[SPOILERS] Cornflakes (Norway) vs. TheArchduke (Japan)

[Cornflakes]

   

   

Captured Yerevan! Had just enough muscle this turn. It didn't really come down to a dice roll, but one fewer attacker and I would have left the city red-lined ... probably would have lost at least 1 warrior and most likely 2 warriors and maybe that injured archer too if I hadn't been able to capture the city this turn. Moral of the story (which I learned in the previous duels) make sure you bring enough force. Each warrior dealt about 2x the damage of each archer. Once the warriors started hitting the HP dropped very quickly.

Also visible in the north, Stonehenge has completed! Timing synced perfectly with the capture of Yerevan For some reason the GProphet appeared in Stavanger rather than Nidaros where the Stonehenge was built. I don't know if founding the religion at Nidaros HS will spread the religion to the Stonehenge city or if it would just auto-spread to cities with Holy Sites. Therefore I'm going to spend a turn transferring it to Nidaros to found in Stonehenge just to be safe. Plus having the HS further north will put more boosted  passive pressure from the holy city further north where I'm probably going to expand in the longer-term. And also it brings Buenos Aires into range of the holy city pressure.

   

   

Founded the religion and took the Jesuit Education & Church Property beliefs as planned. I couldn't come up with a clever religion name, so I chose Confucianism mostly because I like the green color. I immediately buy the library at Stavanger because I expect Hypatia to appear in Nidaros and am trying to save a turn transferring her south. In case you're wondering, the next GProphet costs only 60. I've been in the classical era for several turns now via Horseback Riding so I take this to mean that TheArchduke has not yet entered the Classical Era, otherwise the next GProphet should have been Medieval for 120 cost. Which means I won't be as far behind to Political Philosophy as I expected.

   

Not much happened for a couple turns until T63. Currency was just slightly over 50%, so when the trader completed this turn I was able to finish Currency when I started the trade route. I then harvested the Stone that I'd been holding back for a couple turns. This completed the settler and provided enough overflow to 1-turn the Commercial Hub! Notice in the panel on the right that the Holy Site would take 4 turns to complete, but the CH only 1! This has to be the discount mechanism at work here. I didn't think to check what the actual cog cost was between the two districts. If I remember I'll load up the save again tonight and check the cost. I queued up a 9-turn settler in Stavanger and a 7-turn builder in Yerevan. Tech was set to Apprenticeship which completes in 7 since I now have 5 mines and will get a decent production boost from that tech.

[Cornflakes]

The district discount mechanism's mysteries have been unveiled! https://forums.civfanatics.com/threads/d...sm.625743/ The explanation there is somewhat opaque. After reading the post I had to scratch it out with paper and pen to wrap my mind around it, so I'll restate it here in a way that I was able to understand it better. With the following variables defined:

C = total number of completed districts (regardless of type)
P = placed districts of the specific type in question (e.g. Commercial Hub), this includes the sum of completed and uncompleted districts of the specific type
A = available districts to build, which is based on techs discovered. (e.g. Astrology + Writing + Bronze Working = 3 available districts)

For a specific district type, the discount is available only if both of the following are true:
1) C>=A
2) C/A > P

The discount does not apply to Spaceport, Aqueduct or Neighborhood and thus these districts are not included when counting up A & C. Count only Aerodrome, Campus, Commercial Hub, Encampment, Entertainment Complex, Harbor, Holy Site, Industrial Zone & Theater Square. Thus A has a range up to 9.

For an example in this game, last turn I finished Currency and noted that I had a discount on the Commercial Hub. I had 4 total districts complete (2 Campus, 2 HS) and 4 possible districts to build (Campus, HS, Encampment, Commercial). Thus C=A=4 and C/A = 4/4 = 1 > 0 = P. Both conditions are met for the first Commercial district placed. After placing the first CH, P becomes 1 thus C/A is no longer greater than P. HOWEVER, after completing the first CH at Nidaros, C/A = 5/4 which is greater than 1 and the discount then applies to the 2nd Commercial Hub.

Thus it is not always cheaper to lock in districts ASAP! With the discount of 40% it was significantly cheaper for me to wait 1 turn to lock in the 2nd CH at Yerevan ... we're talking 55 cogs cheaper! Note that the discount mechanism only looks at your districts and techs, which means that it is independent of the other players actions and a perfect opportunity to optimize build/research orders to maximize the number of districts which get the discount. In fact, with careful planning I think it would be possible to achieve the discount on nearly all of the districts built throughout a game. Theoretically you could get the 40% district discount on all districts except whichever kind you have the most of. I'll come back to this later. First, some screenshots of the CH costs here:

   

This is from the current turn T64. Here you can see the 2nd CH placed at Yerevan only costs 80. After playing this turn I went back and re-loaded T63 to confirm that the discount would not have applied that turn. And indeed as you can see in the cutout pasted in the top-left, the CH at Yerevan would have cost 136 if I had placed it down last turn. Contrary to the current convention it may actually be cheaper to wait to place a district rather than locking it in right away! With the 40% discount, you would have to research 66% more techs before the cost increase due to techs would out-weight the discount due to waiting to lock-in. In fact, it seems like it would almost always be better to wait to place down any district that is in the minority in your civ. If you already have exceeded the discount you could achieve in the near future (3rd/4th of the same type) then might as well lock in the cost ASAP. Looking at my situation now and coming up:

I currently have 5 districts completed, 4 available. After researching Apprenticeship I will have 5 available, so the first condition is met and the 2nd condition will be 5/5 so the discount applies to the 1st IZ placed. Completing any district of any type would bring the ratio to 6/5 and the discount would then apply to the 2nd district of any type. Drama & Poetry will be completing a couple turns after Apprenticeship. At that point the first condition will not be met since I only have 5 districts currently completed with 6 available after D&P. I would not get the discount on anything at that point in time. After completing the CH which was just placed at Yerevan I would then meet the first condition, and the 2nd condition would be 6/6 and apply to the first Theater district placed. In that regard it might be advantageous to delay D&P for a bit until I can lock in an IZ somewhere. On the other hand, completion of that CH would bring me back to an even 6/6 so that I could get the discount again on the first IZ.

I've just about maxed out the districts I can place based on my current population. That's not true, I have placed down all the districts that I can with the current population. I'll have to grow cities up to size 7 now to place additional districts, or plant more cities to place additional districts. The discount is a balancing act between science/culture rate, vertical expansion, AND horizontal expansion. If science & culture out-pace expansion  to the point where there are more district types available than completed districts, then district discounts don't apply. Purely horizontal expansion without completing districts make the discount not apply. The discount really encourages diversity in the types of districts construction (i.e. 3rd campus or HS would not get the discount, but 1st IZ/Theater would, and then 2nd IZ/Theater would after the 1st are completed)

Hmmm ... I could lock in the 1st IZ at the new city. That actually sounds like the best move. Lock in the IZ, build a galley Long Boat with +100% production policy, purchase a builder to chop or harvest for max overflow into completing the IZ, and use the last couple builder charges to build mines around the IZ. I'll have that new city planted before D&P completes, and will just wait to place down the Theater at Nidaros until the CH completes at Yerevan. It will probably be 4-6 turn delay which would be worth it since the 40% discount will save me at least 2-3 turns of production cost

Back to the theoretical prospect of getting the 40% discount on nearly all districts. I'll use a specific build & research order as an example, which would also work under similar circumstances with alternate build & research orders.

1) Research Astrology, build 2 HS.
2) Research Writing. C=A=2, C/A=1 so the discount would apply to the 1st Campus.
3) Research Bronze Working. After the campus is complete, C=A=3, C/A=1 so the first Encampment is discounted.
4) Research Currency. After the Encampment is complete, C=A=4, C/A=1 so the first CH is discounted.
5) By this time cities are starting to grow. You build a 3rd HS without getting the discount, which makes C/A = 5/4 thus the 2nd Encampment AND 2nd Campus AND 2nd CH now all get the discount.
6) Research Apprenticeship. The 2nd districts from (5) have already been placed and have the discount locked. C/A is 5/5 before any of the 2nd districts complete so you get discount on the first IZ, after any of the 2nd districts complete the ratio is bumped up to 6/5 and the 2nd IZ gets the discount.
7) Repeat (5) & (6) with Drama & Poetery for the Theater, Celestial Navigation for the Harbor, etc. Eventually you'll build the 3rd HS without the discount to enable the 3rd district of each type to get the discount. In theory you could get the discount on every district except the HS in this example, as long as you continue building the other districts evenly. And with the discount at 40% it almost makes the other districts worth building evenly to maintain the discount, especially early on when there aren't many district types available.

[Krill]

Wtf.

So what happens if you ignored HAS?

[Cornflakes]

I'm not sure what you mean by HAS ... Holy Sites (HS)? The important point is that you get the discount when you are pursuing a more balanced approach and building all types of districts equally. If you want to completely ignore one type of district you will have to build 2 types of districts at full cost in order to get the rest at the discounted price. In my case I have completely ignored Encampments thus far. I'm able to get away with that since I have an "extra" HS that I captured in Yerevan rather than an encampment. Types available was 4, and I had 4 from [2 campus + 2 HS]. Had I built all the districts myself rather than capturing one, the most efficient way would have been [2 campus + 1 HS + 1 Encampment] since the Encampment would have been discounted but the 2nd HS would have been at full price.

The district discount mechanism highlights another stealth-buff to Japan, in that they can leverage their half-price districts to more easily get the discounts on the rest of the district types.

[Krill]

My takeaway from your post is that you could ignore one type of district and still make it work: if the first district type is a campus, second is encampment, third is commercial district, fourth is Theatre or entertainment and fifth is industrial zone all that works out.

The reason I'm asking is that I want to know if the meta of encampment, fighting over the GG and trying to kill a player or CS will be affected be a potentially effective building option. I doubt it though.

[Cornflakes]

An empire-wide update since I haven't done one in a few turns ...

   

Starting in the north ... I found a mid-sized desert, lots of mountains in the east, and a natural wonder of some sort in the west (grants +1/+2 to adjacent tiles). The river which passes by Buenos Aires appears to extend towards the natural wonder, so I might be able to get a decent Petra city in that region. Also desert opens up the possibility of Pyramids, which as analyzed in depth in the Rome game is almost a no-brainer if I can swing it. Maybe luck out with a good coastal location that can chop a ship out with the +150% production bonus and overflow into the 'Mids. The northern coast is dry unfortunately. Might get a decent filler-type city among the mountains though.

The new quest for Buenos Aires is "Recruit a Great Engineer" which I will hopefully accomplish but not in the near future. The first GEngineer grants the eureka for Printing (2 Universities) and has 1 charge which allows a city to build 1 more district than the normal limit. The extra district would be useful in Nidaros, but I'd probably get the eureka anyway through Jesuit Education. Overall a snooze GEngineer and I wouldn't mind if TheArchduke takes that one. While on the subject of GPP, TAD was generating ran the +2 GScientist card for 4 turns, and recently swapped out of it. I think he has now a 3rd HS completed since empire score jumped up again and he's now at +3.4/turn indicating Classical Republic government.

   

Now the south ... Warrior exploring around Tsingy found what will likely be the area for the settler coming out of Stavanger. I'll give a close-up of that area later and discuss specific options for that city's location. The western coast is dry without mountain to make an aqueduct-city viable   No clusters of crabs along the tundra shore either like I usually see. The units in the southeast will explore beyond Auckland, while the archer in the southwest will follow that peninsula to its conclusion. After the Tsingy city it looks like I have a decent city around the incense south of Nidaros and another along the eastern river around the Tea & Stone.

   

Close-up of Nidaros. The monument was a 3-turn build, and it's time to improve my culture rate some. Nidaros is getting +4 towards buildings from the two Industrial CS so I'd like to keep it on buildings for now and let the other cities handle the settlers & builders. The Monument should have a turn off of D&P, and after reviewing the above analysis I'm going to delay Apprenticeship 1-turn in order to complete D&P first and lock in the Theater at Nidaros with the 40% discount. I'll swap research to Shipbuilding after the new city is founded on the coast, which will be just barely a 1-turn research. Nidaros will be up to Size 7 and able to support the 3rd district in a couple of turns. I'll probably go Granary next which will allow 2 more pop without penalty, and then the last 2 for size 10 would be at penalty unless I build a couple of farms (which I may do since I need to get 6 farms somehow for the Feudalism boost).

   

Moving on to Stavanger & Yerevan, badly need some builder labor here! Apprenticeship will give a nice production boost to Yerevan. I'll probably go Finish Campus (7T?) > Builder > Builder. Stavanger I'm thinking Builder > Builder [chop jungle] > Trader > Settler (chop jungle or forest) > Trader. Or something like that.
 
   

Now for that close-up of the Tsingy location. There are 3 potential city sites that I see. Each have their pros and cons:

1) Pros: settles on a tundra tile, has grass forest and fish first ring for the first 2 pop to work at food neutral, whales 1st ring (easy to get Celestial Navigation boost), excellent harbor site 1st ring. Cons: Pushes the stone & copper to 3rd ring, many coastal tiles 2nd ring so in addition to low-tile-yields it creates a long-term vulnerability against naval attack(somewhat mitigated since this is on a southern coast rather than east or west), also outside the Holy City passive religion spread range.
2) Pros: inside Holy City passive religion spread range, stone + copper moved to 2nd ring, off the coast for better 2nd ring tiles and less vulnerable to naval assault. Cons: settle on forests eliminates chop/lumbermill potential, harbor loses city adjacency bonus and is 2nd ring and will require tile buy, tiles available for first couple of pop to work are marginal compared to (1)
3) Pros: inside Holy City passive religion spread range, preserves another forest and can work the forest with 1st pop for decent yields. Cons: harbor tile is 3rd ring, Whales 3rd ring so harder to get Celestial Navigation boost, has more tundra 2nd & 3rd ring (but less coast so basically a wash)

Here's another thought. I could build a Market in Nidaros instead of the Granary. This would reduce my ETA on the GMerchant and essentially make it a lock for me to get it, along with the 2nd CH that will be completing in about 10 turns. I'd get the GMerchant about 10 turns after founding that city, and could use it to pick up 3rd ring harbor & whales. Essentially negating all the cons for (3). I'd build the Harbor as the first district there, so that I could chop the forest & stone through ships to complete the Holy Site & other infrastructure. Might not be worth it to slow down growth at Nidaros though. I'll eventually need a city to 10 pop and this is realistically the best place for it.

[Bacchus]

I note that the two conditions are actually one condition:

FLOOR(C/A) > P

[Cornflakes]

I note that the two conditions are actually one condition:

FLOOR(C/A) > P

No, that condition fails when in fact the discount applies in the example from my 2nd Commercial Hub: C=5, A=4, P=1 ... FLOOR(5/4) > 1 ... FALSE because the left side evaluates to 1

If we change to FLOOR(C/A) >= P it works for the above example, but fails where P=0 and C<A, thus I believe an additional condition is necessary.

[Bacchus]

That is so weird... I'm just trying to wrap my head round how they came up with something like this.

C > A * max(P, 1)?

EDIT: No, that still doesn't work.

[Quagma Blast]

That is so weird... I'm just trying to wrap my head round how they came up with something like this.

C > A * max(P, 1)?

EDIT: No, that still doesn't work.

I think the idea is that they had behavior they wanted to encourage, and ended up making mechanics that did that if you know what they are but are opaque enough that the playerbase missed them (especially with the way price lock-in encouraged setting districts as soon as possible, while the mechanics counted completed districts.)

They wanted people to build districts evenly, so your "underbuilt" districts are cheaper to construct.

They also wanted to encourage early district building/not give people a bonanza of cheap super-early districts, so they set a floor of district development below which you don't get the bonuses.

Finally, they didn't want to punish people for districts that can't be built and might be eras away, so they set their discount mechanics to scale with the number of kinds of districts available.

EDIT:
Okay, I think I've got a reasonable facsimile of the thought process that came up with those formulas in particular.

If you want to build a district, first, calculate the expected number of that type of district you should have, given your empire's overall development. Call this E. Using Cornflakes' terminology, E=C/A. Or, E=C/[max(A, 1)] just to avoid divide-by-zero errors. Basically, you ought to be building every type of district equally, so it looks at how many you actually have managed to build and divides them up equally even if that results in fractional districts. (Mathematically, of course, this is unnecessary, but I think explicitly naming and calculating this stat or something similar was a significant contributing factor to the "but why this formula" question.)

Next: is 1>E? If so, no discount. You shouldn't even have one of that district yet! Of course you pay full price.

Then, of course, is P>E? That is, have you at least started building at least as many of that kind of district as we expect? If not - get out there! Build those, you don't have enough of them. Have a discount to encourage you.