Deadlock Dev Uses ChatGPT to Help With Matchmaking Code

Valve's upcoming MOBA-hero shooter, Deadlock, recently overhauled its matchmaking system, employing an algorithm sourced from the AI chatbot ChatGPT. This revelation, shared by Valve engineer Fletcher Dunn on Twitter (X), highlights the evolving role of AI in game development.

Deadlock's Matchmaking Overhaul: A ChatGPT Success Story

Deadlock's previous matchmaking, based on a Matchmaking Rating (MMR) system, drew significant criticism from players. Reddit threads showcased widespread frustration with uneven team skill levels, frequently pitting experienced players against less skilled opponents. One player commented, "I get harder games with better enemies, but never equally skilled teammates." Another echoed this sentiment, noting the disparity between experienced and novice players within matches.

Responding to this feedback, Deadlock developers announced a complete matchmaking system rewrite. Dunn's Twitter posts detail how ChatGPT suggested the Hungarian algorithm as a solution. He expressed his reliance on ChatGPT, stating, "I have a tab in Chrome reserved for it, always open." While celebrating ChatGPT's utility, Dunn acknowledged potential downsides, noting it might reduce human interaction in problem-solving. This sparked debate, with some users expressing skepticism about AI replacing programmers.

The Hungarian algorithm, a type of bipartite matching algorithm, is particularly suited to scenarios where only one side (in this case, players' preferences) needs to be considered for optimal matching. This contrasts with traditional MMR systems, which may not adequately address team balance issues.

Despite the improvements, some players remain unconvinced, expressing continued dissatisfaction with the matchmaking. Negative reactions to Dunn's tweets underscore the ongoing challenge of balancing player expectations with rapid development cycles.

Here at Game8, we remain optimistic about Deadlock's potential. For a more in-depth look at our playtest experience, check out the link below!