Last Bell Postmortem (Week 13)
What Went Well
1. Core Mechanics and Feel
The grappling hook mechanic was the standout feature. Most playtesters found it satisfying and fun, and its inclusion added verticality and creativity to movement. The combination of sword slashing and fireball casting gave players tactical variety and agency, especially once they became familiar with the controls.
2. Art and Atmosphere
Despite limited resources, the game succeeded in delivering a cohesive visual theme. The gothic visuals, stylized characters, and dark tone contributed to a consistent aesthetic that players found memorable and tonally appropriate.
3. Playtesting and Iteration
Early and frequent playtesting allowed us to identify key issues with controls, tutorialization, and player onboarding. This iterative process directly influenced changes such as adding on-screen prompts, improving enemy feedback, and fixing grappling inconsistencies.
What Could Have Been Improved
1. Tutorialization and Onboarding
Initially, the game lacked any form of tutorial or explanation, resulting in confusion during early playtests. While we later implemented text prompts, a more structured and gradual onboarding system (e.g., a dedicated tutorial level) would have greatly improved first-time player experience.
2. Combat Feedback
Combat lacked audiovisual impact—hits didn’t feel weighty, and enemies didn’t clearly react to damage. Players often felt unsure if their attacks were effective. Knockback animations, enemy hit reactions, and better sound design would greatly enhance the tactile quality of combat.
3. Technical Bugs
The grappling system, while praised for its design, suffered from minor but noticeable bugs—such as inconsistent collision detection, camera jitter, and animation bugs (e.g., looping falling animations even while grounded). These bugs undermined the otherwise strong traversal system.
4. Enemy Behavior
AI behavior was basic. Skeleton archers occasionally shot in the wrong direction, and werewolves lacked variation or reaction patterns. A more nuanced behavior tree or state-based system would make combat encounters feel more dynamic and strategic.
What We Learned
Player expectations matter: Even short games need clear controls and onboarding to prevent frustration.
Feel is everything: Grappling and movement felt fun because we invested in smooth transitions and flexibility. The same philosophy should have been applied to combat.
Early iteration works: The issues we addressed early (e.g., traversal bugs, level readability) saw meaningful improvement because we tested often and adapted quickly.
Tools have limits: GDevelop was useful for prototyping, but we ran into engine limitations when implementing more advanced features like dynamic camera movement and AI. Future iterations might benefit from transitioning to a more robust engine if the scope expands.
Next Steps / Hypothetical Future Development
If Last Bell were to be taken further:
Develop a proper onboarding/tutorial level
Fully overhaul combat feedback with VFX, SFX, and responsive enemy reactions
Expand level design to include branching paths, secrets, and narrative beats
Improve AI and enemy variety (e.g., shielded enemies, flying enemies)
Compose an adaptive soundtrack that reflects player momentum and scenario tension
Consider switching to Unity or Godot for more scalability and polish potential