Logistics rework

[nobodys ghost]

I’m overall a big fan of Victoria 3 and have been since release. It’s steadily improving over time, albeit a bit slowly, but I think it’s only going to get even slower until we get a proper logistics system. The fact is this game is an economic simulator that doesn’t simulate the economy.

The movement of goods and people is one of, if not THE most important driving force of human history. If you want to make game about the economy you need to track where goods and people are from, where they’re going, AND, most importantly for this game, how they get there (and how long it takes). I honestly feel like logistics management should be the one of the pillars of gameplay. It ties into the main economic development loop, the trade system, the military system, the migration system, it should be a huge part of diplomacy. The two big canals should actually matter. There’s a reason humanity poured so much money into them. Railroads were a world changing invention and in Vic 3 they’re essentially reduced to a modifier.

I bring this up not just to complain, but because the sooner in development the dev team works this out the sooner they can start building on it. Reworking the military and trade before tackling this is a mistake because they’ll simply have to throw the old system out once logistics is implemented and that’s a lot of hard work gone to waste. Until then I feel like important concepts will have to continuously be abstracted to compensate, like MAPI. In the long run it will drag the game down with an esoteric system of modifiers that are ultimately disconnected from the core of the game: production and trade.

I think this should be priority number 1 going forward.

 

[TheFlemishDuck]

But what do you propose?

The matter of logistics has typically been regarded afaik as one that must be bypassed trough a abstraction as your more obvious implementations would likely hog too much resources, it's perceived as a performance problem right? A perfect system would be to be able to calculate the exact cost of a good going over land by rail or by river/canal in country A to a port or directly to the ocean to be traded to country B to be brought to its destination by rail or river or canal. What would even constitute the shortest and/or most effecient path if all options were there and prices fluctuate is hard to say, but eitherway thats a lot more calculations than the system currently makes.

My proposition has always been "keep MAPi, introduce a possibillety for mapi circumvention trough rail pm's for individual factories". Other propositions have been made, feel free to add yours.

 

[nobodys ghost]

The team has already said they want to work on logistics, my argument is more one of priority. I don’t want them to spend a bunch of time developing systems they’re ultimately going to throw away so I think it’s prudent to do this first, if they end up pursuing it, regardless of the form it takes.

Personally, I think performance should be compromised for this specific thing. It would turn it from a good game into an amazing one. I can’t really suggest a technical implementation without having the data the dev team has, and honestly I wouldn’t know what to do with it if I had it. The only helpful thing I could say is that PC seems to have found a way to work with large numbers like that by having the calculations treat large numbers of like units as a single unit, if I understand their system correctly. I don’t know if something like that would be enough, but regardless I think it’s worth a pretty hefty performance cost. I hate the abstractions like MAPI, they genuinely make the game harder for me play because the causal relationships aren't very obvious.

Maybe not everyone could run it, but that’s already the case, before anyone accuses me of being cold about it I’m playing the game on a steam deck as my only option, I’d likely be first on the block of player who get cut off. I’d rather the game be better and I wait until I can get a better rig to run it. Additionally, I’d think a significant number of players have updated their hardware since release. Maybe the situations different now.

 

[Prob32]

Agreed!

 

[TitaniumMan91]

Maybe not everyone could run it, but that’s already the case, before anyone accuses me of being cold about it I’m playing the game on a steam deck as my only option, I’d likely be first on the block of player who get cut off.

The only way I could keep playing if they increased the minimum specs is to fully overhaul my computer. What does it matter to me if it's good vs great if I can't even play it?

 

[alexgmcm]

MAPI and Infrastructure work to somewhat simulate peacetime logistics.

War logistics could do with a bit more work though. Apparently there is already a supply system for naval invasions that relies on having sufficient convoys being able to get to the invasion site etc. but it feels like it needs to be dialled up as I've never really noticed it in-game.

Ideally, there'd be a similar system for land units plus dialling up attrition so it matters more and the AI cares about it more.

 

[Helmut von Dandy]

The only helpful thing I could say is that PC seems to have found a way to work with large numbers like that by having the calculations treat large numbers of like units as a single unit, if I understand their system correctly.

This kind of simple suggestion is incredibly naive. It already works like this. You're not trading articles of clothing, you're trading an abstracted amount, and any abstraction they use has to be on a scale that supports quantities consumed by segments of the population and consumed by industry, and able to scale in a way that reflects the growth in productivity from 1836 to 1936. People are also already abstracted into pops by treating each combination of culture, religion, employer and job as a single unit.

It's not the numbers that matter, it's the number of things. For example, if the supply of late game goods increased from 10K to 100K that would probably have a much lower performance impact than having more types of goods in the game, because more goods means more buildings, more buy and sell orders, more trade routes, more taxes, etc. that need to be calculated.

 

[TitaniumMan91]

Ideally, there'd be a similar system for land units plus dialling up attrition so it matters more and the AI cares about it more.

I'm pretty sure there is, if my convoys go negative while at war overseas then my troops wither.

 

[alexgmcm]

I'm pretty sure there is, if my convoys go negative while at war overseas then my troops wither.

Yeah, sorry I meant for land units that aren't overseas.

 

[OxfordNik]

Completely agree that it seems odd that they seem to be re-working trade (external trade, at least as far as we've seen through published dev diaries to date) but aren't addressing the logistics system.

This is possibly tied to the belief that there shouldn't be a 'stockpile' of resources of any kind anywhere, as there was in e.g. Vicky 2 (whereas historically nations in this period did have stockpiles of e.g. weaponry, and logistics hubs were a big thing).

Seems to me that it would make sense to re-work logistics at the same time as trade/military, so that the three systems work in tandem (or four, if you split naval warfare and land warfare - which should be different, and imho deserve separate tabs, even).

If it were up to me, I'd design it something like this:

Logistics - related to (but in addition to) infrastructure/transportation, this would represent the process of moving goods from one location to another. Possible locations for goods would be: origin (where the good is harvested/processed/manufactured), transport hub (where the good arrives to be shipped to another state/destination), and destination (where the good is consumed by factories/pops/armies/naval units).
For simplicity, there wouldn't be a logistics cost to moving goods for processing between producers/consumers in the same state (which would instead be represented by infrastructure), but increasing e.g. port/railroad level in a state would provide additional logistics and hub capacity - potentially with additional production methods allowing for specialisation between boosting internal infrastructure and external logistics.

Trade - for external trade, using the reworked World Market to represent cross-ocean shipment of resources; but use logistics to represent 'land trade' (which seems a slightly under-baked concept at present, at least to me?). I'm encouraged by what I've read so far on this, particularly if large nations like Russia/US (or even Ottomans) can be split into multiple market regions despite being connected by land.

Army supply - I would, personally, like to see "manpower" being a 'good' that barracks/conscription centres produce, and armies consume (alongside the ability to build forts and/or supply stockpiles to be able to 'stock up' on supplies in certain areas where you know you are likely to face conflicts).
That way, you could also represent the logistical challenge of shipping lots of men to the front lines - including via sea lanes for overseas conflicts - since the army being away from its 'base area' would need to use logistics to actually get from A to B to begin with (a challenge for Russia during the Crimean War, to take one famous example), and then to re-supply reinforcements, if needed.
I sort of get why armies are assigned to strategic regions rather than states, though to me it would make more sense for a mobilized army to be physically present in a particular state (meaning it would be using that state's logistics links) - which could also impose some constraints on how many troops you could reasonably keep deployed (and supplied) in a particular area, something that is severely needed imho in colonial conflicts, in particular.
Incidentally, I would also remove the link between armies and navies for naval invasions, and instead use merchant marine to ship navies from anywhere to anywhere - but possibly have the option of assigning a navy to 'escort'/support a naval invasion. It doesn't really make sense that troops would board ironclads/destroyers/dreadnoughts etc to actually be shipped from A to B, since in reality troop transport was carried out using otherwise civilian ships, afaik?
(and while we're at it, can we please separate ammunition used by infantry - bullets - from ammunition - shells - used by artillery/navies?)

Naval rework - the devs have stated that they want to see navies become more 'tangible' (for want of a better phrase), and the way I would personally do it would be to rework the military shipyards/naval bases, along the lines of something like this:
- naval bases provide naval manpower (seamen) and naval logistics hubs for supplying/repairing navies (with different PMs used to represent the level of sophistication of e.g. drydocks to repair different ship sizes, and stockpiles for ammunition/guns/steel for repairs and supply).
- (military) shipyards provide the capacity to build/repair individual ships - though it may also make sense to combine 'standard' shipyards and military shipyards, with the shipyards simply providing merchant marine (which would potentially replace clippers/steamers) unless used to commission the construction/repair of warships
- navies would be roving hubs of ships which can be located in a state/region or in a naval node (potentially with sub-divisions of a fleet based in a particular strategic region being assigned to a state unless/until deployed, and supplied from that state while deployed, in a similar way to armies being supplied with manpower from a particular barracks - with both deployed navies and armies requiring logistics to supply them with ammunition/reinforcements while deployed).

This would hopefully make logistics both more realistic and more meaningful - while making it matter whether you're fighting a war in a home province or halfway around the world

 

[grommile]

If you want to make game about the economy you need to track where goods and people are from, where they’re going, AND, most importantly for this game, how they get there (and how long it takes).

Source (where from) and destination (where to) aren't too bad.

Routes (how they get there), on the other hand, are notoriously expensive to compute.

 

[Prob32]

Routes (how they get there), on the other hand, are notoriously expensive to compute.

Routes shouldn't actually change that often and can just sit as static table values after calculated - once you have your routes it's just a static integer which can then be multiplied by cost of transport

 

[grommile]

Routes shouldn't actually change that often

Whether that confident assertion is true is a genuinely interesting question, which requires you to define how you want routes to be selected, how often you want to audit the cost of an established route vs the cost of a prospective alternative route, and so forth.

 

[Prob32]

Whether that confident assertion is true is a genuinely interesting question, which requires you to define how you want routes to be selected, how often you want to audit the cost of an established route vs the cost of a prospective alternative route, and so

My day job involves data software products, there’s solutions out there for this type of calc. I’m not saying i think it would be easy but I don’t think it’s an insurmountable thing. My whole thing is just agreeing with OP that’d I’d like to see this feature set in the game prioritized over other feature sets

 

[IsaacCAT]

Posting here to follow this thread. I will try to elaborate my ideas for a logistic rework to discuss with the community.

 

[yurcick]

My proposition has always been "keep MAPi, introduce a possibillety for mapi circumvention trough rail pm's for individual factories". Other propositions have been made, feel free to add yours.

• Split the world into trade regions (about two dozen), track interregion trade (including within a single market) about the way the international trade works now, although with more automation (I'm yet to figure out 1.9 trade)
• keep MAPI system for intraregion interstate trade
• turn MAPI losses into shopkeeper profits, add transportation consumption scaled with those profits

This is essentially a computational compromise (as it would still be as expensive to move goods from Berlin to Dresden as it is to move them from Berlin to Cologne), but I'm sure we can't do without one.

 

[marmot01]

The two big canals should actually matter.

Also straits! There are five major straits in Europe (Bosphorus, Dardanelles, Gibraltar, Skagerrak, Suez), and a few more in the rest of the world.

I wrote recently in the trade rework DD:

Will the new trade system allow for interrupting a country's trade routes by blocking a strait that these routes pass through?

A major consideration for Russia for example was that most of its exports were going through the Bosphorus and Dardanelles, and it was ever under the threat of having its trade shut down by a sea power larger than its own Black Sea flotilla blockading the straits. This threat partly drove the whole school of thought within the Russian establishment, that Russia ought to conquer Istanbul for itself. Of course there was the whole "restoring Byzantium" thing too.

 

[za3tarani]

I’m overall a big fan of Victoria 3 and have been since release. It’s steadily improving over time, albeit a bit slowly, but I think it’s only going to get even slower until we get a proper logistics system. The fact is this game is an economic simulator that doesn’t simulate the economy.

The movement of goods and people is one of, if not THE most important driving force of human history. If you want to make game about the economy you need to track where goods and people are from, where they’re going, AND, most importantly for this game, how they get there (and how long it takes). I honestly feel like logistics management should be the one of the pillars of gameplay. It ties into the main economic development loop, the trade system, the military system, the migration system, it should be a huge part of diplomacy. The two big canals should actually matter. There’s a reason humanity poured so much money into them. Railroads were a world changing invention and in Vic 3 they’re essentially reduced to a modifier.

I bring this up not just to complain, but because the sooner in development the dev team works this out the sooner they can start building on it. Reworking the military and trade before tackling this is a mistake because they’ll simply have to throw the old system out once logistics is implemented and that’s a lot of hard work gone to waste. Until then I feel like important concepts will have to continuously be abstracted to compensate, like MAPI. In the long run it will drag the game down with an esoteric system of modifiers that are ultimately disconnected from the core of the game: production and trade.

I think this should be priority number 1 going forward.

so i agree, proper logistics system where movement of goods (preferably transport cost, distance etc is tracked), but how do you suggest to do this specifically? how do you imagine it to work without reworking the entire economy? as of now, there are no physical goods in the game, there are no actual quantities of goods. what you have (for each specific date) a snapshot of current balance of goods demand/supply which determines the price. how do you track the movement of goods if you dont even track the goods?

 

[AlexNereča]

We'll already have distance calculations for trade advantage (and maybe merchant marine consumption) with the new trade centres, although looking back at the dev diary and dev comments I didn't quite understand from the trade centre to where it will exactly calculate...

I'm almost tempted to say that you both need to simulate getting goods from the state to the market, as well as getting goods from the market to the state (abstracted by MAPI currently)

At least I see the number of state-to-state distance calculations we can do are gonna be limited by computational resources so I don't know how much of an option it is (imagine if the performance load you currently get when you enable the ownership/industry supply chain visualisation had to be dealt with every time a new industry in a different state entered the supply chain)

 

[Prob32]

so i agree, proper logistics system where movement of goods (preferably transport cost, distance etc is tracked), but how do you suggest to do this specifically? how do you imagine it to work without reworking the entire economy? as of now, there are no physical goods in the game, there are no actual quantities of goods. what you have (for each specific date) a snapshot of current balance of goods demand/supply which determines the price. how do you track the movement of goods if you dont even track the goods?

I think this is where there's a bit of a disconnect with some people, talking about hub's and spokes and different type of algorithms folks use for path finding (maximum flow extc). I want to see what the world market looks like in practice before updating my suggestions in my signature, I think alot of them are actually out of date somewhat - which is fine I enjoy writing up logistics proposals for the game.

My new view is that a better way to conceptualize the logistics question is to not think about goods physically getting loaded onto boats and trains but rather as an abstraction tool towards where we're measuring sell order impacts and buy order impacts.

Logistics is such a big yet important mechanic to solve because of it's impact on all parts of the game, it should encourage natural growth in places in Russia instead of seeing massive mines in Perm 10 years into the game and it should serve as a key pillar of the entire military system.