Postmortem: They Have My Daughter

They Have My Daughter is a "choices matter" visual novel with eldritch horror, cults, and a touch of Hatchetfield weirdness.

All right, in all honesty I didn't wake up this morning with the goal of writing a postmortem anywhere on my schedule, but I did somehow wind up stepping onto the bookmobile here in my small town today, and checked out Single Player by Tara Tai wherein by page 3 the main character starts working on a 'choices matter' style game... And that inspired this. So here we go!

What Went Right

1. First-Time Successes

   - This was a project of *many* firsts: my first Godot game, first visual novel, first time vibe-coding, and first time integrating AI-generated art. Despite the learning curve, the pieces came together cohesively.  

   - The branching narrative with multiple endings worked better than expected, especially for a debut in the genre. Players could appeal to authorities, rally the public, or go rogue—and each path *felt* distinct.  

2. Hatchetfield’s Unexpected Role 🪓

   - What started as a cheeky reference (using the Hatchetfield map for the Nexus) evolved into a fun fanon crossover. While "The Tide" is more Lovecraftian than Starkid’s "Starry Children", the shared universe hints (especially in certain endings) added depth without compromising the tone. Thanks, STARKID, for making such a wonderful, horrifying world! (I may have 5 specific new endings planned for v1.2... You can imagine!)

3. Atmosphere & Scope

   - The beachside mansion, the cult’s eerie presence, and the looming elder gods created a strong vibe. Restricting the core setting to a few key locations helped maintain focus despite the branching paths.  

What Went Wrong

1. Development Time (of course)

   - Estimated: *"A weekend and a half."* Reality: **A month and a half.**  

   - Writing (and rewriting) branches bloated the scope. Tracking variables for endings became messy, and debugging dialogue trees was a time-sink.  

2. Overambitious Systems

   - The "3D stat system" I wanted initially (karma/tact/ideology) sounded elegant but proved clunky in practice. Instead of organically shaping endings, it often felt like brute-forcing outcomes. I ended up scrapping it, and going with endings driven purely by the players choices and available inventory. Next time, I’d simplify or prototype mechanics earlier. Maybe I can employ something like this in a future game more suited to the idea...

3. Godot Learning Curve 

   - Godot's "scene elements" tripped me up initially—I treated them like Unity’s scenes rather than Unity’s prefabs: modular, self-contained objects. Later, I realized their flexibility (e.g., as reusable UI elements), but early inefficiencies cost time.  

Lessons Learned

1. Branching Narratives Are Deceptively Hard 

   - Even "small" choices compound exponentially. Tools like Twine or yarnSpinner might’ve helped plan branches before implementing them in Godot.  

2. Prototype Mechanics First 

   - The stat system seemed cool on paper but wasn’t fun in practice. Future projects need playtesting *before* full integration.  

3. Embrace Constraints 

   - The Hatchetfield reference forced creative adaptation, and it worked! Limitations (like reusing assets or maps) can spark innovation.  

If I Did It Again Today…

- Use Godot’s scenes more effectively (e.g., modular dialogue components).  

- Simplify stat tracking or tie it to clearer narrative beats.  

- Budget 3x the expected writing time—branching stories *always* grow.  

Final Thoughts

They Have My Daughter was a chaotic, rewarding experiment. It’s janky in places, but the atmosphere and player agency shine. The Hatchetfield connection was unplanned but now feels inevitable—like the cult’s influence seeping into the code. 😵‍💫🤐

Would I do it again? Yes, but with fewer stats and more coffee. ☕☕☕

(PS: This is not the Best ending, but it's quite a good one)