Game dev process

Overview

I wanted to talk a bit about my process and general schedule when it comes to making a game. This will likely evolve over time, and it’s probably also somewhat personal to me, my goals, and my personality.

Starting a new game

The broad overview of how I like to pursue games is:

• Spend 1-2 weeks in documents planning the game design, figuring out the story, and hammering away at any questions. A lot of this follows my notes in General game design#Designing a brand new game.
• Pursue the checkpoint system (see Checkpoint system)
• Stabilize between checkpoints (see Stabilization phase)
• Prep for launch (see Prepping for launch)

Checkpoint system

Checkpoints are small-ish (1-2 months) chunks of work. The general idea is:

• Plan a checkpoint to be the next playtestable unit of work.
• Run playtests at the end.
• Process the feedback from the playtests to be able to plan the next checkpoint.

By doing things this way, you are constantly getting validation about your product.

Stabilization phase

While pursuing checkpoints, you’re probably building up some tech debt, polish-level tasks, etc. I’ve personally found that it’s good to dedicate time to tackling these tasks.

• Pros
  • It spreads out the “un-fun” work.
  • It helps you cut through your issues and TODO lists, which may make you feel less overwhelmed.
  • By dedicating the time:
    • You won’t race through the tasks to get back to the fun stuff.
    • You won’t need to context-switch constantly between this and regular dev work.
• Cons
  • If you don’t pursue the game as a product, most of this would probably become wasted work. E.g. imagine fixing a bug in a game that you totally throw out.
  • It’s probably not the most sound business decision. Many low-pri tasks can accumulate in a game and without really affecting the product.

In general, if a checkpoint is X weeks long, then I think stabilization should be about ¼X weeks. However, it might be worth spending a bit longer at the beginning of the project since you may want to make some video content, create a Steam page, set up a Discord, domain, email address, etc., and all of those take time.

As an example of the kind of work I pursued for the first stabilization fortnight of Stuck in a Broken Tutorial:

• Laid entire localization foundation. This involved writing new tooling since I was using a dialogue system not present in Skeleseller.
• Investigated an incredibly tricky crash in Godot
• Tackled random bugs and TODO items in the rest of the time

Prepping for launch

• At some point, you probably have enough general playtesting feedback and can start polishing everything, marketing, improving your Steam page, etc. I don’t have a hardened process for this section, so I just wanted to mention what this phase is about.
  • Note that playtesting is still important, especially with new people, and especially with people who have no stake in you or your game, that way you get honest feedback from “real” customers.