Decentralised Learning in Systems With Many, Many Strategic Agents