Smoke and Mirrors supply system (No step back)

[minimouse007]

As others have correctly pointed out, this is a system where bottlenecks and capacity matter, not the individual movement of any given quantity of supply. On a strategic level, tracking the production and consumption of supply from N producers (cities, per your example) would be chaotic, as well as computationally prohibitive.

Importance was placed where @podcat and I felt that it mattered - you as a commander should be concerned with taking key points, protecting supply lines, and not overburdening hubs - not managing hundreds of sub-networks and running a tycoon game in the background.

On the topic of supply hubs: these have been the source of some debate, internally. As has been noticed, the cost of these is unusually high. This is because they sidestep having to care about the factors mentioned above - namely, taking key points, protecting supply lines, and not overburdening hubs. I do not want the solution to a poor supply situation to be the construction of additional hubs - it is the 'nuclear option', and exists for dire circumstances, or in the creation of well supplied, fortified lines.

The intended gameplay loop here is to make the player reconsider their tactical approach to a situation: if you are struggling to supply your troops on a frontline, you can improve the railway connections to already present hubs, remove some supply-heavy units, aim to capture additional enemy hubs, or use air supply (yes, it looks like air supply may be in need of a bit of attention). If you can just build your way out of every situation, we didn't really succeed in making supply a strategic concern at all.

Overall, it seems you may want the system to represent the productive supply issues faced by the major powers in ww2. That is not unreasonable, but it is already partially covered by the equipment subsystem. It is also not what we set out to create with this supply system - here we built a system to represent logistical supply concerns with key interaction points for the player.

At the end of the day, if people are already considering this system to be more complex than before, I'm not sure what good it would have done to go even further and simulate the exact production, travel, and consumption or every piece of ammunition/supply.

it feels weird tho, that a seaside supply hub costs only 3k ic, 15% of a normal supply hub... it just runs under a weird name called "ports". also, i ve always used hundreds of transports to supply my barbarossa. so i guess, that will be easier now...

 

[Arheo]

it feels weird tho, that a seaside supply hub costs only 3k ic, 15% of a normal supply hub... it just runs under a weird name called "ports". also, i ve always used hundreds of transports to supply my barbarossa. so i guess, that will be easier now...

Yes, and it is likely the cost for these will increase. Not to regular supply hub levels, though.

 

[OllyOllyOxenFree]

As others have correctly pointed out, this is a system where bottlenecks and capacity matter, not the individual movement of any given quantity of supply. On a strategic level, tracking the production and consumption of supply from N producers (cities, per your example) would be chaotic, as well as computationally prohibitive.

Importance was placed where @podcat and I felt that it mattered - you as a commander should be concerned with taking key points, protecting supply lines, and not overburdening hubs - not managing hundreds of sub-networks and running a tycoon game in the background.

On the topic of supply hubs: these have been the source of some debate, internally. As has been noticed, the cost of these is unusually high. This is because they sidestep having to care about the factors mentioned above - namely, taking key points, protecting supply lines, and not overburdening hubs. I do not want the solution to a poor supply situation to be the construction of additional hubs - it is the 'nuclear option', and exists for dire circumstances, or in the creation of well supplied, fortified lines.

The intended gameplay loop here is to make the player reconsider their tactical approach to a situation: if you are struggling to supply your troops on a frontline, you can improve the railway connections to already present hubs, remove some supply-heavy units, aim to capture additional enemy hubs, or use air supply (yes, it looks like air supply may be in need of a bit of attention). If you can just build your way out of every situation, we didn't really succeed in making supply a strategic concern at all.

Overall, it seems you may want the system to represent the productive supply issues faced by the major powers in ww2. That is not unreasonable, but it is already partially covered by the equipment subsystem. It is also not what we set out to create with this supply system - here we built a system to represent logistical supply concerns with key interaction points for the player.

At the end of the day, if people are already considering this system to be more complex than before, I'm not sure what good it would have done to go even further and simulate the exact production, travel, and consumption or every piece of ammunition/supply.

if you were worried that people would be building their way out of trouble then why not add different levels of supply hubs so that they don't cover such a massive region?

 

[Bremen]

Yes, and it is likely the cost for these will increase. Not to regular supply hub levels, though.

This makes me really nervous. Building new/better ports is basically a requirement for certain nations like Japan.

 

[seldon]

I think part of this is psychological. Under the old system, being low on supply was a really bad thing. NSB normalizes it a lot. A unit being low-ish on supply while standing in their own territory doing nothing, has little to no effect, and is not unexpected. On the other hand, if supply drops below a certain threshold (indicated by the exclamation mark next to the red supply symbol on divisions), then you need to worry.

During offensive pushes, you will encounter significant supply concerns, especially as you say, in Russia/China etc. The game has been largely rebalanced around this - players and AI are expected to suffer from penalties and attrition for a while in these circumstances, until you have had the opportunity to address the situation.

Thanks that is important to understand. I noticed that my armor units had minus 16% attack malus, and big 63% breakthrough malus. The rather pathetic Republican troops only had a 2 or 3 attack factor, so they weren't killing a lot tanks. But was worried about attritions, since I was only seeing a 6% supply status

I think what would be helpful, for those of use to the old supply would be a table that did a comparison
Supply Status
75%
50%
25%
5%
And the list of negative effects in both the old and new system

 

[Mattyboy5]

As others have correctly pointed out, this is a system where bottlenecks and capacity matter, not the individual movement of any given quantity of supply. On a strategic level, tracking the production and consumption of supply from N producers (cities, per your example) would be chaotic, as well as computationally prohibitive.

Importance was placed where @podcat and I felt that it mattered - you as a commander should be concerned with taking key points, protecting supply lines, and not overburdening hubs - not managing hundreds of sub-networks and running a tycoon game in the background.

On the topic of supply hubs: these have been the source of some debate, internally. As has been noticed, the cost of these is unusually high. This is because they sidestep having to care about the factors mentioned above - namely, taking key points, protecting supply lines, and not overburdening hubs. I do not want the solution to a poor supply situation to be the construction of additional hubs - it is the 'nuclear option', and exists for dire circumstances, or in the creation of well supplied, fortified lines.

The intended gameplay loop here is to make the player reconsider their tactical approach to a situation: if you are struggling to supply your troops on a frontline, you can improve the railway connections to already present hubs, remove some supply-heavy units, aim to capture additional enemy hubs, or use air supply (yes, it looks like air supply may be in need of a bit of attention). If you can just build your way out of every situation, we didn't really succeed in making supply a strategic concern at all.

Overall, it seems you may want the system to represent the productive supply issues faced by the major powers in ww2. That is not unreasonable, but it is already partially covered by the equipment subsystem. It is also not what we set out to create with this supply system - here we built a system to represent logistical supply concerns with key interaction points for the player.

At the end of the day, if people are already considering this system to be more complex than before, I'm not sure what good it would have done to go even further and simulate the exact production, travel, and consumption or every piece of ammunition/supply.

"would be chaotic as well as computationally prohibitive" No it wouldn't, ive imagined it all entirely within my head over the course of the dev diaries and i expected it to be this way. I wish both you and podcat loaded up War in the East 2 for just 30 hours of gameplay so you could realize how simple and better it would be to use said supply system in HOI4. It would need some minor adjustments to make it compatible, because HOI4 is an arcade game in vanilla, but it would then allow modding because u added it as a base instead of an actually prohibitive system like you added now that is unmoddable to get the desired effect of an actual supply system with transported goods.

How it could be easily done, is make the states produce a supply amount per population and civilian factories (for an example, it could be more simplified) which then could be scaled based on different modifiers/province modifiers. This would be an actual good/mechanic, like FUEL which is currently produced in states and also already in the game as a base for you, and likewise would be needed in divisions/planes/ships.

Then you would make it be transported via rail to depots, each level of rail would have a limit of 10k supply, 20k, ect, whatever number you end up with, then depots would stockpile the supply over time(with a capacity of supply based on size of depot, eg smolensk is bigger then priprat marshes depot) as every week tick or month it increases in stockpile if the rail is functioning/repaired and connected to a output state.

Then your units within range of the depots(you could even use current depot system range), would then take x amount of supply from the depot, with the amount they take based on the division/plane/ships requirement (THE SAME AS FUEL). The farther a division is away from the depot the more attrition to trucks it could take(like wite2), and have a low supply penalty. Terrain could also play a part (like WITE2) on how efficiently said supply is transported from depot to division, but fundamentally all you need to do is copy fuel and make railroads transport it which stockpile it into depots that are connected to a working railroad, which have a RANGE to deliver it to divisions, instead of teleporting and producing it from thin air like the current model.

So seriously man, no it would not be impossible, yes it would be simple and inherently a better system in my opinion without more complication/choas. Infact if you layed it out like i described, it would be simpler than the system you have now, especially describing how it works to people and UI functionality. This is the end result of 8 months of development which is sad, but it's completely practical to update the system now that you've added mostly everything as you guys did similarly in patches years ago, and it would be a lot less work for you.

Anyway i dont know why i spent this much time writing this, but playing hoi4 from 2016 and having over 5000+ hours made me understand all the core mechanics thoroughly, and it's just depressing now seeing so much missed opportunity in the game, because i enjoyed it so much in multiplayer with hundreds of other people.

 

[Arheo]

"would be chaotic as well as computationally prohibitive" No it wouldn't, ive imagined it all entirely within my head over the course of the dev diaries and i expected it to be this way.

No idea ever survives contact with reality, no matter how cool it might sound in your head. I would love to be able to use this excuse with my TL next time we design something that causes slowdown though ;D.

It may be that I'm not understanding your vision clearly here, but I really don't see the value in simulating transported goods. In a game with such a wide strategic layer, it doesn't add anything to the player decision making - which is really how we should be designing features, not as a deep simulation first (despite how much I love simulation within grand strategy).

playing hoi4 from 2016 and having over 5000+ hours made me understand all the core mechanics thoroughly

I'm glad to hear it.

 

[Sbrubbles]

Seems like a simple solution would be to only allow one active supply hub per state no?

For example, a second supply hub could only buff the ability of the first to spread out incoming supply or serve as a fallback, but wouldn't increase total supply available.

 

[Bremen]

Seems like a simple solution would be to only allow one active supply hub per state no?

For example, a second supply hub could only buff the ability of the first to spread out incoming supply or serve as a fallback, but wouldn't increase total supply available.

Or limit the supply from hubs to the sum total of all incoming rail connections. I can think of several ways to abstract it that would require minimal processing burdens to get a similar effect.

As I see it there's several potential solutions being argued here, or at least several layers of abstraction:

1) As it currently works, where hubs produce supplies and simply require a connection back to the capital, and 5 hubs on the same rail line produce 5x the usable supplies. This obviously makes multiple hubs exploitable and leads to the current system where they are prohibitively expensive.
2) Hubs are limited at the local (state perhaps, or some other area since supplies no longer seem tied to states) level, so that, say, you could only have 1 hub in a region, or the hubs were maxed out in some way based on the sum of connections to the capital. This would be similar to the old system, but with the player able to exercise greater control thanks to the new rail system. Also this would mean hubs/ports wouldn't need to be prohibitively expensive to manipulate the supply system.
3) The capital has infinite supplies but limited throughput; IE if you have a level 5 rail line from the capital to the front, and no other connections, the entire front (no matter how many states) couldn't exceed the throughput of that rail line + local supply production. This is how I expected things to go from the dev blogs, and I can think of some tricks to do it with minimal processing burden, but it would be a lot of developer work.
4) A nation produces a limited amount of supplies and they get shuffled around by the rail network. I don't think this is completely unworkable, but I do agree it seems like it would risk overcomplicating things as well as adding a significant processor load.

Of those, I honestly greatly prefer 2 or 3, but I can live with 1 or 4. If the developers decide to stick with 1, it might be something moddable.

 

[Mattyboy5]

It may be that I'm not understanding your vision clearly here, but I really don't see the value in simulating transported goods. In a game with such a wide strategic layer, it doesn't add anything to the player decision making - which is really how we should be designing features, not as a deep simulation first (despite how much I love simulation within grand strategy).

How can you actually believe it couldn't add anything to the player decision making?? Here's just a few.
- Allow players to build/Increase output of supplies to further supply your or friendly armies, as a small nation like Albania cant supply 2 million men from its industry efficiently.
- Allow players to lend lease supplies, which is essentially a simplified food/ammo/essentials, which countries did historically in WW2 and would lead to much more player decision making.
- Choosing actual paths to advance(because theres actual bottlenecks in a transportation system, al alemein will never have same supply as London), railways and depots in Northern china are going to be far less supportive for your army then along the coast/center where they can TRANSPORT and stockpile more supplies.
- Choosing where to upgrade depots to stockpile more supply (shouldn't be necessary, as when a depot fills with supply it will go to the next one, but nonetheless an option if you believe you are going to need to supply a huge amount of divisions in the depots area or expect winter to be harsh, which should effect supply also)
- Actual logistical bombing, where bombing supply hubs destroys supplies & fuel, and player decision on where to bomb is more impactful, instead of just a wide area which reduces train stockpile (trucks are irrelevant).
- Upgrading ports would be a player decision, because they are a bottleneck in a transportation system, as currently they have no use other than a yes/no check for building depots which spawn supply.
- Stockpiling supplies in an area you think will struggle with supplies in the future (strategy), say convoy raiding or perhaps even being cut off, depots with supply stored in them could still supply units for some time.

Honestly at this point i dont even know why i'm writing this, if you said you can't do it because you tried it already, and it to heavily decreases performance, i'd understand. However saying you cant see how it would be overall better and add more player decisions and further strategic depth, i cant even be bothered to continue writing this, good luck as game director.

 

[Reman]

As others have correctly pointed out, this is a system where bottlenecks and capacity matter, not the individual movement of any given quantity of supply. On a strategic level, tracking the production and consumption of supply from N producers (cities, per your example) would be chaotic, as well as computationally prohibitive.

Importance was placed where @podcat and I felt that it mattered - you as a commander should be concerned with taking key points, protecting supply lines, and not overburdening hubs - not managing hundreds of sub-networks and running a tycoon game in the background.

On the topic of supply hubs: these have been the source of some debate, internally. As has been noticed, the cost of these is unusually high. This is because they sidestep having to care about the factors mentioned above - namely, taking key points, protecting supply lines, and not overburdening hubs. I do not want the solution to a poor supply situation to be the construction of additional hubs - it is the 'nuclear option', and exists for dire circumstances, or in the creation of well supplied, fortified lines.

The intended gameplay loop here is to make the player reconsider their tactical approach to a situation: if you are struggling to supply your troops on a frontline, you can improve the railway connections to already present hubs, remove some supply-heavy units, aim to capture additional enemy hubs, or use air supply (yes, it looks like air supply may be in need of a bit of attention). If you can just build your way out of every situation, we didn't really succeed in making supply a strategic concern at all.

I feel like the intended design choice didn't really translate to the implementation here. It seems like railroad levels and motorization were the intended ways to solve supply issues. They're presented well with tooltips and make intuitive sense, but their effects are completely overshadowed by supply dumps which have frustratingly esoteric mechanics. I tried to understand how supply dumps propagate supply to nearby provinces for around an hour, and I'm only more confused than when I started. For instance, why does Arkangelsk seem to propagate supplies 5x better than Oviedo in Spain, despite both of them being level 3 ports?

The entire system ends up being very anti-strategy, where the correct answer to supply issues is "hope there's a supply dump nearby you can conquer, and hope it works reasonably well". Supply dumps are way too expensive to build on the fly, and motorization/railroads mean little when supply propagation drops off a cliff even just a few tiles from the supply dump.

 

[Demonthese]

Not screenshot. If you were to block the eastern Med sea. The whole theatre will be unable to receive any supplies from London. Even if you construct ports which can be accessed from the red sea, and connect them to the rail network, the game will simply refuse to recognize it as a valid path to the capitol.

If I'm remembering right the game only considers the 10 closest sea routes so there are edge cases where even if you develop a land route from a port you won't get supply. The big one I remember not working was Vladivostok not being a valid lend lease path (or something like that Vladivostok had less logistics utility than other ports). So it's probably related to that.

 

[Arheo]

It seems like railroad levels and motorization were the intended ways to solve supply issues. They're presented well with tooltips and make intuitive sense, but their effects are completely overshadowed by supply dumps which have frustratingly esoteric mechanics.

Yeah, these were the intended primary methods to solve issues.

I tried to understand how supply dumps propagate supply to nearby provinces for around an hour, and I'm only more confused than when I started. For instance, why does Arkangelsk seem to propagate supplies 5x better than Oviedo in Spain, despite both of them being level 3 ports?

Supply dumps don't propagate supply, they propagate potential supply flow. I agree this did not become a very easy-to-understand system, but was necessary to combat the very binary cliffs that occurred without having a distance-from-hub element to supply distribution. We tried it both ways - without this, every war would becoming a tug of war over a specific point between two supply hubs.

The entire system ends up being very anti-strategy, where the correct answer to supply issues is "hope there's a supply dump nearby you can conquer, and hope it works reasonably well".

This is also partially the intention, though. I disagree that it is inherently anti-strategy however. Planning your invasions around where there are supply hubs feels very strategic to me - finding myself fighting an entire operation over securing a bridge was one of my highlights of alpha testing.

Supply dumps are way too expensive to build on the fly, and motorization/railroads mean little when supply propagation drops off a cliff even just a few tiles from the supply dump.

I was on the verge of removing the ability to construct supply hubs at all, at one point. Supply cliffs are hard to avoid, but I'll be taking a look at terrain effect on distribution soon.

Above all, I will reiterate that being under 100% supply satisfaction for a unit should not be considered a failure, unlike previous iterations of the game. This is something I want to work on telegraphing a little better.

 

[Bremen]

Supply dumps don't propagate supply, they propagate potential supply flow. I agree this did not become a very easy-to-understand system, but was necessary to combat the very binary cliffs that occurred without having a distance-from-hub element to supply distribution. We tried it both ways - without this, every war would becoming a tug of war over a specific point between two supply hubs.

I'm curious what you mean by this. The impression I got from the developer diaries was that hubs would directly supply nearby units - that if, say, I have a hub with 25 supply throughput, it has the potential to supply roughly 25 supply worth of nearby units (possibly with some form of diminishment based on the distance from the hub). But based on my playtesting, it does the opposite - if that hub has 25 supply throughput, it will scatter that supply to nearby provinces, regardless of whether it's needed or not. So if I put, say, 10 supply worth of units on the literal province with the 25 supply hub, there won't be enough supply for those units, because it's trying to supply a bunch of nearby provinces with nothing in them. But if I scatter those units around, so there's 2 supply worth in each of 5 nearby provinces, it seems to work fine, despite the fact that logically the same amount of supplies are coming down the rail line.

You use the word potential supply flow, which makes it sound like the hub should direct the supply to the units that need it, but this doesn't seem to be the case. Or maybe my game was bugged/I wasn't understanding it right.

 

[Jays298]

It sounds like the debate is between an organic supply network moving items from origin to place of need vs the use of these supply hubs.

But I think as an abstraction, the hub system is ok. If there was as war going on and I have a giant warehouse complex within a military base, I would move all the goods there preemptively.

To have a long "supply train" implies just in time logistics. Or great shortages. And I think that is abstracted with the trucks / motorization.

The system isn't perfect but it is miles better than what we had before.

Maybe a suggestion is to connect new logistics with shortages and surplus of various production items, old logistics.

Which basically means a supply hub isn't useful if you have nothing to stock it with and the factory / food producer is far from the front. The hub system makes sense if there is a surplus of equipment / supply.

 

[coffeelingfine]

HOI4 is doing more things overall than HOI3 does. This adds to those more things.

You must be a casual gamer since you didn't think about this possibility.

At this point in HOI4's life cycle I especially don't get the "casual" thing

If anything it's more complex than 3 now

 

[seldon]

Above all, I will reiterate that being under 100% supply satisfaction for a unit should not be considered a failure, unlike previous iterations of the game. This is something I want to work on telegraphing a little better.

My go-to source for information about this game and all Paradox games is the Wiki. Right now the Wiki says nothing about Railroads, supply hubs, horse vs trucks.
May I suggest you guys assign someone to update the wiki ASAP. I'd like to see an overview, a detailed explanation, and finally a strategy guide.
How to fix supply problems. Along with the malus comparison I previously asked for.

 

[LordWahu]

Fighting in North Africa, it's extremely difficult to maintain good supply. I was building a large railroad network from the port in Alexandria to the border, and the best I could do was 90% with about 11 garrison units and one cavalry. Looking at the issue, it appears the rails to the front don't actually do much without the supply depot. Despite putting Alexandria to extreme motorized, the hub couldn't reach

Maybe this is a silly suggestion, but maybe rails could enhance the range of a hub? Basically carting supplies to the end of the line and unloading for further transport after.

 

[Idle America]

Not gonna lie, I'm pretty disappointed that supplies still come from the capital. Because Washington D.C. is a famously industrial city, producing all the supplies of the US military.

 

[Arheo]

I'm curious what you mean by this. The impression I got from the developer diaries was that hubs would directly supply nearby units - that if, say, I have a hub with 25 supply throughput, it has the potential to supply roughly 25 supply worth of nearby units (possibly with some form of diminishment based on the distance from the hub). But based on my playtesting, it does the opposite - if that hub has 25 supply throughput, it will scatter that supply to nearby provinces, regardless of whether it's needed or not. So if I put, say, 10 supply worth of units on the literal province with the 25 supply hub, there won't be enough supply for those units, because it's trying to supply a bunch of nearby provinces with nothing in them. But if I scatter those units around, so there's 2 supply worth in each of 5 nearby provinces, it seems to work fine, despite the fact that logically the same amount of supplies are coming down the rail line.

You use the word potential supply flow, which makes it sound like the hub should direct the supply to the units that need it, but this doesn't seem to be the case. Or maybe my game was bugged/I wasn't understanding it right.

It is mostly semantics. The supply hub can be assumed to distribute supply, yeah.

Hubs can support a maximum draw based on the rail connection to them. Let's say 15.

They also distribute a maximum flow value to the provinces around them, which falls off with range (affected by terrain and other factors). This is to prevent cliffs, add some organic tactical gameplay, and avoid over-stacking. This number is the one you see in the province tooltips around a hub.

Once this capacity is used up, additional units in that province cannot receive supply (or the draw will be averaged), and no more supply will be drawn from the node.

To effectively overstack node capacity is actually quite hard:

If I stacked all those units in one of the provinces, most of them would start starving, and the hub just wouldn't be able to supply them all. Consider it a localised version of throughput.

My go-to source for information about this game all Paradox games is the Wiki. Right now the Wiki says noting about Railroads, supply hubs, horse vs trucks.
May I suggest you guys assign someone to update the wiki ASAP. I'd like to see an overview, a detailed explanation, and finally a strategy guide.
How to fix supply problems. Along with the malus comparison I previously asked for.

As a rule of thumb, don't worry too much unless you have divisions that have a red exclamation mark next to a red low-supply icon.

We don't usually update wikis ourselves (these are community curated), but I'll see if I can get some exact info put down somewhere.