Software Engineer | Robot Race

What I did — repair the multiplayer race loop

As a software engineering intern, I investigated post-race transitions and AFK detection in a multiplayer racer. I traced race-completion events to fix a failed transition, and changed inactivity detection from movement distance to player input so using the settings interface would not be mistaken for being idle.

What I think — stability is part of fairness

In a competitive game, an incorrect penalty or a broken transition changes the experience as directly as a balance decision. Playtests helped me connect a visible symptom to the underlying event logic, then define the behavior around what the player was actually doing.

Design process

As a Software Engineer Intern, my primary responsibility in Robot Race has been ensuring multiplayer stability and network reliability. Since this is a fast-paced, competitive racing game, even minor connectivity or synchronization issues can significantly impact player experience. My role involved identifying and resolving critical multiplayer functionality issues, improving game flow stability, and refining AFK detection mechanisms to prevent unfair disconnections.

During an engineering team playtest, we encountered a game-breaking issue where the post-race settlement system failed to transition players correctly, preventing them from moving to the next phase. Additionally, the game’s AFK detection system occasionally misidentified players as inactive, triggering penalties even when they were engaged in actions such as adjusting in-game settings. These issues disrupted the intended gameplay loop, requiring immediate debugging and refinement.

To address the settlement transition failure, I analyzed the relevant network scripts and event-handling logic, identifying inconsistencies in how the game processed race completions. By modifying the underlying script logic, I ensured that the transition now executed seamlessly for all players under different network conditions.

For the AFK penalty system, I redesigned its detection method to shift from distance-based evaluation to input-based tracking. Originally, the system relied on movement distance to determine player inactivity, but this led to false positives when players paused to access in-game settings. By refining the system to detect button inputs rather than movement alone, I significantly reduced incorrect AFK detections, allowing the system to more accurately distinguish between truly idle players and those engaging with UI elements.

This experience strengthened my ability to debug real-time multiplayer systems, optimize server-client interactions, and refine game mechanics to enhance player experience. Beyond technical implementation, it reinforced the importance of playtesting-driven problem-solving—leveraging direct gameplay observations to inform data-driven fixes. As development progresses, I continue working on improving multiplayer responsiveness and stability, ensuring that Robot Race provides a seamless and fair online experience.