[Discussion] Overextension

[gigau]

In the v4.0.3a discussion thread, we have started talking about the overextension of nations. I guess we could discuss certain aspects here.

The vanilla modifier looks at the core percentage, and i like that.

It also looks at the number of cities, depending on the centralization_decentralization slider.... and, well, i don't think i agree. I believe Breat Britain, France and Russia were centralised (or am i mistaken) and held big empires... better that other much more decentralised states. So i wonder if this is a correct way to go about it.

Fuzzbug pointed out that num_of_cities was tough to use given the difference in density of provinces. I can only agree with this statement.

A way of combining the facts i've stated in those two last points could be of rather linking the num_of_cities to the government... possibly. At least partially. What do you think ?

This would lead to the question as to what number of provinces could be allowed for a given government ? An example i often have in my hands off test games is the imperial city of Geneva who likes to expand, sometimes up to 10-15 provinces. Should it be accepted ? Should it get a nasty event ? Or should it be invited to change government and/or tag ?

[Mr. Capiatlist]

Originally Posted by gigau

In the v4.0.3a discussion thread, we have started talking about the overextension of nations. I guess we could discuss certain aspects here.

The vanilla modifier looks at the core percentage, and i like that.

It also looks at the number of cities, depending on the centralization_decentralization slider.... and, well, i don't think i agree. I believe Breat Britain, France and Russia were centralised (or am i mistaken) and held big empires... better that other much more decentralised states. So i wonder if this is a correct way to go about it.

Fuzzbug pointed out that num_of_cities was tough to use given the difference in density of provinces. I can only agree with this statement.

A way of combining the facts i've stated in those two last points could be of rather linking the num_of_cities to the government... possibly. At least partially. What do you think ?

This would lead to the question as to what number of provinces could be allowed for a given government ? An example i often have in my hands off test games is the imperial city of Geneva who likes to expand, sometimes up to 10-15 provinces. Should it be accepted ? Should it get a nasty event ? Or should it be invited to change government and/or tag ?

Obviously an Imperial City with more than one city should be forced to take a new title/government.

Over extension should be a function of the following:
Stability, Prestige, Centralization, NIs, Core vs Non Core and Government Tech.

Great Britain had a large Empire, and though the Isles themselves were centralized, the colonies weren't (until the modern age with a little thing called the telephone). Colonies had forms of limited self government, the 13 colonies especially, one cause of the Revolution was the revoking of Massachusetts' right to self government.

But here is the run down:
Stability is a major contributer, if you aren't stable, you are likely to loose your empire.

Prestige is a middle contributer; this is using prestige as more of a check of success then literal prestige. If a country is doing poorly it should begin to have trouble controlling colonies, non-contiguous and non-core regions.

A centralized government would have more issues with colonies, this can be undone however with:

NIs like Viceroys, Bill of Rights, Cabinet should increase the size of a country before it becomes "Over extended".

Too many cores is more of a measure of expansion rate then of "this land is mine" ideals. Obviously core or not, if a country conquers a province they believe it to be (at least to some degree) theirs. Instead, if a country expands too quickly, or in an unexpected way, it might have trouble exerting authority on the newly conquered regions.

Government tech is mostly just a ticker, so it can be replaced by year checks, but basically as a country evolves how its government can work, it becomes better at ruling large regions.

[Aenas89]

Originally Posted by gigau

In the v4.0.3a discussion thread, we have started talking about the overextension of nations. I guess we could discuss certain aspects here.

The vanilla modifier looks at the core percentage, and i like that.

It also looks at the number of cities, depending on the centralization_decentralization slider.... and, well, i don't think i agree. I believe Breat Britain, France and Russia were centralised (or am i mistaken) and held big empires... better that other much more decentralised states. So i wonder if this is a correct way to go about it.

Fuzzbug pointed out that num_of_cities was tough to use given the difference in density of provinces. I can only agree with this statement.

A way of combining the facts i've stated in those two last points could be of rather linking the num_of_cities to the government... possibly. At least partially. What do you think ?

This would lead to the question as to what number of provinces could be allowed for a given government ? An example i often have in my hands off test games is the imperial city of Geneva who likes to expand, sometimes up to 10-15 provinces. Should it be accepted ? Should it get a nasty event ? Or should it be invited to change government and/or tag ?

totally agree with the first point. It´s good that paradox tried to slow down extremly fast expanding, because it´s unrealistic. But to make the cent. dez. slider one main factor is not reasoned. A dezentralised gov could be one with efficient local politic, but it could also stand for a a corrupt government where local aristocrats or warlords hold all the power. It gives no coercive information about how efficient the goverment is, and that should be the main orientation for an event like overextention.

Actually exact this problem is implemented in MMP with Adminstration efficiency and Administration Ability. From my opinion the sanction system is a bit too harsh, because expanding (especially at the beginning) is very restricted but thats another question, however you should take a look at it.

[Sandam]

ok... so i looked into MMP2 "AE Recalculation" its not that i want the same in MEIOU. MM is way to restricted but what i wanted to find is mechanics "how" it is calculated. i am a noob at EU3 scripting, but from that i learned form C++ i can see this...

any_country = { set_variable = { which = province_count value = 0 } }

this sets province_count value to 0 for every country.

any_country = {
		limit = { exists = yes }
		any_owned = {	# base count
			owner = { change_variable = { which = province_count value = 1 } }
		}
}

as i understand this goes through all countries and all their provinces changing the countries province_count value by +1 for every province that country have. or to tell that short - counts provinces. so i think if we add line with something like limit = { population = 10000 } inside any_owned it would count only provinces with 10000+ people.. so i guess this is the easy way to count provinces and later use province_count as a trigger for triggered modifier... of course MM is not that simple... its just a short part taken from that long file counting administrative efficiency.. but still i dont know how this event fires and how often... MM is just to complicated...

[gigau]

The way to prevent overextension/blobbing will be tough to implement, as there will always be a part that displeases someone.


That being said, given what i read, i see three aspects to the feature that i wish to seperate :

• the triggers. those i think will be the toughest to put in place, especially regarding the size of the realm. The remaining (core percentage, stability, prestige, infamy, government level or type,...) are easy enough. for the size, i see two ways to go about it : we go for num_of_cities or for population.
The first is the easy way. The downside is that num_of_cities counts minor colonies as well, and doesn't desciminate highly populated provinces.
However, the population way will require an event in the way the fief event works, changing a variable with changing values depending on the city size. We could replace multiple fief events by a bigger event calculating the nation variable once in a while. Then again, how often should this be calculated. At one point, i thought about doing this at the change of monarch. However, in case of a long reign, i fear it'd be often enough. Therefore, when ? how often ?

• the modifiers. We can have, as Fuzzbug suggested (more or less), a series of scaled modifers with subtle but increasing penalties like just an increase in the stability cost modifiers, and in the really overexpanding cases a slight cost in tech costs (there again subtle, but present).

• the true penalty. An overexpanding nation would get events, like revolts in non-core provinces, revolt in the colonies, a war of religions, and in latter stages the secession of parts of the realm of different culture groups.

[Broaz]

One other sort of penality that could be nice would be to slow the BB decrease rate and/or the BB limit for the over extended countries. Then the bigger you are, the more likely you are to face many ennemies when you wage war.

[grzegorz444]

AI build to large armies. During my Lithuania campaign after 2 yaers game i saw 2 extremely large moskowian armies: first 22 cav, 10 infantry, second: 15 cav, 8 infantry
It's unhistoric
I don't know what expect after 100 years game


The states are too rich, bulding large army is wery easy. It,s like vanilla

[MacroEconomics]

Is overextension really a problem currently? It seems the map is so large (num of provinces so large) that certain countries which did expand a lot during the era are going to have an incredibly difficult time accomplishing that. I'm thinking of the Ottomans and Russians in particular. I think the central issue is that whatever add-on overextension penalties a mod adds, there is no way to add-on a corresponding ability for the AI to think properly about those new penalties. The AI just continues behaving the way it did in EU3 w/o modification and runs head long into the new penalties. Which I guess is an argument against making the penalties too severe - the human players knows how to avoid the penalties, the AI unfortunately does not...

[Sandam]

well picking up population vs num_of_cities is basically question of performance... i hope in next version fiefs will be removed... i just hope it was slowing game down, not something else. so if the game speed wont improve, more resource demanding population method is out of question... i just think that keeping game simple is better than making huge events which give a strain to PC. but if game runs fine then more accuracy would not hurt.