Records
Baseline arena: greedy against random
Registration 9aa52df6-f9fe-47f5-beab-760ab04c2008 (study baselines/baseline-arena), run 2026-09-18 on AMD Ryzen AI Max (Strix Halo), 32 threads, 125 GB unified memory, Radeon 8060S via ROCm, Linux 6.18. Server commit 7a66da0, research commit a000925 (registration files as committed; this record added afterwards).
Commands:
cd server && cargo run --release -p armies-arena --bin matchup -- --red greedy --blue random --games 200 --seed 0 --rules training --threads 32 --out ../research/runs/9aa52df6-f9fe-47f5-beab-760ab04c2008/greedy-vs-random.json
cd server && cargo run --release -p armies-arena --bin matchup -- --red random --blue random --games 200 --seed 0 --rules training --threads 32 --out ../research/runs/9aa52df6-f9fe-47f5-beab-760ab04c2008/random-vs-random.jsonGreedy against random
| Player | Wins | Draws | Losses | Effective | 95% Wilson | Mean plies | ms/move |
|---|---|---|---|---|---|---|---|
| greedy | 180 | 1 | 19 | 0.902 | 0.854 to 0.936 | 219 | 0.001 |
| random | 19 | 1 | 180 | 0.098 | 0.064 to 0.146 | 219 | 0.000 |
| ending: flag_captured | 190 | ||||||
| ending: no_moves | 9 | ||||||
| ending: quiet_limit | 1 |
Per-seed paired contrast for greedy: mean 0.805, 95% normal interval 0.706 to 0.904 over 100 seeds (86 positive, 9 split, 5 negative). Wall time 0.0 s.
Random against random (companion)
| Player | Wins | Draws | Losses | Effective | 95% Wilson | Mean plies | ms/move |
|---|---|---|---|---|---|---|---|
| random | 46 | 96 | 58 | 0.470 | 0.402 to 0.539 | 569 | 0.000 |
| random#2 | 58 | 96 | 46 | 0.530 | 0.461 to 0.598 | 569 | 0.000 |
| ending: flag_captured | 98 | ||||||
| ending: no_moves | 6 | ||||||
| ending: quiet_limit | 96 |
Per-seed contrast for the first random: mean -0.060, 95% interval -0.176 to 0.056. Wall time 0.0 s.
Provenance
Archived JSON: runs/9aa52df6-f9fe-47f5-beab-760ab04c2008/greedy-vs-random.json (SHA-256 2f70f9e51914efc2eab505826ed1244db905c9cfa9682fcc28328cf0ba6efebe) and runs/9aa52df6-f9fe-47f5-beab-760ab04c2008/random-vs-random.json (SHA-256 caabc62f668a985f0c20bcbd059b8b17888b31439577afe121aba369b0d3ca1a); copies next to this record. Games planned 200 + 200, attempted 400, completed 400, interrupted 0. --seed 0 --games 200 plays seeds 0 to 99 twice each with the seats swapped, as registered. No deviation.