A tycoon game, you produce and sell a product to several regions.

I always hated distribution in those games. You made a factory in region A then make it produce 5 products and send those to region B,C,D,E. Then you make a new factory in region C which again produce 5 products and decide to which region they will ship it (now you have 2 factories and 5 regions), then you replace one product, so you have to adjust factories but then the production and distribution does not match because it has lower demand in region E and the capacity of the second factory is too low because it's not upgraded and... ARGH!!!! At this point the game turns out into a hell where I spend like 80% of my time adjusting wat to send wher (and usually I need to adjust it each turn). It was not what the game was meant to be! Distribution is soooo demanding (and boring), it takes over the whole game!

Ideas how to trim distribution? It could be more abstracted, automatized, whatever. As long as it does not make the player to focus on it (the focus should be on other, more interesting, parts of the game).

You can make it either like ancient Detroit game where Sale Offices can set 3 supply lines (actually recent Project Automata does similar if I'm not mistaken) and they supply in priority order. ( There may be also an option to set priority for supplying )

Or you can do it like Capitalism plus where you use supply units and link buildings, they get according to demand.

So at the end, you get a weighed return instead of a super micromanagement feature.

Ugh... Detroit was one of the horror stories I wanted to avoid :D Didn't like the distribution part there.

As for Capitalism, well, I kind of disliked it as well, and, there was on option to "automatize linking" which was a tempting option. So, in short the distribution there was regarded as boring as well and they automated it (nothing else had a big "do it automatically" button in that game).

To make an interesting resource management ("tycoon") game, the player should be tasked with doing something that cannot easily be automated by the AI: any task that can easily be optimised by game AI (especially game AI written in circa 1997, as with Capitalism) is likely to be tedious for humans, too. Ask yourself - what exactly is the domain of interesting decisions that the player will be making in this game such as to make it fun?

To make an interesting resource management ("tycoon") game, the player should be tasked with doing something that cannot easily be automated by the AI: any task that can easily be optimised by game AI (especially game AI written in circa 1997, as with Capitalism) is likely to be tedious for humans, too. Ask yourself - what exactly is the domain of interesting decisions that the player will be making in this game such as to make it fun?

Exactly. Preferably I would left out distribution completelly. But I need some sort of distribution because players would want several markets (cities/regions/continenst/whatever) and factories in several places :( So, I need to handle somehow the distribution part of the game (preferably in a minimalistic way).

The difference between shipping from different factories to the same destination is cost and delay. Cost differences would be modest and possibly negligible within the same class (train, ship, truck, plane...), delay differences could lead to interesting strategic decisions (e.g. place factories closer to important markets and/or resort to expensive air transport for part of long-distance orders). AI automation can easily plan and order the cheapest transport that delivers goods from point A to point B on time; unlimited offer of transport is a realistic assumption. The player would be involved only to warn when respecting contracts is expensive enough to incur a loss.

Handling of shipping would be different if you want to make a game (or a subgame) about troubled and troublesome truckers (tracked one by one, including truck characteristics and maintenance), about filling and commanding a container ship, or other adventurous and interesting kinds of shipping.

The difference between shipping from different factories to the same destination is cost and delay. Cost differences would be modest and possibly negligible within the same class (train, ship, truck, plane...), delay differences could lead to interesting strategic decisions [....]

Yes... distribution cost is, well, low overall (unless we talk about bulky super cheap goods, but I'm not making a "brick manufacturer simulator", so it's irrelevant :D) as for delays I can't (and don't want to :D) really implement those (would be too messy and would not fit the game or mechanics, unless we go for something super simple and abstracted). So... I wonder what's the point of this distribution thing... OK, I could artifically increase transport cost (as long as you don't make a cellphone sim where it would make no sense :D) so the player want to have factories on various continents...

But I still find it not right. I think the real cost of distribution is tarrifs. Like unless the company is registerd in country X (or some internation trade zone) they get tariffs imposed for their goods. But then, what's the point of building a factory on various continents then? Why not make one centralized factory and distribute the product worldwide? Again, we are talking about high value products (relativelly high tech, not extremelly bulky).

I'm really tempted to make a single factory and then impose tariffs when selling to other regions (plus marketing costs, sales infrastructure, etc). But then players would surely complain they can't build factories in other regions :D

To make an interesting resource management ("tycoon") game, the player should be tasked with doing something that cannot easily be automated by the AI: any task that can easily be optimised by game AI (especially game AI written in circa 1997, as with Capitalism) is likely to be tedious for humans, too. Ask yourself - what exactly is the domain of interesting decisions that the player will be making in this game such as to make it fun?

Exactly. Preferably I would left out distribution completelly. But I need some sort of distribution because players would want several markets (cities/regions/continenst/whatever) and factories in several places :( So, I need to handle somehow the distribution part of the game (preferably in a minimalistic way).

The point that I intended to make is that you should be able to automate all of the tedious stuff (such as the logistics of distribution) without adversely affecting the player's ability to do interesting things because anything that can be automated should not be done by the player in any event. So, the answer that I propose is that you simply automate the distribution (or, at least, allow the player to choose to automate it).