Until you know how to begin, it can seem like even starting a game is an impossible task. We're told to manipulate the mortals using the Calling button, but the feedback on what's going on is so poor, I didn't realise I had even done anything to begin with. The blue dot and lack of a black button on the character indicated that something might happen, but it appeared to me that you must first discover the person by scouting them before any calling could be done, and that was not made clear at all in the opening message.
From there flow general gripes with how information is provided to us with regards to character actions. Generally, any character actions are lost in the scrum of world actions (which, at present, we cannot influence) Perhaps splitting the feedback dialogue into two, one for character actions and one for world actions, would make the comings and goings of the actors more obvious. I didn't realise someone was starting a quest against me until I received a little message saying that they had failed, for instance. On the occasions that I did use an action against another character, I felt that there was very little feedback on what they were likely to attempt, and nothing at all on what they had actually done. While an element of randomness or uncertainty is desirable, the player still needs to understand the possible outcomes of their actions for the game to work. My friends had mixed feelings on taking up even more space with feedback screens, but perhaps this could later be rolled into the quest dialogue (plus the ease-of-coding that might be gained from framing everything within a quest).
On the subject of character interactions, (and this is a minor note) each person needs to possess a tangled web of relationships. While hate is the only one that exists currently, the current single-target hate doesn't provide enough opportunities. Let each person hate two or even three other people. When relationships and character actions evolve, it might be handy to assign these feelings a value, either internally or exposed to the player, which indicates the depth and thus, the possible actions arising from, an attempted character action. I put this here because if it's going to be shown to the player, the UI will need to incorporate a potentially long list of relationships, each with their own values.
Clues are also an annoyance, in that the game clearly tracks them, while the player cannot. An obvious entry somewhere on the city screen would be helpful in helping us determine where to focus our efforts, and as the districts get fleshed out, they might need to take on the roll of clue-holders, rather than the city at large. This would enable players to install an antagonistic ruler of a district in order to keep prying eyes away. Just a thought.
The impact of a Rebellion or Loyalist win is.. nothing? I haven't worked out if this is an issue of it lacking content, or an issue of the player not being informed what the differences are. Of course, this too will need to be made more obvious as complexity grows, especially if regime changes become a more natural element of character interactions with juntas and assassinations taking the place of never-ending food riots. If even these could be represented within the quest framework, we would have a strong idea of the causes and outcomes of these sorts of city/world actions.
So, I've spoken a lot about the "quest framework", so I'd like to elaborate on what I mean. This is a mix of how I'd like to see actions presented, and the means by which they are undertaken. I envisage a popup box for your interactions with the world, in which each element of the world may play a part. Essentially, when you want to make use of something you would select it as the prime mover of a quest. Then you would select a target (a city, a district, a character, a clue or even a less concrete thing, like the rebellion currently underway, or the region in which your foes have marshalled an army. You could then add additional resources, either further pawns, money, eldritch power or anything else in the game world over which you have some modicum of control. The opposition would depend on the target and various circumstances, such as the presence of an army, or the current Doge of the city in which you take your action. For instance, if you want to institute the usurpation of a local Doge in order to place your favourite minion, Valerie, (a dyed-in-the-wool politico) at the head of the table, you would first open the Quest Dialogue, place your devious, political minion to be elected (or depending on how complex this gets, someone else who will install your favourite minion) in the actor box, and the current Doge character, Doge Vincent (or again, complexity allowing, the very position of Doge) in the target box. The Vincent's wealth would fill one defender box beside him immediately, so you react by placing the angry peasants (that the current food shortage has generated) in an attacker box beside Valerie. The Doge pops the mercenaries he hired as a result of a previous action, in another defender box beside him. Doge Vincent, sadly, has more than just positive resources on his character card - you select the Scandal you spent a bit of time and money creating (the result of a previous action you took with a local harlot as the actor and the Doge as the target) and slot that in your third attacking box. The Doge has no further options, as you've spent a while slowly eroding his control over the city council, and just for good measure you put one of those very council members in one of your boxes - you could add more, but why expose your precious resources to risk when you've probably already got this. Sitting in town there's a rebellious cleric, but the Doge doesn't use him and neither do you - he's too anti-establishment to be helpful to the Doge and he's too wary of the insidious evil that your (unfortunately revealed) politician agent works for to be anything but a danger to your plans.. for now. Had your politico avoided a previous quest which unearthed her as an agent of the darkness, she could have made use of this wayward clergyman without risking her plans, and with additional actions later he might be swayed or tricked to your benefit. Obviously in a straight-up fight Valerie would've gotten her face done in - The mercenaries are much better fighters than peacekeepers, and the manufactured scandal would have been of no use at all. Add to that that the Doge is famous for his swordsmanship, and that Valerie has a flat 0 in a fight, and perhaps Doge Vincent could've averted this whole crisis with a spot of martial law (after all, evil prospers when good men do nothing), or perhaps the repurcussions would have only given you another opportunity further down the line, with Doge Vincent having acquired the "Tyrant" trait.
I think that unifying all quests, events and action into this single Quest Framework would ensure readability and provide an easy way to implement and represent the conflicts that go on during the game. In this manner the heroes attempting to uncover a clue can be easily handled by assigning one hero as the actor and his party members as attacker resources with the clue as the target, and whatever you happen to put in their way being the defenders. The rebellious peasants would work much the same way - angry peasants as actor, the district, character or city being their target and the various circumstances of the game being played either as attacker resources or defender resources with an outcome decided by the appropriate values of the resources on each side and a small amount of RNG.
For visualisation, since it was a bit of a mess up there, I've attached a little Paint drawing for your viewing pleasure. Obviously I'm a shit artist.
One of my friends pointed out that this setup seems very Cultist Simulator-y to him. Fair point.