Community playtest · Released 22 October 2026
Battleward Table — Public Beta 0.1
A virtual battlefield featuring a human-supervised AI opponent for Drums at Dawn.
Create a battlefield, move your units, measure distances, and try the developing solo opponent. This is an experimental release: we’re inviting players to help find problems and improve the game.
Before you play
- The AI needs human supervision. Some compulsory effects, routs and complicated situations require a player ruling.
- Bugs and unfinished behavior are expected. Ten complete human-supervised acceptance games have not yet been verified; this is not a finished or certified autonomous opponent.
- The AI does not learn from your games or retain tactical lessons. Shared learning and player-specific tactical memory are future work.
- The engine is v6.0.11, using the Drums at Dawn v11.2 Beta integration. Other rules PDFs on the site may show older versions; do not mix editions.
Getting started
- Open the beta. The battlefield is immediately below Quick Play.
- Choose a demonstration from Table setup, or open Load Game beside Quick Play to resume a battle. Use a complete game JSON, rather than a map-only file.
- Open Help / Controls beside Quick Play for the simple manual, including moving units, editing terrain, turns, AI controls, saving and loading.
- Use − and + for deliberate zooming, and hover over a unit to see its name.
- Follow the phase and AI prompts. Resolve any referee-required instructions before continuing.
Keep a backup
Browser saves belong to the browser and device you use. Export a complete saved-game JSON regularly, especially before resolving a difficult combat situation. Clearing browser data can remove local saves. A save made in a downloaded prototype may need to be exported and opened again on this website.
Help us improve it
Tell us what you tried, what you expected, and what happened. Include Beta 0.1 / engine v6.0.11, your browser and device, the turn and phase, and the units involved. Keep the saved-game JSON and a screenshot if they help reproduce the problem; remove personal information before sharing.