I HATE MELDING

[BobbinT]

Last night I literally failed 34 attempts in a row to meld a level 3 materia to the fourth slot of my secondary tool, at an 11% success rate.
I then tried my luck on another meld at 35%, failed 6 in a row, and decided it was a bad night to do melds.

If I remember my statistics right, the mathematical probability of an event with an 89% chance happening 34 times in a row is 1.9%.
Probability of 65% failure triggering 6 times in a row is 7.5%.
Probability of both happening consecutively is 0.14%.

Yeah, really bad luck. For the record, up until that point, I've had what felt like a relatively fair distribution of successes and failures, but over a million gil worth of materia went down the toilet last night. The worst part is I didn't do that many failures before getting a success. I walked away from those melds with nothing to show for it.

Something I learn from other person the other day:

"Statistic means nothing for individual"

Even if you have like 80% chance, it could take away dozens of tries. That's how bad RNG is. lol

[Kos_Mos]

Something I learn from other person the other day:

"Statistic means nothing for individual"

Even if you have like 80% chance, it could take away dozens of tries. That's how bad RNG is. lol

Thats called the law of large numbers, the larger you sample size the more accurate the percentage becomes. If you synth 100 ingots with a 1% hq rate and get 0 hqs its back luck. If you make 10k brass ingots though you should get around 100 of them.

[BobbinT]

Thats called the law of large numbers, the larger you sample size the more accurate the percentage becomes. If you synth 100 ingots with a 1% hq rate and get 0 hqs its back luck. If you make 10k brass ingots though you should get around 100 of them.

Haha... sorry though. Even if it says 5% chances, which should mean you get one out of 20 tries, it still could mess you up beyond said chances. As what @scarebearz just posted, his/her failure are way below 11% and more like 2% roughly. Whenever you trigger another chance (or bet), it won't go up like 2/20 -> 3/20 -> and so on, it will always be 1/20 chances.

I understand about your large samples theory, since I made my thesis about data mining for those. It needs to be repeated for same quantity though. And still... in the end, it just doesn't matter for individual. Good if you get result on said chances or better, but be prepared when it goes worse than you think.