1.3.3: Influence of Infrastructure on Industry in Numbers

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Status
Not open for further replies.

Fyras

Sergeant
93 Badges
May 21, 2014
95
67
  • Stellaris
  • Hearts of Iron III Collection
  • Europa Universalis IV: Res Publica
  • Semper Fi
  • Europa Universalis IV: Pre-order
  • Cities: Skylines - After Dark
  • Europa Universalis IV: Cossacks
  • Cities: Skylines - Snowfall
  • Europa Universalis IV: Mare Nostrum
  • Crusader Kings II: Charlemagne
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Field Marshal
  • Steel Division: Normand 44 - Second Wave
  • Crusader Kings II: Monks and Mystics
  • Stellaris - Path to Destruction bundle
  • Stellaris: Synthetic Dawn
  • Europa Universalis IV: Third Rome
  • Hearts of Iron III
  • Crusader Kings II
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: Rajas of India
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sunset Invasion
  • Crusader Kings II: Sword of Islam
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Hearts of Iron III: Their Finest Hour
  • For the Motherland
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: Rights of Man
  • Europa Universalis IV: Mandate of Heaven
  • Cities: Skylines - Mass Transit
  • Steel Division: Normandy 44
  • Warlock 2: The Exiled
  • Cities: Skylines
  • Europa Universalis IV: El Dorado
  • Stellaris: Leviathans Story Pack
  • Stellaris: Digital Anniversary Edition
  • Tyranny: Archon Edition
  • Crusader Kings II: Reapers Due
  • Hearts of Iron IV: Colonel
  • Crusader Kings II: Way of Life
  • Stellaris: Nemesis
  • Pillars of Eternity
  • Europa Universalis IV
  • Europa Universalis IV: Common Sense
  • Crusader Kings II: Conclave
The new bonus to construction speed based on infrastructure adds another factor to think of to the game: Just as a reminder, for every level of infrastructure in a province you get a 10% boost to construction speed. (so at 10/10 infrastructure your speed is doubled).

So the fellow general might ask himself: "Is it worth it to build up on infrastructure purely to boost construction speed?"

I asked me myself this question und after the better part of the evening I managed to come up with a excel spreadsheet to calculate the influence of infrastructure.

(If you just want the tl;dr; scroll to the end of the post).

The base formular with the new construction-system for a full 15 line in IG-days:


CodeCogsEqn1.png


The "Speedbonus" is depending on your Trade and Economic Laws, National Foci, Advisors and Construction Technology. Depending on your picks it may also vary between the different industries. (i.e.: War economy rewards a +20% to MIC construction but 0% to everything else). The infrastructure buff only applies to industries (CIC, MIC and NIC) not to itself, AA, forts...

So using this formular we can start make some calculations:

Many speedboni are affecting infrastructure and industries equally (Trade Laws, Construction Tech...). The economy laws though are impacting the equilibrium heavily as their boni (or in fact mali) only apply to industry. For our analyses we will pick Civilian and War economy:

In Civilian Economy there are certain sweet spots, where it is actually beneficial to build infrastructure (if you don't have any higher INF provinces available) to boost overall construction speed:

neu_civ_cic4.png


At an infrastructure level of 4 (or of course lower) it is actually going to be beneficial to build at least 1 infrastructure. The higher the number of new CIC the higher the bonus of any additional infrastructure. In comparison the numbers are a hell lot closer with an already build infrastructure level of 5:

neu_civ_cic5.png


If you plan to build more than 5 CIC, infrastructure is (even though lesser extend) viable in war economy:

neu_war_cic10.png


So depending on how many CIC you are going to be able to build in a set province, infrastructure is always a viaable option!

But what is about MIC? The cost is considerably lower and also the economic laws are influencing them more positively. Let's take a glance at Civilian Economy:
neu_civ_mic7.png


At an infrastructure level of 4 (or below) we are getting the sweetspot at 7 MIC (or more). Otherwise just plain building them is more efficient!

During war economy there is no reason in a normal setting (1 infrastructure provinces with 10+ building spaces are not in the game) to ever build infrastructure as a speedboost:

neu_war_mic7.png


Because of the even lower IC cost of NIC, they are always better of being build right of. (In a 4 inf level province the sweet spot starts at at 10 NICs - and those provinces are already very rare).

The tl;dr therefore is:

If you don't want to calculate everytime you are building new stuff, you can use these rules of thumb:
  1. During Civilian Economy for CIC: Building Infrastructure is viable if the difference between INF-Level and planned number of CICs is at least 2. The higher the better. You can safely increase the INF-Level to the number of planned CICs for certain net gain. If the planned number of CIC at least 7 you can rise INF level all the way to 10 savely.
  2. During Civilian Economy for MIC: You can raise your INF-Level by half the existing INF-Level, if the starting level is below 5 and you build at least 7 MIC.
  3. During War Economy for CIC: Building Infrastructure is viable if the difference between INF-Level and planned number of CIC is at least 3. Above INF-Level 4 you can savely increase the INF-Level to the number of planned CICs for certain net gain. If the planned number of CIC is at least 7 you can raise INF level all the way to 10 savely.
  4. During War Economy: It is almost never beneficial to build infrastructure to boost MIC construction speed. The examples are very unlikely.
  5. If you need to build infrastructure because of other reasons, try to build it in provinces with higher industry cap to minimize your effective CIC loss.
  6. If there is not a strategic reason (like SU building their industry right on the German border...) in the way, always build industry in provinces with the highest infastructure possible. The order should be CIC>MIC>NIC for long term investment!
---

PS: I'm by no means a statistical analyst in RL, so there might be mistakes... If you find one - please share it! :)
 

Attachments

  • 133 Infrastructure 2-0.xlsx
    19,7 KB · Views: 527
Last edited:
thanks for sharing this clear analysis
 
Good job on this! one thing you might want to add is the continuous focus for construction engineering as that adds +10% to infrastructure building.
 
Good job on this! one thing you might want to add is the continuous focus for construction engineering as that adds +10% to infrastructure building.

Right... forgot about that. Might add it tomorrow. Good thing it doesn't change the analysis in any important part: By the point you are reasonably able to take that focus, you should be already at least at partial mobilisation... The 10% won't be a game breaker then.

Thanks for the feedback though!
 
I hate to be the guy who disagrees with something that you clearly put some effort into, but I must.

Just now I tested building 1 infrastructure at construction tech 4. (since construction tech proportionally increases both infrastructure and factory building (of all types) the fact that I am at a higher level should not drastically alter the results (though the actual numbers will shift). building 1 infra at construction tech 4 takes almost exactly 30 days. Even at the base level of technology it is 3000/75 = 40 days (assuming you are figuring for 15 CIC on the project. Again this is proportional so fewer CIC changes the numbers but not the actual result).

Unless I misread what you said, how can building 5 CIC take 100 days longer if you first upgrade the infra by one when that upgrade itself only took 40 days?? One CIC at base tech level and infra 4 is 10800/(75*1.4) = 103 days (rounded up slightly). for infra 5 it is 10800(75*1.5) = 96 days. so you save 7 days on each. Building 5 of them gets you back 35 of the 40 days spent. It ends up costing you 5 days to upgrade first not over 100.

Your initial formula in the first white box combines too may elements. I suppose it almost works if you cancel most of them out by setting the quantity to zero on things you are not building. It may still hold true that building infra will not save you a lot of time in most cases, but what it costs you is nowhere near as high as claimed and the cost becomes less in later years or with other construction bonuses. The savings is also proportionally less.

My take on this is that in a province with more than 5 unused slots (don't forget to factor in the expansion slots from raising concentrated/dispersed industry) you will eventually come out slightly ahead. Now this is for CIC which is a more expensive item. For MIC you do not break even until your 7th building post infra (so ahead on the 8th) upgrade instead of coming ahead on the 6th.

The bottom line is that it is not always best to raise infra first but there are certainly parts of the map where it will be. Much of the Japanese home islands fit this case.
 
Go back to where you are building 5 CIC in 4 infra.
No advisors (+0%), no techs (+0%), Civilian Economy (-30% Factory construction), Export Focus (+10% Factory output), 15 CIC input.
To build 1 CIC you need 10800/(75*(1+0.1-0.3))/1.4=129 (128.57) days. To build 5, you need 643 (642.85) days.
To build 1 infra you need 3000/(75*(1+0.1))= 37 (36.36) days.
To build 1 CIC in now level 5 infra you need 10800/(75*(1+0.1-0.3))/1.5=120 days. To build 5, you need 600 days.
Sum up 1 Infra and 5 CIC gives you 637 days - 6 days advantage, over the other!
Wait! There is more!
All those shiny new factories are producing stuff too, right?
So, what do we get in additional factory production after 643 days for both tests?
Without building 1 infra, you get:
1 factory each for: 514 (643-129), 385, 257, 128, 0 days. To sum it up: 1284 IC/days not spent on building 5 factories.
Building 1 infra first, you get:
1 factory each for: 486 (643-37-120), 366, 246, 126, 6 days. Or 1230 IC/days. Additionally you get those 15 initial factories for 6 last days! Bringing total to 1320 IC/days!
Thus, it pays off to build that 1 infra
 
I hate to be the guy who disagrees with something that you clearly put some effort into, but I must.

Just now I tested building 1 infrastructure at construction tech 4. (since construction tech proportionally increases both infrastructure and factory building (of all types) the fact that I am at a higher level should not drastically alter the results (though the actual numbers will shift). building 1 infra at construction tech 4 takes almost exactly 30 days. Even at the base level of technology it is 3000/75 = 40 days (assuming you are figuring for 15 CIC on the project. Again this is proportional so fewer CIC changes the numbers but not the actual result).

Unless I misread what you said, how can building 5 CIC take 100 days longer if you first upgrade the infra by one when that upgrade itself only took 40 days?? One CIC at base tech level and infra 4 is 10800/(75*1.4) = 103 days (rounded up slightly). for infra 5 it is 10800(75*1.5) = 96 days. so you save 7 days on each. Building 5 of them gets you back 35 of the 40 days spent. It ends up costing you 5 days to upgrade first not over 100.

thanks for the input. You are right, there is a mistake in the sheet. After my short pre work review, it seems that I somehow have an additional 2 inf costs in every calculation... The formula itself should be accurate though.

Guess that happens if you stay 100% in theory. Will crunch down on it later today!

Thanls!
 
The post is now up to date! Calculations in Excel are now working as intended (or so I hope)! At least I can now safely reproduce all your hand made examples! So optimize away with THIS new version!

Thanks again to your feedback!
 

Attachments

  • 133 Infrastructure 2-0.xlsx
    19,7 KB · Views: 464
Thanks! This seems in line with my quick analysis too. Yours is a lot more detailed though. The last upgrade from 9-10 seems to be the least worth one. For dockyards it seems to take a lot of them for it to break even.
 
Status
Not open for further replies.