The last few days I’ve been toying with various “Drunkard’s Walk” algorithms.
Well, technically I spent about an hour toying with various Drunkard’s Walk algorithms, and then about 2 days fighting with the obnoxiously convoluted and irritatingly documented “matplotlib“. I’m probably going to switch to plot.ly, but that’s not really germane to the topic at hand which is that matplotlib sucks more eggs than Paul Newman in Cool Hand Luke. Wait, I was talking about drunk walks.
Above is a graph of the first one I set up, which I lovingly call “shell game”. Basically, if you start with $5, and have a 1/3 chance of winning $1 and 2/3 chance of losing, how long will it take to double your money, or to lose it?
Now, obviously, that’s a pretty solved problem in math, I know. My plan was to run the simulation 10 times, and then put all 10 on the same graph. I’m still working on that last part (see above, re: Paul Newman). And then I have a few others (75% chance of losing $1 and 25% chance of getting $2, etc) I wanted to try.
To what end? I don’t know yet. But I figured that I could get that info, so I might as well do it.
Battle Royale
I also find myself slowly starting to drift back towards programming projects that will again put me roguelike-adjacent. For example, I sketched out some pseudo-code for a “Jumble” auto-generator, which I’ll probably implement this weekend. There’s not a lot of overlap there for the outside observer, but I know there is.
I’ve also spent a LOT of time researching various procedural map generation techniques (which is sort of related to the Drunkard’s Walk exploration, but loose lips sink ships, after all.)
Currently I’m toying with the idea of a Battle Royale roguelike. By which I mean, a B@L Royale roguelike (pronounced B-At-L, of course). A couple of months ago, I read the Japanese novel via Scribd. Okay, technically I listened to it as an audio book, but it was read by Mark Dacascos, so that’s pretty cool.
Lately, there’s been a lot of Battle Royale “inspired” games – Fortnite, PlayerUnknown’s Battlegrounds, H1Z1, etc. But after reading the book, I think that a lot of them are missing a crucial element. So much so that I actually wonder how many devs read the book as opposed to just watched the movie or something.
See, the thing that makes Battle Royale the novel different from these games is the psychological aspect of it. Much of the novel is spent exploring the mindsets and reasons for the various characters trapped in this life or death, kill or be killed situation. It’s a lot of why I’ve seen various people call it the Japanese “Lord of the Flies”. But you don’t get any of that in a video game where you’re playing against 100 strangers. It’s simply a sociopath simulation.
A few years ago, a man attempted a Battle Royale 7DRL called BattleRL. He didn’t finish it, and he was definitely going in some different directions (hewing much closer to the novel in narrative feel rather than in implication), but I think he had some interesting ideas.
Two challenges that I’ve started putting some thought into:
- How to provoke that feeling of empathy, not just the desire to “score high”? In the novel, obviously, the main character had his need to look after Noriko, and his desire to save as many classmates as he can. Right now I’m trying to imagine some Prisoner’s Dilemma situtations and maybe create AI’s that will cooperate and/or betray the player?
- How to add some dynamism. In the novel, the sensible move was to stay put as much as possible, since moving meant being caught outside of safety. But that’s not particularly interesting to play.
As you can see, I still have things to consider. An I’m not really planning to make it a 7DRL, so I have time to consider them. But how it’ll pan out is still something of a mystery to me.
