Mass melding statictics. {You can have this}

[Ralliare]

Statistics are a funny thing, if you want to prove to a scientist a 25% chance, you're going to have to do it about 5000 times or more. Ergo, just go with the numbers they tell us...

[Raldo]

From this data, I glean that triple melds have a 100% success rate.

[Cynthis]

Wait did you really blow 25 melds in a row? Damn.

I got my double-melded NQ hat on the 49th attempt and most of those were grade III materia with between 16 and 24% chance. I feel the OPs pain. This was spread over about a month though so I am in no way trying to support the theory that the chance diminishes after a number of tries in one day.

[Deceptistar]

Why are people so quick to bashing him for posting his results. He never once stated that the figures he got were THE general %, but more HIS personal % he got. Its posted so that whatever WE do in relation to it can 'help us optimize' aka think and refine it if we think we need to. Technically its no diff. than BG's Legatus Challenge.

Its an experience someone posted and to each diff. person the info may be worth following, refining, or just plain worthless.

[Fievel]

Why are people so quick to bashing him for posting his results. He never once stated that the figures he got were THE general %, but more HIS personal % he got. Its posted so that whatever WE do in relation to it can 'help us optimize' aka think and refine it if we think we need to. Technically its no diff. than BG's Legatus Challenge.

Its an experience someone posted and to each diff. person the info may be worth following, refining, or just plain worthless.

What?

No

[indira]

sometimes you get 2 in a row, then you get nothing.
melding 2+ is random i dont trust the percentage its shows.

[Dzian]

There is somewhere on google docs A bloody huge spreadsheet some jps did on melding and spirit bonding. i really wish i still had the link but it was some pretty mass testing. something like 5000 dated tar voyagers belts and thousands of other common sb items converted to materias giving a % on each type. I really wish i had that link still but maybe someone can post it.

the melding part of that sheet was typically just moon phase testing and provided no proof that moon phase influences success.

statistically you need a sample size big enough that each individual result cannot have a big impact. example afer 8 attempts your success was 0% after 9 attempts it was 11.1% after 10 your success was 20%. after 12 attempts your success was 33% each individual synth has a fairly large impact on your results. to be accurate you'd need at least 1000 attempts at which point each attempt could at most provide a 0.1% deviation in your results.

out of 50 i count 10 "o" successfull attempts thats 20% success rate. 50 attempts you have a 2% deviation per synth of your results. so you only really have 3 anomolies in your results. the trouble is that each of those skews your results by 2%

been ages since i did this stuff at college....

[Rau]

someone needs to post that copypasta about how "told" that OP has been told.

[Ricky]

Why are people so quick to bashing him for posting his results. He never once stated that the figures he got were THE general %, but more HIS personal % he got. Its posted so that whatever WE do in relation to it can 'help us optimize' aka think and refine it if we think we need to. Technically its no diff. than BG's Legatus Challenge.

Its an experience someone posted and to each diff. person the info may be worth following, refining, or just plain worthless.

personal observations aren't valid in science and mathematics though
I know he isn't claiming to be a scientist but still

[AmyNeudaiz]

I've seen someone get a gamblers crown on 3rd try and make a triple STR IV on his First try right after that. This makes your "statistics" now void.