Ok, so how does the shop algorithm actually work?

[Gakato]

Well I'm not that mathsy, though I'm pretty sure there is no simple way to calculate the cumulated probabilities for every single item to appear in a dungeon. At least for me it was much easier and faster to just count the chosen items in one billion dungeon runs and calculate the probabilities based on that.

[William]

[Tinker]

[dislekcia]

That little square root really messes with permutations, doesn't it

[Gakato]

[Astral]

This is a pretty interesting topic

Like most statistical problems this one isn't difficult either, just tediously long to calculate. That's why Gakato's approach is clever, it's much easier to simulate it and collect the data.

First, I have a question. Shouldn't rule 6 say 'subtract from n'? Otherwise we can get a -1 index, and the list goes 0-37.

Btw I started constructing an equation for this problem, but without some programming it will be useless. I can already see that the final form will be monstrously long, containing plenty of sums to expand and around 3600 unique factors.
You couldn't even showcase such a big equation, let alone work with it. That's for 8 shop dungeons, with each extra shop added the number would increase exponentially.

I will try to make a table collecting the exact probabilities, but can't promise a condensed equation.

[Tinker]

[Gakato]

[Phoil]

[Astral]

No probs and thanks for clearing that up

Though that means at best we can only generate the item in the slot n-1-0, so it's impossible to pull the last one. Has anyone found a Zot in the shop without elite items prepped or the Witchalok Pendant when elites were on?


I don't want to hype you up too much Tinker, it's taking long to solve, I'm only done with the mechanics of the 1st 4 shop yet. And it's only getting harder from here. Also a little tweaking is necessary, cause I assumed we subtract the random number from n, not n-1.