Indeed, one of the big advantages of IOT over video games is its freeform capacity, things that escape easy quantification and so a computer can't really do them. Long ago I - as many others - realized that ultimately, this subforum should play to the advantages of the format, rather than try to emulate a Paradox or Civilization title. Mechanics that serve to structure rather than be the alpha and omega are the goal. The advantages of the format are freethinking GMs who can use their fiat to keep things from getting too crazy, freethinking players who are both cooperative storytellers and clever rivals, and above all the uniqueness that human interaction can yield compared to the dry, scripted behavior of AI.
Of course, it's a theoretical advantage. A game can be fantastic with a very disciplined GM and structured yet not omnipresent rulesets, or become infamous when failing on these measures. I'd know that better than anyone.
I'm not surprised people choose the predictability, if often broken predictability, to the broken unpredictability of human GMs.
***
Meanwhile, not to toot my own horn too much, but that ties into some things I'm trying to capture with the new Multipolar Politics. Spending is largely automated now, so most of the turn by turn gameplay will be on policies and diplomacy.
With regards to war, conquest is extremely difficult, particularly against another player, so I expect to see it in a more limited fashion than other games I've held. I have often heard many players say that the threat of war is actually more exciting than war itself, so I'm hopeful the military mechanics leave war as a possible menace, but one that rarely breaks out.
And, handcuffing myself for the benefit of the player base, NPCs are almost entirely reactive to player-initiated actions towards them. Using modifiers to model pre-existing relationships and national values with dice rolls that mitigate any unconscious biases of mine, I'm hopeful to deliver a somewhat less crazy experience than the original Multipolar World.