Posted September 24, 2023 by EsoDev
Alright, this time, the update is a day early, and that's because I'm at the doctor's tomorrow and then at work for three more days. Still, nothing to share in terms of lore, as I just don't have the time or energy to deal with that.
So, what's on today's menu? A few things, big or small, and I'll be going over them in no particular order.
Firstly, let's talk about Karma and probability. It's an established bit of the lore that Karma can alter things in your favour. I've decided to make this more concrete. A random chance can now be marked as either "good" or "bad" for the PC, and accordingly, it will be modified by Karma in two different ways. Firstly, your current fluid Karma always affects random chances, similarly (actually uses part of the same code) as it affects skills though the "luck" modified. However, additionally, if you are using bursts of Karma (the clovers below your choice menu, basically), you will be able to use them to affect random chance, too. When active, they not only influence skill but also temporarily affect all probability rolls in your favour. This will pay dividends, design-wise, the more interactions we have.
Secondly, we finished the vending machine for the moment. As we've been nudged to do last week, we've included the possibility of items getting stuck in the machine. This leads to a little bit of expression for the player. Let us know if you want some other options to be added here. (And yes, this does use the system described above, where good Karma will save you from any snack-stuckage.)
Thirdly, back when I showed off how the bank works, I mentioned we wanted to do a few more things with it. Well, for some of that, I needed a working simulation of a stock market. Unfortunately, things went badly at first, as I started with realistic models that just didn't work for the purposes of the game. See, the issue with realistic models is that they also include possibilities of the market collapsing in various ways, and for a video game, I needed something that looked right but didn't necessarily have the full range of possibilities. Thus, I moved away from those models and set out to design my own that would have the desired properties but also held some stability for gameplay. What I came up with was a Higher Order Markov Chain based on a Beta distribution (some more details are on my Discord if you want to know) that gives good results in testing. This is how it behaves for four subsequent runs, each over the span of 1000 days:
And finally, we will go a bit towards research and design. We're currently trying to make right with the park location and make it everything it was supposed to be way back when we first came up with it... and to do this, we've been doing a lot of research on plants (mainly trees and flowers) and animals (mainly birds). Yeah, we're researching plants and animals; the places development takes you can be quite strange. Anyway, this is related to plot/lore implications and... well... we hope it will be quite a bit more interesting than it might seem at first.
Partially, though, this research will be swung towards modular narration to build universal, self-adjusting descriptions. These descriptions then become reusable and spread to other parts of the game. It just so happens a lot of what we have in the park happens to be about trees and flowers.
We've also continued with the usual improvements to CSS and minor code fixes that come up here and there.
Anyway, I'm still in a bad position with my usual work, so we'll try to do what we can, but the next significant changes might need a while. I might want to find something smaller to work on because, right now, that would fit my limited free time a lot better.
We do have some food and drink things on the roster.
Honestly, though, at this moment, the biggest issue we're facing is just the lack of man-hours dedicated to writing and data input to keep up with our backlog.