I'm playing a game right now, and  I've already twice made a calculation to see if it makes sense to attach new territories. The last calculation looks like this:

It was calculated what the production costs are for an industrial district. You can clearly see that the more territories you connect, the longer the production of a district takes. It is clear that the more districts the city has to maintain in the end, the more expensive the new districts are. 

Connecting districts to a city is only a must if you:

- Collect stars

- Want to have a territory where no unwanted battles take place

- No space for new districts

Otherwise in my opinion it is hardly worthwhile to attach new territories! 

I only attach the minimum it takes for stars until I get the ability to combine cities, it only makes it more expensive to make districts due to inflation.

I'll generally Attach territories if and when I have a new front in my battles that I want to drop a horde of troops from.  
Build a troop training area with gold, set up a quick batch of troops for the next turn and continue to push forward. Once the battles over I can remove if needed, but it tends to work decently.

There's an issue of outposts not being a claimed territory but rather a disputed one, so unless you plan to go to war with an a-hole neighbour that refuses non-aggression pact, it's better to at least consolidate your borders and save yourself a headache.

And there's also local pollution to keep at the back of your head, if you want to industrialize before inventing green energy, you will need to spread your FIMSI generation around and maneuver your districts around natural reserve spots. Even on low the penalties can end up crippling your city, there's a lot of people that moan pollution is too harsh while evangelizing that makers quarter is be-all and end-all, because industry is the king, and I have a sneaking suspicion that's the thing they don't take into account.

@RebelsDawn @DNLH  Thank you for your input. These are good reasons to connect more territories but as I see it, these are more cases for late game, where we are normally strong in snow balling. In my opinion in early and mid game you have to resist to attach new territories. 

This behavior we can probably clarify through history. Strong and big cities are created later. In the beginning you had rather smaller cities and many autonomous territories, which were not significant. So I clarify this now to understand it better and improve the game immersion.

This method of scaling is not intuitive for new players, like the district one. Punishment for making more district or building lux resources xD give me a break man

Ciabat wrote:

This method of scaling is not intuitive for new players, like the district one. Punishment for making more district or building lux resources xD give me a break man

I agree with your statement

I still think that devs should try making a hard cap on how many districts one can build, that would expand by number of attached territories and with tech, instead of trying to put in a weird soft cap by district cost scaling. That would be a solid incentive to attach territories rather than current system that discourages it and I think it would also make people think more about specializing cities. Only issue is that it would further weaken EQs that are already mainly an aesthetic choice to build, but maybe they could be outside of that cap, if they have cap on their own anyway.

Bigger issue is that they'd need to teach AI how to properly build their cities without ability to just spew out more garrisons/commons quarters.

Diegoyya wrote:

Early game I attach to have more population to generate more influence per turn. Also more population to make more army, and more production for soldiers as well. You also attach to expand your borders and limit others not to traspass if they are not expansionist. There are many strategic reasons to attach. Not to mention that certain uniqute districts like mayas, or celts give benefits for attached territories.

You don't need to attach permanently for getting population. Territory can produce itself the population and you can attach (shift population from territory to city) and in the same turn detach. I agree based on  strategic reason you proposed but economically is not worth it.

Variant 4 is making 99 more production than variant 1, and I don't think your calculation accounts for the benefits of that for building infrastructure and units. To make it a more fair comparison, I would add on the cost of building enough maker's quarters to get to 318 production/turn in variants 1-3. The idea being that in variant 4, you might only want to build 1 maker's quarter, whereas in variant 1, you may want to build several.

mrpmg wrote:

Variant 4 is making 99 more production than variant 1, and I don't think your calculation accounts for the benefits of that for building infrastructure and units. To make it a more fair comparison, I would add on the cost of building enough maker's quarters to get to 318 production/turn in variants 1-3. The idea being that in variant 4, you might only want to build 1 maker's quarter, whereas in variant 1, you may want to build several.

The data in the table comes directly from the game. I have not changed them. This means that the increased production has been taken into account.

DNLH wrote:

I still think that devs should try making a hard cap on how many districts one can build, that would expand by number of attached territories and with tech, instead of trying to put in a weird soft cap by district cost scaling. That would be a solid incentive to attach territories rather than current system that discourages it and I think it would also make people think more about specializing cities. Only issue is that it would further weaken EQs that are already mainly an aesthetic choice to build, but maybe they could be outside of that cap, if they have cap on their own anyway.

Bigger issue is that they'd need to teach AI how to properly build their cities without ability to just spew out more garrisons/commons quarters.

I wish they would give us the ability to develop an individual district more than once with some drawbacks in later eras.

This would help curb urban sprawl to the point where a city spans a whole territory, eventually ruining much of the beauty in the game. 

Wald