I am writing a grand strategy game that is generally played from a zoomed-out, "strategic" perspective. With battle encounters, I want players to be able to zoom in, if they choose to, letting them see a "tactical" map with individual units moving about on a 2D battlefield.

Think Total War, for argument's sake.

Now there may be several battles happening in parallel in different locations of the world. The tactical view, moreover, will be quite heavy on AI (for pathfinding needs and other considerations), as well as running at a more granular clock speed (minutes in the tactical view vs hours in the strategic view). It is planned to be relatively compute intensive.

The question is how to resolve unseen battles - those battles that the player does not choose to view, or cannot for FoW reasons. One option is to create a second, less complex, abstract mechanism that resolves unseen battles using fewer computations (akin to Paradox games), and then to tune both mechanisms so that they give comparable statistics. Is this a viable strategy?

My original idea was to run the tactical AI in the background for every battle that occurs, but I suspect that will not scale up very well.

(Since I mentioned them already, does anybody know how the Total War series does it?)

  • \$\begingroup\$ Auto-resolving tactical battles is no exact science. Lots of games have auto-resolvers with very weird anomalies. Those mostly annoy most players by subverting their expectations, but occasionally allow some exploits. \$\endgroup\$
    – Philipp
    Aug 22, 2022 at 12:07
  • \$\begingroup\$ So I'm assuming they do use simpler statistical models to auto-resolve? \$\endgroup\$ Aug 22, 2022 at 12:21
  • \$\begingroup\$ Depending on how your combat mechanics work, how detailed the simulation is and how much influence the player has over it, there is often no other way. \$\endgroup\$
    – Philipp
    Aug 22, 2022 at 12:50


You must log in to answer this question.

Browse other questions tagged .