advance mindset in relation to crafting.

And if it's too much to take in information from people that have already walked the path, I encourage you to get out there and start testing. That's how we did it. See ya in 2 years!

[sevenpointflaw]

Updated with more data. Lightningday is nearly complete now.

[blackmajick]

on a side note i think the only thing missing from your test is time, personally i think that time has way more of a factor than any direction you face.

[sevenpointflaw]

Time? As in time of day? Time spent crafting consecutively? Time of the year? Or is that purely speculation on your part, which I'm guessing because I've never heard elesewhere that "Time" affects crafting.

[Myaku]

Originally Posted by sevenpointflaw

Time? As in time of day? Time spent crafting consecutively? Time of the year? Or is that purely speculation on your part, which I'm guessing because I've never heard elesewhere that "Time" affects crafting.

I have heard Time theories before, but I don't think it has more of an effect than direction. I have done limited testing, and there are some times that gave abnormal HQ%, but nothing that I would call significant with respect to the sample size. I haev heard anywhere from 30-min blocks of time to 3-hour blocks of time. My personal theory is that time operates much like Moon Phase - New and Full vs. Midnight and Noon.

[Nakti]

The "Eruntalon Code" threads on Alla talked about time within a Vana'diel day affecting HQs, but I haven't seen any other serious discussions regarding it. Maybe time affects HQs, maybe not (it affects fishing for example). My personal guess is that if time is a factor, it's not more a factor than direction. If the test synths are all done at random times, time shouldn't be too much of an influence (if it's an influence at all).

sevenpointflaw, keep up the good work!

[sevenpointflaw]

Well, just look at the existing data. Like I said, any data that has 25+ entries is done over the course of an in-game day. Whether or not I get 25 synths or 50 synths is a matter of how bored I am. Either way, there is no clear trend in time over the course of a day. The only "possible" trend you can see is in Watserday, where I get an abnormally large squence of losses and breaks; but I have a feeling that trend will average out with more watersday testing.

And honestly, the only other thing that Time of Day affects is pops for mobs. SE has a nasty habit of recycling code. Very lazy programers. And while I don't like to discount the possility, honestly, the idea that "time" is a factor is frankly probably bunk.

Opinion aside though, look at the data. If you can point out where time of day shows a distinct trend that can be distinctly accounted for, I'll listen.