Lately, I've been playing a lot of mystery and detective games. My gateway drug was the Danganronpa series, which I already fell in love with at my first Class Trial. I'm already a sucker for good character design, even if the characters' personalities seem a little bit over the top (and if you played DR: Trigger Happy Havoc, you've got to agree with me on that), but the trials were when the game shines the most: the animated next, presentation, the expressions, plot twist after plot twist. I can't overstate how hooked I was.
Ever since, as a long time game jammer, I was waiting for the opportunity to make my own detective adventure. This compo was when I finally gave a shot at detective games, but with my own twists.
Why "Bloody Baron"?
The general setting of the game is a kingdom where a political assassination has happened. And pretty much all the possibilities point to someone who's a noble. The king, very upset, decides to lock out all the suspects in a building with no escape, until the killer themselves is executed.
Therefore, Bloody Baron alludes to one of the most famous serial killers in history, the Bloody Baron of Brittany.

The game design
My main inspiration for this game was Werewolf/Mafia, a party game that is the bane of my existence, because I love it so much, yet I'm absolutely terrible at it.
Detective games are complex. There must be a mystery and it must be engaging. There must be barriers towards solving this mystery, but they shouldn't be too overwhelming. There must be a motivation to do it. When it comes to Danganronpa, the motivation is pretty clear, you want to survive. And that's true, in general, the most basic motives are the ones the player is most likely to go with.
Because I already needed to handle this complexity, I decided to put everything else out of the way. Literally. Everything else.
And that's when the idea of text adventure came to mind. I had played some good ones in the past and wondered, this could work.
Some concepts were borrowed from the DR design, indeed, like the nighttime. The nighttime is supposed to make you approach it in a different way, you can't always be as careless, but maybe, exploring during nighttime could give you important hints to whom the killer is!
As for procedural generation, I'm no stranger to it. I attempted it at my first compo entry, Dream Roamers, in 2017, with a fairly good result. It can get cheeky to promote your game on the premise of procedural generation, though, even when it's something like generating the map, so we can put that aside, and focus on what the game needs to be: an exciting and dangerous mystery.
Character designs
I'm a big fan of the guideline "show, don't tell". So instead of going with a huge lore dump at the start, the plan was to give the player the pieces to glue together and find out the story behind Bloody Baron. Even though that was the plan, showing and not telling is sometimes not as easy as it sounds.
Characters were designed on the run. Even if I'm kinda proud of my work, it could've gone so much better. Many of them had potential and were underdeveloped simply because I ran out of time. Chio, as a war criminal, should be the prime suspect right off the bat, but maybe he isn't really the murderer. But I couldn't get that across simply because time ran out and I had to ship the compo entry without anything on his description. Same for Odette, and Larissa, which were characters I was proud of designing but the players probably won't see them shine for what they are.
Since this is a game by me, catgirls are necessary. Luna, the catgirl, was actually the first character that came to mind, as some kind of comic relief to the whole situation.
I can't help but make lore and characters for my jam entries, and that has thrown me on the dirt, too many times. It happened in the Ludum Dare before, it happened in the GMTK game jam this year (yes, there were catgirls there too) and it will happen again.
Finishing the game at the very last hour, and the first feedback I got.
This game had plenty of breaking bugs when I first submitted it. You couldn't reach nighttime without the main thread panicking. Inserting out of bounds input caused the IO to hang indefinitely. Voting in the trials had a chance to crash. Those were the first things I've heard from players, and frankly, after spending my weekend locked down in my room doing one thing and barely leaving to go to the bathroom, it is just a little bit disappointing.
The process of bouncing back was mainly using debuggers and trackers and finally doing some play testing, which I didn't do much before. After I tightened up some screws, people actually seemed to like the game. I got a lot of positive comments after that.
About Among Us and my other inspirations
Among Us was definitely a factor in building the hype for my game. I can't really say it was a direct influence, because I never really played Among Us. Even then, I always liked the concept, and have been looking forward to give it a go for some time.
While Among Us and Werewolf were so much about the mind games between people, and Danganronpa is about concrete clues that the murderer lefts behind when committing a murder, the intention of my game was to be about logical thinking and building trust among the people that are there locked with you. I can say for certain that all of these games played a part in my decision to make this entry, but they aren't the same as my entry. I feel like I succeeded in making something that bears its own flavor.
TL:DR, and advice for new jammers
In general, I managed to make a game that had an interesting concept, and people will agree that it was fun indeed, but was buggy as hell at first. What I learned with that was to spare more time to play testing. It's just as important as polishing.
As for the things I got right, I think my choice to remove everything out of my way and make a text adventure so I could focus on building the mystery was definitely a factor. So, it's not always about making the scope simple, but picking your fights. And don't pick them all, or you're gonna get smashed.