I usually play VP civ V on deity, marathon, huge map, abundant resources, continents with atlantic/pacific, raging barbs. I think these settings break the game to a degree, but I am not sure if this is intended or not.
Some observations:
1. There are some general resources in the game: food/population, (free) production, culture, gold, science, faith, happiness. In general, four of these (culture, gold, science, faith, production) scale at most linearly in the number of cities you have. Clearly, the actual growth is sublinear because it takes time to get buildings up and build settlers, but in the ideal situation where most buildings are produced "instantly", these resources scale linearly. The remaining two resources (food/pop and happiness) do not scale linearly. Food/pop is roughly constant in the number of cities you have up to a point, then which it decreases and then saturates. Happiness serves as a cap on the number of cities you have, since being very unhappy is crushing for production
1a. Out of these resources, production and science are the most important to victory. Production allows for better war and quicker adoption of buildings that leads to more science overall, and science leads to better war and better buildings. Faith, culture, and gold are important insofar as they grant more production and science. Faith and gold can be seen as "immediate/banked production", while culture is less immediate but confers greater (and more unique) production and science bonuses. Food and happiness (and GPP/wonders to some degree) can be seen as mitigating caps on the number of cities - wide empires necessarily have fewer of these per city.
1b. Vox populi has less population scaling than the base game. Importantly, most science scaling is done per-city, rather than per-head. This means that the food/pop decrease from spamming cities is is irrelevant compared to the additional yields that each city gets.
2. Two things in VP break these linear scaling laws. Factories and Corporations (with the order tenet) scale quadratically. This means that by late modern era, cities in a large (~100 city empire) will run out of things to build because their production is so high. If the culture corporation is adopted, then a large (~100 city) empire can finish a whole tree in the late game in the time it would otherwise need for a single policy. Notably, 2/3 of science, culture, production can be scaled quadratically.
3. After flight and landships, and especially after carriers, it becomes impossible to hold land without also fighting with similar weapons. Moreover, such wars, because both sides are fighting in a similar manner, become very attritional: the side with more units is likely to win, since good defensive terrain (forts/citadels/railroads) cannot be held, and there is no "rear area" to retreat and heal, as there have been in previous eras (no one unit can stand up to 10 bomber strikes, preceded by a massive fighter sweep). Therefore a surfeit of production inevitably leads to military success, and lacking production entails being conquered. There is no cost-efficient defensive war.
3a. It is also very difficult to fight a cheap naval war for the following reasons. There is no rough terrain or cities in the sea, so ships can rarely survive attack from more than three opposing units. The sea is also very large, which means that flanking maneuvers are usually possible to allow such force concentration. Boarding action also means that ships cannot flee to "rear areas" to heal. Therefore clashes at sea tend to be more lopsided than on land, and once command of the sea is gained and a navy is set up outside all enemy ports, it is near impossible to regain such control without an inland lake from which to reconstitute a navy. Therefore to win such a naval war requires extensive military buildup, and losing such a naval war can be devastating. In VP, it is very difficult for land forces to defend cities against a navy, so navigation and industrialization essentially force naval/technological arms races to avoid having to abandon coastal cities. (The AI likes to frigate and cruiser rush with ~20 units).
4. Public works means that happiness is no longer a hard cap on number of cities.
(1) Implies that if buildings could be purchased instantly, more cities would always be better. However, up until the industrial era, this is not the case, as a newly founded city may take 100 turns to catch up in the build order, and delay science and culture during these turns. Therefore people usually limit the amount of cities, and it is beneficial to focus on city quality. However due to (2), factories allow for cities founded at any point in the game very rapidly catch up. It is not uncommon for a 100 city empire to build a factory in 8 turns, then build one building per turn for the next 20 turns and then become nearly as strong as the capital. The consequence of (3) implies that in the late game, whoever has production is likely to conquer the world. In previous eras, this was mitigated by greater distances (deep water technology), and fortifications, that are overcome with stronger naval and air units. In the endgame, deity AI are likely to have one civ steamroll their entire continent and then attack the player. Therefore if one wants to survive past the modern era, it is recommended that they spam as many cities as possible (as coal will allow).
Therefore, it seems to me that the only strategy that can guarantee survival is to spam cities and go wide (settle cities three tiles apart). If not, then one will lose the late-game war to another civ with more cities. To spam cities effectively and go wide requires lots of land. Since it is comparatively very easy to conquer civs early, then that implies that we should adopt a strategy of war starting from the very beginning. If we do not kill off enough neighbors to carve out a large (~100 city) empire, in the late game a steamrolling AI from the other continent will wipe us out. Since our end goal is a wide empire with an initial goal of war, the necessary strategy writes itself.
In terms of policies, I prefer authority/progress - fealty - industry. If I have more than three other civs on the same continent, I take authority, otherwise I will take progress. The benefits of authority (culture from kills) combined with a big army allows very quick early culture, to balance out the later culture penalties from having a wide empire. Fealty is a no-brainer for a wide empire. I prefer industry over rationalism since production is very important for cities to catch up in the build order, wide empires often struggle with gold. For religion, I prefer either the pantheon that gives faith from kills, or the pantheon that gives science per city, followed by two two faith buildings (I like churches and mandirs, though teocallis are good if you are warring) and zealotry (for more coal). Reformation beliefs of divine teaching/jesuit education (science purchasing) is extremely crucial (In my last game, I was able to build 100 research labs immediately upon getting plastics), otherwise you may fall behind midgame in technology. If you can't get that, culture purchasing is also good. To the Glory of God is a reasonable alternative if you aren't fast enough to get those beliefs.
Generally, this strategy is limited by the amount of coal on the map, so in the late game, I recommend selling your train stations and seaports if there is not enough coal to go around. Similarly, save horses for agribusinesses in only essential GPP producing cities. The endgame great airsea wars usually begin around turn 700-800 on marathon setting. The AI is very bad at building dense cities, and usually your cities end up being a factor of (1.5^2) denser than their cities. This means that you can usually get around double their amount of cities with similar land, which eventually translates to 4x production.
To give an example, my previous/current game as zulu. I killed my closest neighbor with an archer rush. This was relatively slow, since they were able to get walls in time. I then got compbows and killed my next three closest neighbors. I then took crossbows and killed another civ (they had 13 cities) and crippled another civ across the ocean. I took the capital of that civ with impis. One of the civs on the other continent steamrolled the rest of the civs on that continent and by industrialization, there were only us two civs, one holding each continent, and they were 6 techs ahead. I was unable to win an early naval war against them and had two cities that were constantly taken and retaken by ironclads. We are now in the atomic era and due to instant lab purchasing and culture corporation (free techs in order), as well as scholars in residence + science focus, I was able to catch up in techs. I reached carriers at about the same time, and have a fleet of about eighty planes. Due to brandenburg + teocallis and zulu's xp bonus, I am able to get 4.5 promotions immediately. I was lucky this game, since I was able to get an early religion, combined with terracotta and statue of zeus, made wars very quick and profitable. I was unable to get globe theater, but uffizi allows for one extra specialist per city which is helpful. Most AI cities are settled in suboptimal locations, so I puppeted them for a few hundred turns and then razed them.
Essentially, it seems that wide games are forced on this high difficulty to survive the late game domination wars. The only other strategy would be to win before factories come online, but this is very difficult on huge map and deity. Deity AI generally has very good culture. Diplo victories are also very difficult because AI will kill your city states. Nukes can potentially allow for smaller civs to defeat bigger civs, but a large wide civ can tank the destruction of 10 cities without losing significantly. In addition, nukes cannot be produced fast enough to overcome conventional arms. I don't see how any civ can defend against the endgame production of a large empire without a similar resource base.
I don't know if this is an intentional railroading towards a late game preference for large empires, or rather the consequence of wanting to make factories and corporations impactful. In any case, I think this is overtuned, since it means that most wonders and great people don't matter that much after factories arrive on the scene. Moreover, it railroads the endgame into a fixed state of total war. Perhaps it historical, but it is perhaps confining.
Some observations:
1. There are some general resources in the game: food/population, (free) production, culture, gold, science, faith, happiness. In general, four of these (culture, gold, science, faith, production) scale at most linearly in the number of cities you have. Clearly, the actual growth is sublinear because it takes time to get buildings up and build settlers, but in the ideal situation where most buildings are produced "instantly", these resources scale linearly. The remaining two resources (food/pop and happiness) do not scale linearly. Food/pop is roughly constant in the number of cities you have up to a point, then which it decreases and then saturates. Happiness serves as a cap on the number of cities you have, since being very unhappy is crushing for production
1a. Out of these resources, production and science are the most important to victory. Production allows for better war and quicker adoption of buildings that leads to more science overall, and science leads to better war and better buildings. Faith, culture, and gold are important insofar as they grant more production and science. Faith and gold can be seen as "immediate/banked production", while culture is less immediate but confers greater (and more unique) production and science bonuses. Food and happiness (and GPP/wonders to some degree) can be seen as mitigating caps on the number of cities - wide empires necessarily have fewer of these per city.
1b. Vox populi has less population scaling than the base game. Importantly, most science scaling is done per-city, rather than per-head. This means that the food/pop decrease from spamming cities is is irrelevant compared to the additional yields that each city gets.
2. Two things in VP break these linear scaling laws. Factories and Corporations (with the order tenet) scale quadratically. This means that by late modern era, cities in a large (~100 city empire) will run out of things to build because their production is so high. If the culture corporation is adopted, then a large (~100 city) empire can finish a whole tree in the late game in the time it would otherwise need for a single policy. Notably, 2/3 of science, culture, production can be scaled quadratically.
3. After flight and landships, and especially after carriers, it becomes impossible to hold land without also fighting with similar weapons. Moreover, such wars, because both sides are fighting in a similar manner, become very attritional: the side with more units is likely to win, since good defensive terrain (forts/citadels/railroads) cannot be held, and there is no "rear area" to retreat and heal, as there have been in previous eras (no one unit can stand up to 10 bomber strikes, preceded by a massive fighter sweep). Therefore a surfeit of production inevitably leads to military success, and lacking production entails being conquered. There is no cost-efficient defensive war.
3a. It is also very difficult to fight a cheap naval war for the following reasons. There is no rough terrain or cities in the sea, so ships can rarely survive attack from more than three opposing units. The sea is also very large, which means that flanking maneuvers are usually possible to allow such force concentration. Boarding action also means that ships cannot flee to "rear areas" to heal. Therefore clashes at sea tend to be more lopsided than on land, and once command of the sea is gained and a navy is set up outside all enemy ports, it is near impossible to regain such control without an inland lake from which to reconstitute a navy. Therefore to win such a naval war requires extensive military buildup, and losing such a naval war can be devastating. In VP, it is very difficult for land forces to defend cities against a navy, so navigation and industrialization essentially force naval/technological arms races to avoid having to abandon coastal cities. (The AI likes to frigate and cruiser rush with ~20 units).
4. Public works means that happiness is no longer a hard cap on number of cities.
(1) Implies that if buildings could be purchased instantly, more cities would always be better. However, up until the industrial era, this is not the case, as a newly founded city may take 100 turns to catch up in the build order, and delay science and culture during these turns. Therefore people usually limit the amount of cities, and it is beneficial to focus on city quality. However due to (2), factories allow for cities founded at any point in the game very rapidly catch up. It is not uncommon for a 100 city empire to build a factory in 8 turns, then build one building per turn for the next 20 turns and then become nearly as strong as the capital. The consequence of (3) implies that in the late game, whoever has production is likely to conquer the world. In previous eras, this was mitigated by greater distances (deep water technology), and fortifications, that are overcome with stronger naval and air units. In the endgame, deity AI are likely to have one civ steamroll their entire continent and then attack the player. Therefore if one wants to survive past the modern era, it is recommended that they spam as many cities as possible (as coal will allow).
Therefore, it seems to me that the only strategy that can guarantee survival is to spam cities and go wide (settle cities three tiles apart). If not, then one will lose the late-game war to another civ with more cities. To spam cities effectively and go wide requires lots of land. Since it is comparatively very easy to conquer civs early, then that implies that we should adopt a strategy of war starting from the very beginning. If we do not kill off enough neighbors to carve out a large (~100 city) empire, in the late game a steamrolling AI from the other continent will wipe us out. Since our end goal is a wide empire with an initial goal of war, the necessary strategy writes itself.
In terms of policies, I prefer authority/progress - fealty - industry. If I have more than three other civs on the same continent, I take authority, otherwise I will take progress. The benefits of authority (culture from kills) combined with a big army allows very quick early culture, to balance out the later culture penalties from having a wide empire. Fealty is a no-brainer for a wide empire. I prefer industry over rationalism since production is very important for cities to catch up in the build order, wide empires often struggle with gold. For religion, I prefer either the pantheon that gives faith from kills, or the pantheon that gives science per city, followed by two two faith buildings (I like churches and mandirs, though teocallis are good if you are warring) and zealotry (for more coal). Reformation beliefs of divine teaching/jesuit education (science purchasing) is extremely crucial (In my last game, I was able to build 100 research labs immediately upon getting plastics), otherwise you may fall behind midgame in technology. If you can't get that, culture purchasing is also good. To the Glory of God is a reasonable alternative if you aren't fast enough to get those beliefs.
Generally, this strategy is limited by the amount of coal on the map, so in the late game, I recommend selling your train stations and seaports if there is not enough coal to go around. Similarly, save horses for agribusinesses in only essential GPP producing cities. The endgame great airsea wars usually begin around turn 700-800 on marathon setting. The AI is very bad at building dense cities, and usually your cities end up being a factor of (1.5^2) denser than their cities. This means that you can usually get around double their amount of cities with similar land, which eventually translates to 4x production.
To give an example, my previous/current game as zulu. I killed my closest neighbor with an archer rush. This was relatively slow, since they were able to get walls in time. I then got compbows and killed my next three closest neighbors. I then took crossbows and killed another civ (they had 13 cities) and crippled another civ across the ocean. I took the capital of that civ with impis. One of the civs on the other continent steamrolled the rest of the civs on that continent and by industrialization, there were only us two civs, one holding each continent, and they were 6 techs ahead. I was unable to win an early naval war against them and had two cities that were constantly taken and retaken by ironclads. We are now in the atomic era and due to instant lab purchasing and culture corporation (free techs in order), as well as scholars in residence + science focus, I was able to catch up in techs. I reached carriers at about the same time, and have a fleet of about eighty planes. Due to brandenburg + teocallis and zulu's xp bonus, I am able to get 4.5 promotions immediately. I was lucky this game, since I was able to get an early religion, combined with terracotta and statue of zeus, made wars very quick and profitable. I was unable to get globe theater, but uffizi allows for one extra specialist per city which is helpful. Most AI cities are settled in suboptimal locations, so I puppeted them for a few hundred turns and then razed them.
Essentially, it seems that wide games are forced on this high difficulty to survive the late game domination wars. The only other strategy would be to win before factories come online, but this is very difficult on huge map and deity. Deity AI generally has very good culture. Diplo victories are also very difficult because AI will kill your city states. Nukes can potentially allow for smaller civs to defeat bigger civs, but a large wide civ can tank the destruction of 10 cities without losing significantly. In addition, nukes cannot be produced fast enough to overcome conventional arms. I don't see how any civ can defend against the endgame production of a large empire without a similar resource base.
I don't know if this is an intentional railroading towards a late game preference for large empires, or rather the consequence of wanting to make factories and corporations impactful. In any case, I think this is overtuned, since it means that most wonders and great people don't matter that much after factories arrive on the scene. Moreover, it railroads the endgame into a fixed state of total war. Perhaps it historical, but it is perhaps confining.
