Analysis of 1.2.5 on Building Cost and Overbuilding

[phoenixician]

Hi everyone,

Here is my analysis of patch 1.2.5's effect on production and I would like to discuss it with you all. FYI: I am not a math major, so I may make a mistake somewhere.

Patch 1.2.5 introduced the following change to building cost:

• Building Cost Progression:All Building costs now scale with the number of other Buildings in that Settlement, and Cities in your Empire.
  • The cost to produce and purchase a Building increases by 5% for each non-Warehouse Building in that Settlement.
  • The cost to produce and purchase a Building increases by 10% for each City other than your Capital in your Empire.

This change undoubtedly makes building significant longer, especially if you expand your cities' numbers or infrastructures. Currently, there are 2 currencies to afford buildings: Production & Gold. With the change in place, in terms of number, the value of each currency relative to the base building cost goes down with each new building or city.

For example, if my very first city in antiquity (capital) has 10 production, and I want to build the first building, library (90 production), thus it would take 9 turns. The value of 1 production is 1:1 in this case. But...
...If I already have 9 other cities, in my newest 10th city with 10 production, to build a library (which would be 180 production because cost scaled to 200% at this point), it would take 180 turns. Value of production is 1/2 (or 50%). While your 10 productions remain the same number, its value has gone down to 5 relative to the building's base cost.
...If I already have 19 other cities, in my newest 20th city with 10 production, to build a library (which would be 270 production because cost scaled to 300% at this point), it would take 270 turns. Value of production is 1/3 (or 33%).

If we were to demonstrate the value of each currency corresponding with increasing cost, we would get a reciprocal graph:

As you can see, the value of production actually starts to plummet the moment you lay down your very first specialty building or non-capital city. However, this rate does slow down as the building cost gets higher.

The cost of buildings (in terms of production and gold) can be mitigated by efficiency modifiers.

1. What is efficiency modifier?
An efficiency modifier increases your currency by displayed %. For example, if you have Science node 2 (+15% production toward building), the game recognizes that every 1 production = 1.15 when you build. Thus, we can find how much production and time saved with the following formula:

Cost saved = Cost - (Cost/[Effiency modifier])

Although most are additive with each other, there are some exceptions, which are multiplicative, because they modify the production value at the settlement level.

2a. Production efficiency modifiers - Universal
This type of modifier applies when you're constructing new building and/or overbuilding. Most notables are:

• Science node 2: +15% production toward building
• Resource - Coffee (Modern only): +5% production toward building and wonders
• Science City State Suzerain bonus: +5% Production towards constructing Buildings for each Scientific City-State you are suzerain of.
• Great Britain's (GB) ability - Workshop of the World: +25% Production towards constructing Buildings
• Russia's unique tradition - Emancipation Reform (ER)*: +15% Production and -15% growth in Cities.
• Mughal India's (MI) ability - Paradise of Nations*: -25% Production in all settlements (Malus)

*These are special cases. They modify the production value at the settlement level first, then apply other modifiers multiplicative. Thus, the formulas for Russia with Emancipation Reform slotted and Mughal are respectively as follows:
Cost saved (Russia) = Cost - (Cost/[1.15*Effiency modifiers])
Cost saved (Mughal) = Cost - (Cost/[0.75*Effiency modifiers])

Honorable mentions:

• Songhai's unique tradition - Mud Brick: +50% Production towards constructing Gold Buildings
• Hawaii's unique tradition - Kapa: +50% Production towards constructing Culture Buildings
• Benjamin Franklin's ability - The First American: +50% Production towards constructing Production Buildings.

We can then summarize the cost/turn saved with these modifiers in the following table:

Table 1: Building cost/turn saved by attributes above are presented as negative values when compared to a generic Civ that does NOT have any of them. In Mughal's case, because of their penalty, we can think of them having to invest more production (as presented in positive value).

2b. Production efficiency modifiers - Overbuilding
This type of modifier applies when you're overbuilding a specialty building from previous era. Before we go further, I want to explain the current overbuilding feature is implemented as following sequence:

1. Pick building eligible for overbuilding (older era, early tier)
2. Building is then demolished completely (thus, "refund" the additive 5% building cost penalty and also yields & maintenance cost associated with old building. This complicates the formula a bit, but it still remains a linear equation)
3. New building in progress in place of the demolished --the new building MUST not be changed at any point from here till next turn (that is, you get only 1 chance to make the choice and must invest at least 1 production into the new building by next turn), or the entire process will NOT count for the overbuilding purpose.

These modifiers are additive with the universal modifiers since they do not alter the production level on the settlement screen. Currently, there are 3 specific overbuilding modifiers:

• Exploration-age policy - Commune: +20% Production towards overbuilding.
• Modern-age policy - Civil Engineering (CE): +30% Production towards overbuilding.
• Majapahit's unique civic - Wayang: +25% Production towards overbuilding.

Thus, the cost/turn saved by having these attributes and policies slotted can be summarized in the following table:

Table 2: Building cost/turn saved by attributes above are presented as negative values when compared to a generic Civ that does NOT have any of them. In Mughal's case, because of their penalty, we can think of them having to invest more production (as presented in positive value). In Mughal's case, because of their penalty, production is only saved when they have both CE slotted and Node 2.

2c. Overbuilding
Given the above overbuilding sequence, you can utilize overbuilding to remove unwanted old buildings. You can even remove the urban tile itself (IF it contains only Antiquity buildings) in Explorage age by the doing the following sequence:

1) Pick an unwanted urban tile that contains antiquity buildings.

2) Choose any building on the production queue then overbuild. Click Next Turn. Note: Doing this will lock the new building to that urban tile in next turn, so choose carefully.

3) Now that there are production invested into new building, you can safely overbuild other building with any building. Then cancel it from production queue. Note: Doing this will NOT lock other building.

4) Since all buildings that are "In Progress" do not count as fully buildings, the urban tile is then "removed" in next era.

Note: As you can see, even if that urban tile was leading to a farther urban tile, it can still be removed by doing this. That tile to the south-east of the lone quarter is the potkop improvement.

4a) Nuance: That urban tile must not contain any wall. The wall from older age will prevent urban tile from being removed since all walls count as buildings.

*The medieval wall prevented urban tile from being removed.

*No wall makes it possible to remove the urban tile entirely.

2d. Okay great. Now that we have these numbers, is it more efficient to lay out urban tiles and overbuild than to build everything from the ground up?
-In general-, yes, it is more efficient to overbuild (with the overbuilding boosting policies and attributes) than it is to hard-build everything from the ground up, even if we only consider the production value.

To demonstrate this, I made a table to compare building cost before and after overbuilding (again, with appropriate policies slotted).
Given that all tier 1 and 2 respectively have their uniform costs. For the sake of simplicity, we treat % numbers as % of each building's base cost (and for the following example). Most left column (brown) is the unaltered cost of building, then costs of building modified by other attributes in other columns.

Example: If I am Simon Bolivar, playing as Nepal, and I have Node 2 unlocked. I have 6 non-capital cities. In my newest 6th city currently, I only have 1 unique ageless building (therefore, the current scaled building cost is at 165%) in Modern Age:

• If I were to build 8 buildings from the ground up, the cost of new buildings will increase by 5% (of base cost) for every existing building + # of cities, thus I would have to invest 1270% worth of production cost (143% + 148% + 152% + 157% + 161% + 165% + 170% + 174%).
• If I were to plan for overbuilding in Modern by laying out urban tiles with exactly 8 overbuildables, I would have to invest 1076% worth of production cost (134%*8).

If you repeat this math at any other different building scaled cost with any # of overbuildables, the overbuilding cost will be cheaper than building everything from ground up.

2e. One observation: Buildings from previous eras in towns increase build cost, buildings from the current era do not. (Possible bug)
Example: I currently have non-capital 4 cities, my building cost would scale to 140% from having 4 cities alone. In this town, there is only 1 overbuildable (library).

...But if I overbuild that library to school house, the cost actually decreases.

3. Gold Economy in 1.2.5
In relation to gold, 1.2.5 also made the following changes to Gold & Silver:

• Gold: +1 Gold and Happiness in all Settlements. (Previously: +20% Gold towards purchasing Buildings.)
• Silver: +1 Gold and Food in all Settlements. (Previously: +20% Gold towards purchasing and upgrading Units.)

Unsurprisingly, this is a big nerf to the gold economy from prior. These were all efficiency modifier that can be obtained per resource node on the maps, which added up quickly by Exploration Era. Thankfully, there are still a few efficiency modifiers around (such as Economic attribute node 5). Before we go further though, I want to talk about Mughal India.

3a. Mughal India as a special case
Mughal's ability is as follows: +75% Gold from all sources. -25% to all other yields. Simple yet true to every single word. Mughal's ability applies global multiplier to all yields (1.75x for gold, 0.75x for everything else); the net results for food & production only display at the settlement screen. This has many implications. For examples:

• You may think that every 1 production in towns would equate to 1.75 gold for Mughals. This isn't true. Every 1 production of yield displayed on a Mughal town's tile would be multiplied by 0.75 first then 1.75 (=1.3 Gold).
• Mughal's global multiplier is literally multiplicative with Expansionist attribute node 8 (+15% Yields in Towns with a Specialization, or +30% in Distant Lands, or just 30% in ALL towns in Modern Age).

3b. Gold efficiency modifiers
Though similar to production modifiers, gold modifiers are additive with each other, but they are far fewer in availability.

• Economic node 5: +15% gold toward purchasing anything.
• Great Britain's (GB) ability - Workshop of the World: +25% gold towards constructing Buildings
• Mughal India's (MI) ability - Paradise of Nations*: +75% Gold from all sources
• Mughal India's unique tradition - Mayūrāsana: +25% gold towards purchasing anything.

Thus, just like with production, we can calculate gold saved by having any of these efficiency modifiers:

Table 3: Gold cost saved by attributes above are presented as negative values when compared to a generic Civ that does NOT have any of them.

 

[phoenixician]

Here are my thoughts on this patch:

1. Tall vs. Wide

Since the release of civ7, discussions about which civ is better suited for tall vs wide have been mostly absent, and understandably so since there hasn’t been any factors that would nudge the player toward one over another. In fact, there have been barely any discussions about that topic since civ5 days. I think I only came across one during Maya release in civ6, but I am not sure if Maya civ6 would be counted as tall or wide in civ6 context.

In fact, let me ask you all this: What is “tall” vs “wide” in Civ7? Do we go by Civ5’s standards and define “tall” as having =<4 cities or >4 cities as “wide”? Or do we go by 3 cities since that’s the threshold where you get extra bonus from science & culture specialist nodes? Can there be “tall-adjacent” that is somewhat in the middle between wide & tall? I think ultimately at this point, we can at least agree on vague definitions like “tall = few & big cities” & “wide = many but smaller cities.”

With that being said, I do not think these nerfs to building cost completely killed the “wide” play. Referring back to my graph of production value over increased cost, the steeper decline actually occurs at lower scaled cost. For example, the slope of the line at cost 150-200% is steeper than 200-250%. I am not saying any drop of value is negligible, either. Since production and gold remain the 2 currencies to obtain buildings, you can’t deprioritize them altogether.

Production in the Exploration Era still remains strong for a number of reasons even if the buildings have effectively doubled in cost. First, resources that grant production such as cotton, ivory, gypsum, tin, and tea still remain abundant (especially now that merchants are available right out of the gate–you can just trade for any of them right away). In addition, the camels are still active in this age, so you can easily power up new cities more easily. Secondly, rough, vegetated, and wet terrains all gain 1 production at this point, whereas many of them (not on desert or plain) did not have any production at all in Antiquity. All of these factors are still strong enough on their own that you probably don’t even need to utilize the node from Science attribute or masteries.

It all comes crashing down in the Modern Age and I think this is when most people have trouble with. All Modern Age buildings have roughly twice the cost of Exploration age buildings (about more than 4 times the Antiquity age’s). Resources that give production at this point are llamas (which still only gives 1 production), ivory, wine, tobacco. You may notice this already, but the productions on these resources do not catch up with the building cost throughout the ages. In addition, although the aforementioned terrains again gained extra 1 production, the impact of this gain is not as impactful as in the Exploration Age. Masteries still give +1 production to rough, vegetated, and wet tiles, but the impact is not as big as they were in previous ages.

If you did not plan your attributes and overbuilds to take advantage of the Civil Engineering card, you may struggle with the building cost a lot. I think the most effective way to combat this is the combination of overbuilding and getting the coffee resource. Coffee is a powerful, empire-wide resource that adds up quickly (like pre-1.2.5 gold & silver, albeit not at the same potency). In fact, I consider Coffee, Tea, and Kaolin as the most important and sought after resources in Modern Age now.

Civs with Unique Buildings

In my opinion, civs with unique buildings (especially those in Exploration era) were impacted more negatively than the ones with unique improvements.These buildings now cost more than regular tier specialty buildings (and yet still provide same yield #), can’t be overbuilt, non-uniform rule of adjacency, provide gold or happiness (which can be found elsewhere and with less cost), and some of their unique quarters’ effects (again, most of Exploration Era civs especially) are just not that good to justify the cost. Here are some examples:

Chola - Five Hundred Lords: Strict placement rule for Anjuvannam (being next to coast), buildings provide gold & happiness. Unique quarters provide negligible effect.

Majapahit - Pura: Unique quarter effect recently buffed but 10% gold toward converting towns to city is still negligible in my opinion. Yields are mostly happiness, with one that provides culture adjacency.

Spain - Plaza: Strict placement rule for Casa Consistorial (being next to coast), buildings provide gold & culture (ok). Unique quarters provide a small and static effect.

I am not saying all other unique quarters do not share the same problems; there are some exceptions to the rule (Maya, Maurya, Assyria, Abbasids… etc…) since these provide strong yields such as production & science. Most are quite average to weak (Egypt, I am looking at you). The production nerf, in a way, negatively impacted these particular civs more than others since they now suffer 10% should they choose to build them. If their effects do not justify the cost, you can choose not to build them, but then it also means one part of that civ’s toolkit may as well be blank. You may think the ones with production bonus help off-set the nerf; to an extent, yes, but it’s not enough when you compare them to civs with efficiency modifiers (especially the ones in the Modern era since you still need to invest a hefty production into them).

2. Big Winners

There are big winners in this patch: Great Britain, Russia, and Mughals. By “big winners”, I don’t mean they get buffed, but rather they were least impacted by the nerf. I also don’t mean they’re now Tier S, but rather they’re able to mitigate the increasing building cost enough to even go “wider” than other civs.

2a. Great Britain

Referring back to the tables to the original post, you will notice that a fully buffed Great Britain will mitigate significantly more cost than any generic civ counterparts to the point where they can afford to go wide.

For example: If I were to play Ada Lovelace, with Science Node 2 & CE slotted, “tall” (2 non-capital cities) with 10 overbuildable buildings (170% cost). If I were to play Siam, the building cost is effectively reduced to 114%. But If I pick GB, the cost is reduced to 97% and in fact, I can afford 3 more cities to reach the same cost as any other generic Civ.

Their purchasing power is only exceeded by the Mughals. They’re versatile as both “tall” and “wide” in this regard.

2b. Russia

In terms of production, if you have Emancipation Reform slotted, your cities will be neck and neck with Great Britain in terms of producing buildings, though GB still edges out just by a bit without requiring researching unique civics. This is fitting considering most of their kits involving mostly buffing up cities in science, culture… etc… Their purchasing power though is also not as good as they do not have anything to help in that regard.

2c. Mughals

Prior to 1.2.5, it feels like everyone else was the better Mughal than the Mughals themselves. That is how strong pre-1.2.5 Gold was, especially once you start hoarding them. Also, that bonus remained the same throughout the ages.

1.2.5 restored what is unique about them: Economy & purchasing power. They now have the strongest gold economy in the Modern age. Every 1 gold that you see on the map, policies, are essentially ~2.5 (1*1.75*1.4 = 2.45 ~ 2.5) if you have both the Economic node 5 & their unique tradition slotted.

The standard ratio for production vs gold is 1:4 (for any generic Civ) in cost. For the Mughals, with all modifiers available, the ratio is essentially 1:1.6 (4/2.5 = 1.6). That is roughly 2 production to 3 gold. Essentially as the Mughals with their unique civic researched, you can think of towns being able to feed cities production in form of gold. Still though, their cities still have to work themselves if they want to finish projects (Science victory comes to mind). While the Mughals struggle with finishing projects and research, I think the gold economy allows them to gain tempo quicker than other civs, which can be utilized to snowball.

3. Lastly, about the overall bigger picture...

I think one thing this patch has done tremendously well is that it added impactful opportunity cost in the form of cost incurred by expansion (internally or externally). Impactful & consequential costs have been largely missing, though not completely, in Civ7 since the release. Prior to 1.2.5, the closest things were the maintenance costs of buildings, specialists, and units but those costs were easily mitigatable that you often noticed them at all. You practically encountered bonus without any negative consequences associated in every turn that any efforts to improvise your moves in each seem to result in marginally better outcomes. I think costs do not necessarily only stress the players, but rather they can make the gameplay more interesting by forcing appropriate solutions to mitigate the costs. Now, I don’t mean “costs” here as negative narrative thrown out randomly like crises, but rather the negative consequences of your actions in game (in addition to the fitting benefits that you’re getting from the same said actions). If costs are easily mitigated and benefits clearly outweigh costs, then the choice becomes a clear yes. If costs are too severe and benefits are too little, then the the choice becomes a clear no. If both costs and benefits are too little to consider, then why bother? But if costs are severe, and benefits are also huge, I think it would create an interesting dynamic where one would ask "wait... how can I mitigate the costs?". And I don’t mean to only include costs in civ & leader designs, there should be more cost in the fundamental gameplay features in Civ7--they should be impactful and the benefit should be well worth it accordingly. This is definitely a big step in that direction to me.

 

[Krikkit1]

Cool analysis.
Seems like the "removing urban districts" is a bug (as with the cost changes in towns not being applied.)

Sort of related questions... you mentioned the Mughal multiplier is multiplicative with factory resource bonuses. Are those bonuses global as well, and are all of the global multipliers (Himiko Shaman bonus, Noli MiTangere, etc.) multiplicative with each other and the factory resources, or are the Mughals a unique case

 

[phoenixician]

...

Sort of related questions... you mentioned the Mughal multiplier is multiplicative with factory resource bonuses. Are those bonuses global as well, and are all of the global multipliers (Himiko Shaman bonus, Noli MiTangere, etc.) multiplicative with each other and the factory resources, or are the Mughals a unique case

I might have to re-test the factory resources again, since the last I checked their interactions was pre-1.2.5. Since 1.2.5, I was only able to test Elective Republic's celebration effects & Glass Armonica, which are also categorized as global modifiers and actually additive to Mughal's ability.

 

[Manpanzee]

Interesting analysis. Something not noted: The Limestone resource is an additional building production modifier, giving +10% hammers in both Exploration and Modern. The economy changes have now made Limestone a high-tier premium resource in my view.

Previously, I kind of had the impression that Limestone was a relatively rare resource with a lower spawn rate. However, I've been managing to hook up a lot more of it in recent games. Not sure if its rarity actually changed, or if I'm just noticing it more because I'm looking more actively.

 

[phoenixician]

Interesting analysis. Something not noted: The Limestone resource is an additional building production modifier, giving +10% hammers in both Exploration and Modern. The economy changes have now made Limestone a high-tier premium resource in my view.

Previously, I kind of had the impression that Limestone was a relatively rare resource with a lower spawn rate. However, I've been managing to hook up a lot more of it in recent games. Not sure if its rarity actually changed, or if I'm just noticing it more because I'm looking more actively.

Good catch. I forgot about Limestone. It is definitely a lot more powerful now. In the games that I've seen, limestone has been relatively rare also. It also seems to change locations throughout the ages. I am hesitant to lay down unique improvements indiscriminately (on plains) since they can prevent certain resources from spawning.

 

[MutilationWave]

I had never considered UIs preventing resource spawn. Hmm I was really spamming them in the antiquity age I just wrapped up. I need to think about this, thanks.