Dev Blog #1 | An Action Scene Which Became a Game

It’s been a while since I last posted new content here. As you may already know, I’m hard at work on my second solo-dev title, Tenjutsu.

Well, not so solo actually, as I’m working with Daniel aka Dnilox, who helps me with the character animations and in-game chara-design, and with Devolver Digital, who take care of most of the marketing stunt work.

To sum it up, I like to call my game a rogue-fu. A mix of smooth martial arts combat in a roguelite structure, that takes place in a vast open Japanese city.

Like many indie developers, my game began as a game-jam entry in Ludum Dare. You know, those weird events that feel like 2000s LAN-parties, where a bunch (thousands, actually!) of people “gather” to create games in a super limited amount of time, with a theme constraint. It usually happens over the course of a weekend. It’s both super fun and super challenging. You promise yourself you’ll never do it again, then you do another one.

Anyway, long story short, the jam went great. I got lots of very positive feedback, and it encouraged me to push the concept a little bit further. You know, let’s make a short project, I mean the game’s already there right, what could possibly go wrong with a 6 months project? That was 3 years ago.

So Tenjutsu started as a linear beat-em up with a 10-second loop constraint, but it naturally evolved to a rogue as I added more weapons and some randomness to the mix.

That’s kind of surprising, as after working on Dead Cells, I basically thought I was done with this type of game… So I guess, never say never?

Core Pillars

If you’ve heard about game design before, there’s this general rule of thumb that says: define your game “pillars” early, print them on paper, and put them on display in front of you, so you never forget them. Pillars are 2-3 core concepts that really define a game. They can be anything, from a few words (eg. “ Thrill of the Unknown ” in Subnautica) or a piece of music that conveys a particular feeling.

In my case, it was a few scenes from the John Wick movies, and specifically the museum combat in John Wick 3: Parabellum.

TRIGGER WARNING: violence & blood

These scenes really captured what Tenjutsu is about:

• Organic combat with opportunity as a key mechanic,

• Dancing around instead of focusing on one foe,

• A good variety of player actions again, based on opportunities and threats during combat.

I usually sum up that pillar with the “ one-vs-many combat ” concept.

Unforeseen consequences

Of course, these ideas all came with an awful lot of specific challenges:

• How do you create an AI that creates such dynamic encounters in a procedural setting? I mean, that’s a thing to carefully design combat in a fixed setup (eg. Sifu, love that game) and that’s another topic to generate random encounters that are both interesting and non-repetitive.

• How do you make a “one-vs-many” combat readable while being intense? Things can happen from any direction, and not all players share the same attention or split-second reflexes.

• How do you let players progress in such a game? Should they level up and become more powerful through stats, or obtain new weapons and techniques, or just git-gud? And then, how do you keep the early game interesting once players have tasted more “advanced” combat situations?

That’s a lot of questions, right?

There are definitely many things