I would like to be able to program governors and save them as well as assign them avatars and have a track record or historical one follow them from game to game.

Please don't overdo the governor stuff. I can't bear managing governors in the same than in the Total War games (4x + Sims).

I would like NPC/AI governors in an RPG/4X game where they had real personality, dialogue, story developments etc... and I posted about doing that earlier but the more I think of it I just don't think that such a degree of RPG could be worked into this without hijacking the straightforward 4X goal.

I do like the idea of units/heros/whatever that provide bonuses to cities where they are stationed or nearby, and those units could be, ah, "persuaded" to cease to exist like any other, but I don't know how much further I'd go than that.

Assign them avatars, follow them through games... sure, sounds cool.  But I don't want to have to pay attention to an AI personality in each city or region of cities.  If "things don't work right" because of characters at that level then managing an empire on a really massive map could bog down.

Maybe there's a way without hitting the issues I'm thinking of, though.  Stardock has surprised me before.

I'm with the dissenting opinion here. Whenever I have the option to allow Governors to manage cities/colonies/planets/deli shops for me, I always turn it off. I'd much rather take the time to micromanage everything myself than to have one of my key cities not producing what I need it too because I'm only limited to generic build queues like 'military' or 'civil' or 'whole wheat'. Gah! No one wants a whole wheat pastrami sandwich!

I can see governers taking something like the grand vizer did in MoM.   Just an AI that manages your building.   Nothing more, and that would be fine.

Aren't you forgetting that the grand vizier in MoM was horrible? If you turned him on, all of your towns will act just like the rest of the computer controlled towns. If I could get him to stick to buildings that would have been perfect. But leaving him on always resulted in huge amounts of redundant troops. I could garrison a town with 6 regiments of elite longbowmen and he would still build basic cavalry instead of a shrine.

Along with "template" governors we should be able to customize our own. We must teach them what we wan't them to do, but only once. So at the beginning you could have a "colony" governor who builds a preset list of buildings ending in defensive walls. Once that is done he either turns himself off or changes into a "fronteir town" AI who upgrades the defenses and builds garrison troops. You could assign tasks in chains such as "do this and this and this then inform me when you are done" or "do this and this and then raise this army and send the troops to the next town over". You could also assign standing orders for governers of different types to automatically change into others when certain conditions are met.

I could have explained that better but hopefully you got the gist of it. This type of control is macromanagement, it works more like directing an orchestra composed of many instruments. You provide each musician with a different sheet of music, but on concert night you only have to wave your baton to set the rythm with which they play.

Compare that to redundant micromanagement which is akin to playing the piano with your hands and an accordion with your feet at the same time.

Standing orders could also apply to unit production. For example:

Order 31 is as follows: "each town must raise and maintain a single unit of spearmen not to exceed 10% of the total population. Up to half of the soldiers may be militia if there are not enough workers to fill the current production quota. They must be armed with equipment up to the current standard, if the resources required are not immediatly available then the next best local option will be taken until supplies can be brought in from other towns"

What this means in terms of game mechanics is this:

1. Order 31 is a standing order for every town you own to maintain a standing army of spearmen. If your town has a population of 2000, then soldier will be drafted automatically until the number reaches 200.
2. Up to half of the soldiers recruited can be militia. What this means is that these soldiers will go back to thier civilian jobs when required. Since they don't spend as much time on the training ground they are not as experienced as full time soldiers. But on the other hand since they must spend a minimum of time on the training ground, they are not as efficient as full time workers.
3. Normally idle civilians who don't have permanent jobs are recruited first. If the quota is not met then your active workers will be enisted as militia. So under Order 31 smaller towns without a worker surplus will rely heavily on militia while large well developed cities would have none unless you upped the percentage.
4. The current standard of arms in this example is a shirt of maille armour, helmet and spear. The best material you have access to is steel but steel is rare and expensive, so iron is the common standard.

This would play out in a predictable manner all over your empire. Lets say you founded a new town on choice land that is cut off from your others by a dangerous river. Cargo shipping over the river is impractical and can't handle heavy items, going around takes a long time. The new town doesn't have the industry to supply iron and experienced armourers to make maille. Since they can't go unequipped for the time it takes to build a bridge over the river, they fulfil Order 31 by using local bronze instead of iron and leather armour.

Now this might be a long and convoluted example but it is one of many ways we could condense tasks which make micromanagement a slog.

Here's the deal, though.  If the AI could manage your towns, it should be the same AI that the computer uses to manage its towns (give or take some additional intelligence depending on your difficulty level chosen).  If the management AI sucks, then it does for the computer opponents, which means there's an AI coding issue that needs to be addressed.

In essence, we could help trouble-shoot the governor AI by pointing out issues that would be more difficult to see if it was just the computer managing it.  Not only would we get a better governor, but the computer opponents would be better and at worst, just make it an opt-in system so you can still do your own micro-management if that's what you find enjoyable.

If you try to use player interaction to improve the behavious of AI opponents then that wouldn't work. The only way to truly test the AI is to give it total control and observe how successful it is based on the results of its own descisions. If it is acting "stupid" then you need to give it access to more and or better information so that it can make better descisions. If you have any hand in the process whatsoever then the end result can only be in essense a stupid version of you. You are modifying the AI to behave in the way you yourself consider intelligent. If you take that process too far then you no longer have a true computer opponent, smart as it may be.

Not that this is a bad thing! Far from it. Imagine the flavour and depth we could add to the system if we could create whole custom races and empires and the AI to control them for us by proxy.

I like to Micro my towns until the game enters the mop of phase.  At that point I would like to use some sort of governor.  A script controled governor AI that we can edit would be best.

Sammual