Skip to content

AI/Infra: Implement dynamic engine depth scaling based on opponent ELO #761

Description

@chinweobtagaz

Description

If a 1200 ELO player plays against Stockfish level 20, they will lose every time, resulting in a poor user experience. The engine must scale its difficulty dynamically.

What Needs to be Done

  • Map human ELO ratings to engine depth and skill parameters.
  • Implement an API middleware that dynamically sets the engine's config based on the matchmaking payload.
  • Adjust parameters like Skill Level, MultiPV, and Depth dynamically.

Acceptance Criteria

  • Engine plays at a roughly equivalent ELO to the requesting user.
  • Lower skill levels result in faster compute times (saving server resources).
  • The transition in difficulty feels natural, not just random blundering.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions