In The Name Of Science, 100 Dial Key #Fo Results

Thanks for the test.

I was thinking about testing it myself and see if heavy metal pouch rates were worth it, but ended up getting a bunch of standard dial keys instead, seems it was a good choice.

There's 28 things (I think) the fo keys can give, and although half of them are pure trash, I think it's worth it over sp gobbie keys

But only because I would never be lucky enough to pull an o sash or new sroda stuff with sp gobbie keys

17

https://www.bg-wiki.com/ffxi/Dial_Key_Fo

Aye, that's incomplete. I just counted what he had above

The screenshots show 28 different items.

That bg-wiki page does not account for M1+M2, H1+2, etc. or the 2 types of each of plutons, beitetsu and boulders.

it's also missing frayed sack (Mel) and such.

cool, someone have a bg account to update it?

Just wanted to put this here from anniversary event totals in case anyone was interested in distributions. Just seems kinda rando to me, but I'm sure someone might like to see

Thanks for posting this. I have always thought that something was off about the way that this game handles "random" outputs in certain areas, and this is just more evidence of how arbitrarily weird it is.

To match the dataset of 28 items above, I ran simulations of 22660 random selections of the numbers 1 to 28, and the count of each number falls within the approximate range of 740-880 almost every time. However, in the Dial Key #Fo table we see an erratic distribution, and it's clear that there is not an equal chance of obtaining each item.

Off the top of my head, this suggests one of a couple different possibilities.

1. The items are intended to have an equal chance of being obtained, but there is something severely wrong with the way the game handles RNG and/or there are unknown confounding variables that severely skew the results.

2. One of the devs manually input all of the probabilities to be the way that they are and they are working as intended. This is fine in theory, it's just quite strange given that some items in the same product category have vastly different probabilities (e.g. Frayed Pouch (A) and (D)).

Yeah it's goofy. Die/residue/cotton/linen clearly run lower chance than pluton cases, bayld boxes, frayed sack (a), and frayed pouch (m1) in my sampling. Almost like they didn't want as many die/residue/Alex from them and just shoved some of their drop chance into into the other items, ha

Guess I should have noted the moon phase and direction I was facing! *oof*

Maybe someone could record /random 10000 times (or whatever) and see if there are RNG problems in other parts of the game.

Been playing around with the numbers some more.

The more data points the better, but you could see some statistically significant differences between data sets with a much smaller sample. Even 1000 might be good enough based on the standard deviations that I'm seeing.

very similar rng numbers as Ave