So I got to thinking (which usually only ends in much disappointment) about what makes a good strategy game. Detail? Yes. Fidelity is a must. Unit variety. It can't just be copy pasta with one unique unit per side. But, scale. One would think that the bigger the better, yes? Lest things become a Starcraft spam fest. Yet that isn't the case - games like Supreme Commander show that once you get to that number of troops onscreen things get unmanageable without Korean level micro. You end up just mindlessly flinging your guys at the enemies in hordes at a time, crossing your fingers that they'll come on top. Even in 'realistic' titles like Wargame there aren't really any tactics involved, so this flaw becomes exceedingly apparent in an AOE type game like Glest.
Alright, so how could that be rectified? Well, why not focus on the macro level of things, as per real militaries? For example, an officer would not charge headfirst onto the battlefield and delegate orders to the lowliest of privates. He would instead hang back, observing the overall tactical situation and dictate commands to his subordinate officers. They will then comply and relay their instructions to their own underlings as necessary. So let's say a player in Glest/Megaglest/GAE/what-have-you is that officer, and while he always has the ability to instruct units directly, it'd be easier to appoint a subcommander to a control group (CTRL+#) and issue orders to a group - for example, 'hold position', 'assault', etc, and have an AI do it's damnedest to figure out what you want and carry it out. The glorious comrade leader can now relax and focus on building up his base, only pausing to reinforce his control groups as necessary instead of frantically clicking about.
My question is whether this is even possible - to pretty much give the player a subordinate AI responsible for a selection of units, yet also strictly adhere to any 'stances' given to it. Would you need one for each control group, at that?
