Transportation vs Infrastructure

[Vohen]

Reading the recent DD on the trade rework, something struck me as strange for trade centers: the lack of transportation usage.
This makes sense if they are on coastal states, as that would be the new merchant marine good, but on an inland one?
No, they need to get those goods to the coast somehow before the MM can take it to the world market.

But they do use infrastructure, which is what the game uses to represent this I guess.
Problem is, infrastructure is free, while transportation is not. This creates issues, especially for railways, where buildings are using their output for free, making them unprofitable while still very much necessary.
Labor saving PMs for transportation are very odd in this regard too. Their only effect for the building is to save labor, but with them deactivated the building is still using railways (or the infra they create) for free.

Ideally, labor saving PM should use transportation even on the first level (with the amount increasing as you decrease the required laborers).
To compensate, peasant farms should also produce some token amount of transportation, and maybe even give urban centers a boost to their transportation PMs, representing the infra source from population. State traits, tech and other modifiers could give a throughput bonus here as well.
Automobiles by themselves shouldn't give infra, but be used in the last levels of the labor saving PM. Pops would still consume them as usual though.
(Obs: this would make automobiles very good at saving labor, as not only would they be on a labor saving PM, they'd also reduce the demand for transportation, making railways less profitable, firing people and downsizing in the process).

Essentially, wherever there's infra usage, there should also be transportation input to go with it.
This would create a more or less direct correlation between infra and transportation, where low market access implies a high cost of transportation, being very low access where there's a shortage.

My problem with the way it is now, is that there are a lot of goods being transported essentially for free, the trade center case just made it more clear to me.
Am I missing something here? Or does someone else also feels the same?

 

[yurcick]

I agree with you all the way.

To be honest, I think the solution is more or less apparent: MA losses should not be deleted from the economy, but should go to early middle strata shopkeeper jobs, which would consume transportation.
However, the hint that logistics rework is upcoming sometime, “but not in 1.9”, is all we have that’s close to acknowledging the issue.

 

[Vohen]

I agree with you all the way.

To be honest, I think the solution is more or less apparent: MA losses should not be deleted from the economy, but should go to early middle strata shopkeeper jobs, which would consume transportation.
However, the hint that logistics rework is upcoming sometime, “but not in 1.9”, is all we have that’s close to acknowledging the issue.

I hope so, but I don't remember this being mentioned in their future plans chart.

Though to be fair, what I mentioned is very much doable with mods.

 

[IsaacCAT]

Are not merchant marine the transport cost that trade centers will pay in 1.9?

 

[TheFlemishDuck]

The way i always sugested it, is that transportation should allow "MAPI circumvention". AKA, instead of a labor saving PM, the factory uses rail to buy and sell at the market price, not the local price, while a factory that doesnt buys and sells at local price.

MAPI to me as a mechanics seems to represent "inabillety to sell at market price due to distance and/or transport cost". It's a functional tool to distinguish imho that which is not connected to the rail network and neither required a investment in rail and that which is. Instead that the money would be lost to Mapi when its trade between provinces, having "the factories connected to rail" would mean the money goes to the transportation services, which would make sense. You can reward the player for having provided rail access to it's industry, albeit that if you haveto few for demand then the price is high and depresses factory margins so you also have to provide a given quantity depending on how much you need. So that would also perhaps stimulate some micro, as you might think the priority for one industry to have said access might be greater than for another when supply is limited. It's a cost but also a gain, one where you need to balance the cost to get the gain which requires investment.

 

[Vohen]

Are not merchant marine the transport cost that trade centers will pay in 1.9?

Over seas yes, but they don't apply on land.
As the DD showed, even landlocked states like Brandenburg will have trade centers using MM.
Which is fine in and of itself, MMs are produced in ports, so one could infer that the goods traded in Brandenburg are being loaded into/out of that port.
However, how those goods reach that port/TC is the question, which is why it should use transportation.
And I'd argue that it should apply to coastal states even, as they tend to have cities and resource extraction inland as well, so their goods would also need to reach the coast.

The way i always sugested it, is that transportation should allow "MAPI circumvention". AKA, instead of a labor saving PM, the factory uses rail to buy and sell at the market price, not the local price, while a factory that doesnt buys and sells at local price.

MAPI to me as a mechanics seems to represent "inabillety to sell at market price due to distance and/or transport cost". It's a functional tool to distinguish imho that which is not connected to the rail network and neither required a investment in rail and that which is. Instead that the money would be lost to Mapi when its trade between provinces, having "the factories connected to rail" would mean the money goes to the transportation services, which would make sense. You can reward the player for having provided rail access to it's industry, albeit that if you haveto few for demand then the price is high and depresses factory margins so you also have to provide a given quantity depending on how much you need. So that would also perhaps stimulate some micro, as you might think the priority for one industry to have said access might be greater than for another when supply is limited. It's a cost but also a gain, one where you need to balance the cost to get the gain which requires investment.

That's a good point, perhaps ideally the prices a building sells/buy goods for should be a weighed average for local and market prices.
So if a building produces 15 units, with local demand of 10 and local price of 6, and market price of 8, the actual unit price would be (6*10 + 8*5) / 15 = 6.66. Same deal with inputs.
Only issue would be on the few goods that can be produced by more than one building, but I think it'd still be an improvement over the current system, where money is just deleted.

 

[IsaacCAT]

However, how those goods reach that port/TC is the question,

They use each state infrastructure.

Infrastructure is a capacity and it is used to simulate transport while "transport" is a cost to automate.

They are two different things and the choice of words is misleading.

I understand that logistic networks will have different infrastructure usage per state depending on goods transported between states, from origin to destination. But infrastructure is not a discrete cost but a capacity.

Thus, you do not pay directly for infrastructure aka transport. You pay for automation (weirdly named transport).

Infrastructure - Victoria 3 Wiki

vic3.paradoxwikis.com

But I agree this could be improved, but the cost of opportunity must be weighted against other issues that may bring more benefits compared to costs (including computational)

Devs are already thinking about a rework but this is what it is now.

 

[TheFlemishDuck]

That's a good point, perhaps ideally the prices a building sells/buy goods for should be a weighed average for local and market prices.
So if a building produces 15 units, with local demand of 10 and local price of 6, and market price of 8, the actual unit price would be (6*10 + 8*5) / 15 = 6.66. Same deal with inputs.
Only issue would be on the few goods that can be produced by more than one building, but I think it'd still be an improvement over the current system, where money is just deleted.

Something like that, or the using factory chooses to sell where its most profitable for it, which might be indeed a part to the market and a parti locally, as rail prices and transport availabillety might also dictate.

That would still work though withought nessecarily having a "network of rail", and i'm kinda interrested in ideas to which this could be perhaps implemented in a fashion that is not too hard on the CPU, so my thought was using certain "rules" that dictate if you can build a railroad somewhere:
1: Your first railroad must always be build in the capital
2: any other railroad must be adjacent to a province that has a railroad, otherwise that province cant build a railroad yet
3: any overseas territory requires a port to build the railroad from, and requires that atleast 1 port in the nation already has a railroad connected to it to be able to build that railroad in the port
edit: 4: if a railroad touches a foreign railroad, a larger land connections is established for the purpose of trade

Similarly, if a province has a railroad to it, that factory can sell to market price instead of local price, and if there is a port or tradeport that has a railroad in the nation, any factory connected to rail in the nation can also sell at global market price. With such conditions i think you would hence always have a existing rail network between a port and the factory using the rail before it would be able to sell at global market price.

 

[Vohen]

Thus, you do not pay directly for infrastructure aka transport.

That in and of itself is weird, isn't it? I don't think transportation should be free.
Also, it's the reason railways will struggle for profitability in many cases.

Devs are already thinking about a rework but this is what it is now.

I see, fingers crossed for next year.

Something like that, or the using factory chooses to sell where its most profitable for it, which might be indeed a part to the market and a parti locally, as rail prices and transport availabillety might also dictate.

Selling locally should still be prioritized imho, it's the main component for concentration of production chains and specialization (and was the whole point of introducing MAPI in the first place), so it's ultimately a good gameplay feature.

That would still work though withought nessecarily having a "network of rail", and i'm kinda interrested in ideas to which this could be perhaps implemented in a fashion that is not too hard on the CPU, so my thought was using certain "rules" that dictate if you can build a railroad somewhere:
1: Your first railroad must always be build in the capital
2: any other railroad must be adjacent to a province that has a railroad, otherwise that province cant build a railroad yet
3: any overseas territory requires a port to build the railroad from, and requires that atleast 1 port in the nation already has a railroad connected to it to be able to build that railroad in the port

Similarly, if a province has a railroad to it, that factory can sell to market price instead of local price, and if there is a port or tradeport that has a railroad in the nation, any factory connected to rail in the nation can also sell at global market price. With such conditions i think you would hence always have a existing rail network between a port and the factory using the rail before it would be able to sell at global market price.

If we're talking about an in depth overhaul, HoI4's railway model is pretty interesting, and gives a "railway tycoon" vibe which I quite like.
Honestly, I find it insane how HoI has a better railway system than Vicky, but oh well.

 

[IsaacCAT]

If we're talking about an in depth overhaul, HoI4's railway model is pretty interesting, and gives a "railway tycoon" vibe which I quite like.
Honestly, I find it insane how HoI has a better railway system than Vicky, but oh well.

I have read a bit about the HOI IV system.

I agree that clicking the map to build railroads and see them growing a network must be gratifying.

However, in HOIIV supply is radial from hubs and it has no concern on financial viability.

Main differences I find with Vic3:

A) transport of goods is more complicated to calculate than distance from hub to troops. It involves many point to point changing flows

B) transport companies will have to pay for transportation and all trade will have to be explicit. Like trade centers from market to world market, transport companies will harbor all trade routes.

For these reasons, I do not feel going full logistics is a good idea.

On the other hand, I would love to have railroad companies. I will post a suggestion for railroad buildings to be only possible to build by railroad companies that will add all the different states railroads buildings. An oligopoly.

Infra and transportation will work as now but a new lens will allow players to build a companies railroad networks. Visualizing each company network through states.

Multiple companies can share the same states as companies will have each different levels.

The government will be able pay for the rails and sell it to the company it choses to.

 

[Vohen]

I have read a bit about the HOI IV system.

I agree that clicking the map to build railroads and see them growing a network must be gratifying.

However, in HOIIV supply is radial from hubs and it has no concern on financial viability.

Main differences I find with Vic3:

A) transport of goods is more complicated to calculate than distance from hub to troops. It involves many point to point changing flows

B) transport companies will have to pay for transportation and all trade will have to be explicit. Like trade centers from market to world market, transport companies will harbor all trade routes.

For these reasons, I do not feel going full logistics is a good idea.

On the other hand, I would love to have railroad companies. I will post a suggestion for railroad buildings to be only possible to build by railroad companies that will add all the different states buildings.

Infra and transportation will work as now but a new lens will allow players to build a companies railroad networks. Visualizing each company network through states.

Multiple companies can share the same states as companies will have each different levels.

The government will be able pay for the rails and sell it to the company he choses to.

It absolutely would need some adaptations, but I'm not referring to the supply of troops from hubs, but the railway network itself.
More specifically, how it deals with bottlenecks in the network, the analogy here would be that states are the supply hubs and the railway network connects them to the market capital (in HoI4, just the capital).
No need to radiate supply from states (hubs) in this case, the access to the railway network (with bottlenecks considered) by itself would determine market access.

So having a railroad in Kursk means very little if you cannot connect it to St Petersburg, and having a lv10 railway there will leave bottlenecks if you cannot connect to other lv10 railways all the way to the market capital, and consequently market access will be limited.
Of course, there are also ports and such, so connecting Vladivostok to St Petersburg would happen by sea at first, until you build the Trans-Siberian railway (this is something HoI4 models very well).
I don't have a problem at all with railways being a building and working within the building framework, but they could also be represented on the map and be integrated on an actual network.

 

[exi123]

They use each state infrastructure.

Infrastructure is a capacity and it is used to simulate transport while "transport" is a cost to automate.

They are two different things and the choice of words is misleading.

[...]

But this also the root cause for the total weirdness of infrastructure and transportatrion. They are two different things for sure, but a rail without train is worthless and a train without rail is also worthless. Why was this split at all?

All kind of buildings use infrastructure, which is fine. But they do not pay for it. There is only the indirect way of paying taxes which in return can be used to build infrastructure... okay fine.

But what about private industries that use vast amounts infrastructure but "pay" MAPI? The money is lost, why is this not connected to transportation? Urban centers produce transportation, rails produce transportation, let them profit from the local industries. Industries could potentially have a pm that uses transportation, which in return reduces MAPI as it is used to send the goods to the market. An industry without transportion PM could even not be able to sell outside of the current state, which would be a really cool mechanic to build a local resource industry that only feeds the profitable factories within the state.

City centers and railways are also two of the worst performing businesses. Urban centers are never really running well (why do we even have the last two PMs here, I never get them profitable even in the super late game), even in super wealthy states and railways run on super low and tight margins that they are never really invested in.

 

[TheFlemishDuck]

I think we can not expect that the devs will implement a system where transport costs will be calculated on distance. The suggestion i made only considered transport cost as a sort of supply you have locally and connected to the market, really that considers more the cost of loading and offloading things on trains and leasing them, irrespective of how far they go. Though i considered that you could also add a modifier "lenght of transport network" which would add something to the average cost the higher it is.

I think an adjacency requirement is something that might be simple to implement, as in "a railroad cannot be build if its not adjacent to another province with a railroad, or when its not the market capital, or when its not a port connected to the market capital by port. This prevents that the player would be placing down a disconnected set of railway and only in the provinces he's developing at the time, which doesnt feel very immersive to the concept of having a rail network and connections. With adjacency requirements, the network has to start somewhere and you need a continuous line of provinces connected to the network. That seems something that seems easily enough to implement and feels right, sure that might make it harder to have your railroad in Siberia early but then again it would feel good if the player would be kinda forced to build slowly from Moskou to Vladivostok, having to wait for each first level to complete to build the next one. Lets say that you can add the next one in the queue though if you already put the connecting one in the queue before it, but you have to complete the one before the other can be completed. Building transcontinental rail created such iconic projects back in that time it would be nice to have them imho.

I generally agree that it would be more fun and immersive that railways in game would be railways with a network and an advantage to bringing products to larger markets, where Mapi can seemingly be such a usefull tool to distinguish between that which is not connected to it and that which is though with the cost of transport added but withought the loss that Mapi otherwise creates.

I wouldn't add bottlenecks, in the system i propose 1 level is enough to expand the network however the size of the amount is more dependent on how much loading and unloading the stations must do. Lets say that all rail connections are bi directional, in principle that allows a lot of capacity because trains can just drive behind eachother over the whole lenght in each direction, what you dont want is them having to wait in stations when they dont need to and not needing to wait longer than they have to if they have too, so you want a station of a certain size that is able to take the volume that must be loaded and offloaded, or with other words some stations have 2 tracks others have dozen for example. Thats what i think of the railway building, its primarily a station including the bi-drectional rail that runs the lenght of the province to the networks it connects to, a larger building means a larger station but the bi directional rail doesnt nessecarily turn into 4 or 8 rails, and afaik thats usually how railways work anyway one bi directional line can take huge loads but its more the control systems and they who control it from stations that will need to be good to manage a lot of time tables i guess. And its also then related to the sheer demand of the factories, the more they need to transport by rail the more railway workers needed for loading and unloading, similar to ports.

 

[exi123]

I think we can not expect that the devs will implement a system where transport costs will be calculated on distance. The suggestion i made only considered transport cost as a sort of supply you have locally and connected to the market, really that considers more the cost of loading and offloading things on trains and leasing them, irrespective of how far they go. Though i considered that you could also add a modifier "lenght of transport network" which would add something to the average cost the higher it is.

I think an adjacency requirement is something that might be simple to implement, as in "a railroad cannot be build if its not adjacent to another province with a railroad, or when its not the market capital, or when its not a port connected to the market capital by port. This prevents that the player would be placing down a disconnected set of railway and only in the provinces he's developing at the time, which doesnt feel very immersive to the concept of having a rail network and connections. With adjacency requirements, the network has to start somewhere and you need a continuous line of provinces connected to the network. That seems something that seems easily enough to implement and feels right, sure that might make it harder to have your railroad in Siberia early but then again it would feel good if the player would be kinda forced to build slowly from Moskou to Vladivostok, having to wait for each first level to complete to build the next one. Lets say that you can add the next one in the queue though if you already put the connecting one in the queue before it, but you have to complete the one before the other can be completed. Building transcontinental rail created such iconic projects back in that time it would be nice to have them imho.

I generally agree that it would be more fun and immersive that railways in game would be railways with a network and an advantage to bringing products to larger markets, where Mapi can seemingly be such a usefull tool to distinguish between that which is not connected to it and that which is though with the cost of transport added but withought the loss that Mapi otherwise creates.

I wouldn't add bottlenecks, in the system i propose 1 level is enough to expand the network however the size of the amount is more dependent on how much loading and unloading the stations must do. Lets say that all rail connections are bi directional, in principle that allows a lot of capacity because trains can just drive behind eachother over the whole lenght in each direction, what you dont want is them having to wait in stations when they dont need to and not needing to wait longer than they have to if they have too, so you want a station of a certain size that is able to take the volume that must be loaded and offloaded, or with other words some stations have 2 tracks others have dozen for example. Thats what i think of the railway building, its primarily a station including the bi-drectional rail that runs the lenght of the province to the networks it connects to, a larger building means a larger station but the bi directional rail doesnt nessecarily turn into 4 or 8 rails, and afaik thats usually how railways work anyway one bi directional line can take huge loads but its more the control systems and they who control it from stations that will need to be good to manage a lot of time tables i guess. And its also then related to the sheer demand of the factories, the more they need to transport by rail the more railway workers needed for loading and unloading, similar to ports.

They really do not need to reinvent the wheel here. For me, the game already offers basically everything we need, it just needs to get connected together. Railways need to be stacked when huge industries need to get connected, from iron and coal to steel and all the like. This is well represented with infrastructure. That huge urban areas provide lots of infrastructe and transportation is also fine. Building in already heavy developed areas should always be the first option, going out to new resources, connecting with railroad should come later, pushing urbanisation further into new territories.

Resources can keep their labour saving PMs which helps to exploit them in less populated areas. We just need a better connection of infrastructure, transportation and industries. In the end, industries need to pay their share, ideally connected to urban centers. The absolut easiest way would be that later prodution PMs also need transportation to simulate the increased density they are offering with them. The transporation PMs for urban centers should work like the secondaries for other industries: With increasing levels, the amount of services is lowered, so city centers can finally get some more profits and we maybe come to the situation where the higher PMs are needed to produce enough services to match demand.

And later, the game could maybe track which states are connected to each other with rail. With better PMs for rail MAPI should fade away within rail networks, closing the gap to 100% but never eliminate it at all. This would lead to super connected markets in the lategame.

 

[47pik]

I don’t think any of this can realistically be solved until there’s some kind of logistics system added.

Hopefully soon. This year seems unrealistic, but, you never know. The lack of this system is arguably the single biggest issue remaining once trade is fixed in 1.9.

 

[IsaacCAT]

They are two different things for sure, but a rail without train is worthless and a train without rail is also worthless. Why was this split at all?

To simplify. If you have to calculate all the movement of goods the game will run to a halt.

But what about private industries that use vast amounts infrastructure but "pay" MAPI?

MAPI is not transport. Buildings do not pay MAPI. MAPI is there to simulate the local buy/orders influence to prices. It Is about information bias.

People thinking MAPI as a cost of transport has it all wrong. MAPI simulates state marketplaces were prices are set using buy/sell orders information that is weighted with the national market buy/sell orders and the state buy/sell orders. It is information access, not transport.

 

[yurcick]

People thinking MAPI as a cost of transport has it all wrong.

You are technically correct but overall wrong.

MAPI is not a cost by itself, it’s a factor to overall cost increase that can be best described as MA losses (not MAPI losses, this pair of words combined makes little sense).

MA losses are costs of transport, temporary local storage and finding the optimal buyer and seller. MAPI is (mostly) an informational factor, you’re correct in that. But when MAPI is low it simply means these logistics and managerial costs are high, it doesn’t introduce any “additional cost” out of thin air.

In other words, no, people do not have it wrong, the losses that manifest as the difference between seller’s price of sell and buyer’s price of purchase are first and foremost logistics. Which can scale with informational factors, nothing wrong with that.

UPD: Imagine that your task is to run across a maze. The calories spent when you do so are very physical, they’re what your muscles and brain burn while you do the job.
However, the exact length of your trip, and exact calorie spend, depend on whether you know where to run precisely or whether you have to explore the maze first. The worse you know the maze, and the slower you are to understand its patterns (which are purely informational factors), the more physical resource you will spend.
So, low MAPI is “not knowing the way to go”, and having to overspend on various physical things, mostly transport.

 

[IsaacCAT]

MA losses are costs of transport, temporary local storage and finding the optimal buyer and seller. MAPI is (mostly) an informational factor, you’re correct in that. But when MAPI is low it simply means these logistics and managerial costs are high, it doesn’t introduce any “additional cost” out of thin air.

You define Market Access (MA) components very well but when MAPI is low the solution is not to lay another rairload to lower transportation costs.

The nature of information makes it not a cost but something that needs to be treated completely different.

Businesses are making less profit due to MAPI but that does not mean that it is a cost, it lowers your sell price. It is on the revenue side of the equation.

I say that because people think of MAPI as a cost that can be reduced. For example, building more railroads.

If you have lower transportation costs in a state, that will not change the sell price in that state of your goods. The only way that lowering transport costs help you is by outcompeting the local competition and making them go out of business, thus increasing the sell price.

The only way you can affect MAPI is with information technologies or publicity. The later is not yet in the game.

 

[IsaacCAT]

Thanks to this thread, another suggestion for the game:

Thread 'Publicity for MAPI'
https://forum.paradoxplaza.com/forum/threads/publicity-for-mapi.1734347/

 

[TheFlemishDuck]

People thinking MAPI as a cost of transport has it all wrong. MAPI simulates state marketplaces were prices are set using buy/sell orders information that is weighted with the national market buy/sell orders and the state buy/sell orders. It is information access, not transport.

Not really. Observingly its not so because the consequences of lack of infrastructure or railroads leads to "more mapi", so it has to be kinda consistent then in it. A province that lacks infra to be at 80% market access just has +20 MAPI more. How do you make that as information access if the modifier of Mapi occurs/amplifies when there is a lack of railroad? I dont see how you necessarily could increase the penalty of information access indeed so significantly just due to lack of infrastructure/railroad, whereas a province at 50% infrastructure has extreme price differences with others but on the local level they can be self sufficient and then its neither an issue, it makes it rather logical that its about lack of railroad capacity as it were.