I HAVE NO IDEA WHAT IM DOING - ROLL
Video Statistics and Information
Channel: Angory Tom
Views: 44,486
Rating: 4.9414959 out of 5
Keywords: Angory, Tom, angor, yogscast, walkthrough, playthrough, funny
Id: 90lYvB7Mv3M
Channel Id: undefined
Length: 28min 11sec (1691 seconds)
Published: Tue Oct 19 2021
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.
This is the sort of game that makes me itch to try and work out the mathematically optimal strategy... It really feels like it is doable. Seems Tom thinks so, too. Here's my take on the basics:
The 'add X pips to the face of a die' upgrades are actually really easy to figure out. Doesn't actually matter what face of the die one buffs, in isolation. The average value per rolls is the sum of all faces divided by the number of faces, putting more points on a single face ups the average no matter which face one puts it on. At the number of rolls we're working with chance isn't really a factor, it's all about the statistical averages.
The value of a new die is also pretty straightforward: At minimum, it's (number of remaining rolls) x (average value of new die roll). Obviously you can upgrade the new die, but that's still pretty straightforward.
It becomes a non-trivial optimization problem when you add in the more complicated buffs and how they affect each other. At this scale the individual rolls don't matter, but figuring out the exact optimal strategy... Yeah, that's an undertaking.
That said, it's possible and the way to do it is pretty clear. If I had some major free time I'd be able to program a Python script that could produce the optimal sequence of upgrades, and even one that could advise your per-move actions. (So long as you were willing to input the results of every single roll into the script so it knew what was going on.)
EDIT: Assuming the store and/or costs aren't randomized. If it does randomize either of those, at best I could make one that advised optimal actions on a per-roll basis.