Command

dodgem [options]

Options

  • -c, –create Create the evaluation database in MongoDB for the selected board size (-n).
  • -e, –evalmap Create the evalmap JSON.GZ from MongoDB and write it to –evalmap-path.
  • -l, –level INT CPU level for the first player (1–4). Default 3.
  • -g, –gote INT CPU level for the second player. 0 = use first player level; otherwise 1–4. Default 0.
  • -n, –num INT Board size (3–5). Default 4.
  • -p, –play Play games (requires -l and optionally -g and -r).
  • -r, –rep INT Number of games to play in –play mode. Default 10.
  • -s, –status Show MongoDB status summary for the selected board size.
  • -t, –traverse [KEY] Interactively traverse the MongoDB tree from the given key. If omitted, starts at “ini”.
  • -v, –verbose INT Verbose level. Higher values print more info (0–5).
  • –mongo-server URI MongoDB server URI. Overrides config.
  • –evalmap-path PATH Path to evalmap JSON.GZ. Overrides config.
  • –gui Launch the Tkinter GUI.

Examples

  • Play 50 games on a 5x5 board at L3 vs L2: 
    dodgem -n 5 -p -r 50 -l 3 -g 2
  • Show DB status with more detail: 
    dodgem -n 4 -s -v 3

Notes

  • Level 4 uses MongoDB lookups during search. Levels 1–3 use depth-limited search and evalmap.
  • The default evalmap is bundled; you can regenerate it from your MongoDB via -e.