KING PONG
A custom nodal editor, data-driven combat systems and network optimisation on a VR table-tennis game.
The game
KING PONG is a VR table-tennis game by Iconik, published with Plug In Digital and released in February 2021. Return every ball to score, use special power-ups called Ultimates to change the rules mid-rally, and play solo arcade modes or competitive online multiplayer in a Tron-styled arena.
My contributions
- Developed the AI behavior trees and designed a custom node-based editor for them, so opponents could be tuned without going through a programmer.
- Developed the “Ultimate Attack” combat systems using finite state machines and a data-driven approach.
- Optimised network replication through Mirror and PlayFab for standalone headset performance.
Why it was interesting
A table-tennis AI has to be beatable in a way that feels fair, which is a design problem before it is a programming one. Hence building an editor designers could drive themselves, rather than hard-coding difficulty. The data-driven approach to the Ultimates paid off for the same reason: new attacks become data, not a code change and a rebuild.
The networking side had the tighter constraint. The Quest is a mobile chip rendering two eyes at 72 Hz, and replication competes directly with that frame budget, in a game where the entire experience is a fast-moving ball whose position both players must agree on, instantly.