Start here
Armies research notebook
Armies is a Stratego engine, a game server and a ladder of players. The engine plays the competitive rules at 84 million random moves per second on this workstation's 32 threads. The players climb from random play to the method of the 2025 Ataraxos paper, which beat the most decorated human player 15 to 1 with 4 draws after training on 16 H100s for a week. The question here is how far a single machine with one GPU can get along that path, and which ideas transfer: learned setups, advantage filtering, dynamic damping, learned beliefs, test-time search, or DeepNash's regularised dynamics instead.
The ladder
| Rung | Player | Promoted when | Status |
|---|---|---|---|
| 1 | random, greedy | The baseline arena runs and greedy beats random | Complete: greedy 90% against random |
| 2 | expectimax over sampled hidden configurations | Beats greedy on 200 paired games | Complete: 69.5%, promoted; the sweep keeps 16 worlds at depth 2, and belief rollouts (19% against it) stay a baseline |
| 3 | ntuple learned leaf under the same search | Beats the hand-written leaf | Planned |
| 4 | Ataraxos pipeline: setup, move and belief transformers with damped self-play and search | Smoke run beats greedy, then scaled runs beat each lower rung | Smoke complete: pipeline works, 60% against random, rule not met |
| 5 | DeepNash's R-NaD update under equal compute | Compared head-to-head with rung 4 | Planned |
| 6 | kluss subgame solving without common knowledge, then an opponent portfolio and a learned belief | Beats expectimax at equal time per move; the portfolio's crossing budget; the belief's likelihood and match | Proposed, delegated |
| 7 | A NashDreamer infoset model as a belief | Within 5% of the belief decoder's held-out likelihood | Proposed, bounded |
Research ladder standings
| Player | Elo | Interval | Games | W / D / L |
|---|---|---|---|---|
| Expectimaxexpectimax | 1535 | 1 to 1 | 460 | 343 / 7 / 110 |
| Greedygreedy | 1393 | 1 to 1 | 540 | 303 / 8 / 229 |
| Rolloutrollout | 1283 | 0 to 0 | 100 | 19 / 0 / 81 |
| Ataraxos smokeataraxos-smoke | 1090 | 0 to 0 | 80 | 18 / 26 / 36 |
| Randomrandom | 1000 | 0 to 0 | 700 | 127 / 219 / 354 |
| N-tuple leafntuple | 0 | 0 to 0 | 0 | 0 / 0 / 0 |
Read the concepts
Start with why Stratego is hard and the rules the engine enforces, then the engine. The search rungs are determinization and expectimax and n-tuple evaluation. The paper's method runs from information states through self-play reinforcement learning, dynamic damping, the setup, move and belief networks, and test-time search. The alternative is DeepNash, and evaluation explains how both papers and this ladder are scored.
Approaches
Each rung has a page with its question, plan and decision rule: baselines, expectimax, n-tuple, Ataraxos, DeepNash, subgame solving without common knowledge, world models, the pipeline on one machine and initial placements. The research program orders these into theories with decision rules and a compute plan, and the evidence boundary says what is measured today, and game records and replays explains the replay library the site plays back.
- 2026-09-18complete
- 2026-09-18complete
- 2026-09-18complete
- 2026-09-18complete
- 2026-09-18complete
Five registrations frozen and run on September 18, 2026; each links to its report and has a run record.
Run experiments and contribute
Build and test the engine with cargo test --workspace in ../server.
Play a registered match with the arena, for example
cargo run --release -p armies-arena --bin matchup -- --red greedy --blue random --games 200 --rules training --out ../research/runs/baseline.json,
then write the record and report as described in filing an experiment.
Training commands are in the quickstart. Validate
pages and assets with python3 scripts/check_content.py before handing over.