How many teams should there be (2, 3 or more than 3)?
This is important because it affects the dynamics of the game. Having 2 teams in competition seems most intuitive, but having 3 or more teams may introduce more complex dynamics such as temporary alliances and deterrence, which may also be a source of suspense and fun. To test this, we could build a prototype for 2 teams and a prototype for 3 teams (each team could have 2 people) with a simplified version of the rules. My guess is that 3 teams would be preferred because it introduces so much more strategy that would not be present in a 2-player version, adding more complexity to the game.
Should players on a team have uniform roles or specialized roles?
Again, game dynamics will be affected. Uniform roles places all team members at a fundamentally equal standing, but having specialized roles (e.g. a “general” who holds the ultimate authority on whether to launch a nuke, an “ambassador” who is the only player that can communicate directly with the other team) can inform strategy and also introduce more potent challenges to the game experience. Specialized roles might create imbalances though (perhaps some roles are more important, or more boring to play, than other roles). To test this, we could build a prototype with roles and one without roles, and see which is more fun to play. My guess is that uniform roles would be better, because the burden of decision making is shared more equally among team members, and also because the rules would be simpler (which is desirable for a party game because it’s easier to teach).
Should there be a time limit for deliberations? If so, how long?
Having a time limit or not (and the duration) could potentially make or break the gaming experience. If players do not feel the time pressure to decide whether or not to launch a nuke, it detracts from the immersive thematic experience of the game, and may also cause the game to drag on. However, if the time limit is too short, players may not be able to come to informed decisions and much of the game result will be left up to chance instead of skill (which is probably undesirable). To test this, we could adjust the deliberation times for various rounds using a timer to see what amount of time is most enjoyable. My guess is that deliberations that last around 30 seconds to 1 minute would be most ideal.