SHDL#02 Character-based Design

The hook of the game is experimenting with different character skills and combos, specially in co-op, so let's talk about how the whole game was designed to enhance it.

As often with my concepts, this game idea came from some cool assets I had that I wanted to play with. In this case, a nice handpainted character pack with some class-based variations my wife gifted me.

Since they were actual characters, not only different clothing and weapons on the same base model, I thought that using different characters should feel different as well, and to me that's always best expressed through overall playstyle, not only changing damage/health values and 'elements' like RPGs usually do, so the game would need a lot of mobility variation and cool skill effects to make them really stand out from each other.

I always try to fit my controls in a Dualshock joystick even if I dont intend for the game to have joystick controls, just to keep the design intuitive and limit the number of buttons the player needs to memorize, which together with the buttons spent on basic functionality like jumping, basic attacks and interaction meant I only had a couple buttons for skills, so they needed to synergize well, do multiple things at once and have a lot depth. (No simple nukes or heals that waste a button on a single action, for example)

The characters being the focus of the game, I didnt want to spend too much energy on environments so I went back to my trusty procedural generation to give me some content for levels. I had this old system I did some time ago for generating tile-based dungeons door-to-door like Warframe that I really wanted to give a new try so I went for it with some very modular asset packs and managed a simple generator that creates branching arms from a center tile. The tiles being floating island are 'open', in the sense that they have no walls obstructing view or movement, which keeps the focus on the characters both visually and mechanically. More on the generator here.

The level design being mostly a bunch of branching paths with set-pieces at the ends could easily serve a hack-n-slash game, specially an endless one with focus on just executing on the core loop for fun, so I went with a dungeon-crawler design for the game. Levels are generated, you spawn at the center, explore the whole map defeating enemies and defeat a boss to advance to the next. The camera angle that I thought best served this dynamic gameplay was a faux-isometric, like an RTS or MOBA, fixed in order to simplify navigation. Both the camera angle, the depth of the characters and the simple controls kept pointing me in the direction of MOBAs so I decided to take them as inspiration for the character progression and skills.

MOBAs have just a handful of skills, where each does several different things at once and their synergy, specially between different characters, can bring out a ton of emergent gameplay with just a couple of actions, which is always interesting for an indie solodev without time to invest into content. I ended up deciding on 4 skills per character and a run-based level that unlocks them, MOBA-style, so that youd have time to experiment with each and understand how they work as well as giving the player some progression outside of the level loop. This also helped the game to settle as a rogue-lite, totally run-based where no character progression is kept outside of runs, which is something I usually do anyway.

With the controls being so tightly tied to a joystick configuration and the skills having a lot of depth and synergy, local co-op was a no brainer, specially for a dungeon-crawler.