Hearts of Iron IV - 33rd Development Diary - Supply

[Denkt]

However, removing fuel and supplies is a choice that I can only reasonably understand from Mursolini's suggestion that the aim was for something deeper, but time ran out so there was a need to shoehorn in something that would 'do the job'.

I don't think that is correct, they did delay the game and one reason could be that they did not find out any way to make HOI3 supply system superior to the supply system shown in this DD. The developers know what they wan't HOI4 to be, that was the first DD posted. They have actually played HOI4 as well as HOI3 and they know which supply model is superior. Atleast we should give this new system a chance.

 

[Heatth]

You can't import, give or buy supplies because there is no such thing as supplies. There is a measure of how 'in supply' you are based on your location, but you never build supplies, just equipment.

"Supplies" might not be stockpiled, but they are still pretty much a resource in the map. Look at the screenshots. Korea, for example, can provide 10 "supplies", and can import 50 more, although, currently, it is only using 1,4.

My question is, what if, instead of integral part of Japan, Korea was a puppet or even an independent nation? Could they import Japan's unused supplies? It would be very weird if they couldn't, but the DD didn't mention that.

 

[ozmono2005]

I see what you are saying, but I don't like it. Plain and simple, don't like it. It seems to me to be a solution in search of a problem. What is wrong with having "supplies" and "fuel", produced and transported separately? If you have read my posts you will know I also want separate "Lubricants" and "Spare Parts Panzer IV" and "Spare Parts MG40", etc. I love detail, and the more the merrier.

For me, and I see I am not alone, when you make things "simple" you break immersion for a lot of us. We want to worry about these things, we want to micro them. Things like fuel consumption are things the real world commanders, planners, and leaders had to worry about. I want to have to worry about them too. I want to have to make the hard decisions about who gets the scarce fuel. I'm sorry I won't get a chance to.

I don't want to micro them personally, especially not down to that detail, it would mean even more fiddling with production lines than at the worst points in the series. I do understand and share a degree of disappointment, even though what you want isn't shared by me. I don't think it's as bad as people are making out, the important thing for me is their is still a need for oil/fuel and that you can still cut off supply lines and create chokepoints. That said it's far from ideal. I raised concerns that this might be the case when they mentioned no money and no stockpiling of resources and it seems as though those concerns may have been somewhat valid. I think this and the way they deal with resources (and lack of money) are the weak points in the design although I don't think any of them are deal breakers.

 

[BarrosRodrigues]

"Supplies" might not be stockpiled, but they are still pretty much a resource in the map. Look at the screenshots. Korea, for example, can provide 10 "supplies", and can import 50 more, although, currently, it is only using 1,4.

I don´t think so; "supplies" seems to be just a "flag" to determine combat ability/effectiveness and also influence attrition.

 

[donagel]

One main thing you are missing is that there is now local supply production. In HOI3, no matter how big the pocket was, it was completely out of supply all the same, so players often tried to create huge pockets surrounding a large amount of area, so that everyone inside would die. In HOI4, since the supplies are produced by location, that doesn't work quite as well. Sure, it will cut some of supply, but if the region you surrounded is big enough to support itself, it will do so.
.

Sure, and maybe I am just assuming, but if you cut off a large pocket with a large force, then it will wither as the "local supply" will not support the numbers. If you cut off a large pocket with a small force, then it will nevert wither, but its a small force so you would just destroy it. Once again, positve thinking.

Also, don't hijack the thread, but I feel the need to say, thinking of the people of Paris tonight.

 

[Axe99]

They have actually played HOI4 as well as HOI3 and they know which supply model is superior. Atleast we should give this new system a chance.

The question is, superior for what? Unfortunately, even unmodded, HoI3's supply system provides a (substantially) superior abstraction of logistics in the 1936-48 period than does assuming the ratio of replenishment for equipment/parts, fuel and supplies is constant in all situations at all times. It's just maths (and pretty simple maths at that, we're not talking differential equations, integrals or polynomial regressions here), the supply system as it is, where there is only two forms of replenishment (equipment and manpower) instead of four (reinforcement, manpower, supplies, fuel) does not have the variables required to do an abstraction that will do an effective job. As I mentioned above, it'd make more sense to take out the tech tree, and the distinction between light/medium/heavy tanks, thank to make the abstractions that have been made in terms of supply, in the context of a grand strategy game of the second world war (and period).

I'll give it a chance, for sure, and I have no doubt I'll play it to death, but if the answer to my two questions (can me mod supples, fuels and stockpile limits in, and will the AI handle it) is yes, than I'll be modding it in first chance I get (well, if someone doesn't beat me to it first, in which case I'll help them get their mod balanced).

 

[WSnova]

Wait, I might be getting dumber with age but let me see if I understood the supply system:

-Supply is not tracked, its abstracted into provinces.
-So you have supply region Korea, lets assume its all plains with a base of 10 supply each with 10 regions, so the supply region Korea supports 100 base supply.
-VPs and ports give you a bonus supply,like Centers of trade in EUIV? SO lets say Korea has 1 VP an 1 level 10 port. Assuming the VP gives you 20 and the port gives you 30 then the new supply cap is 150 no?
-But you get a bonus if you have an interrupted connection with your capital? SO assuming no partisans and no subs (I guess convoys aren´t tracked anymore too?) lets say that gives me a modifier for +50%?
-So my new Cap would be 150*1.5=225 supplies.

So in theory, if I trap a Division in supply region Korea, cut it off with my Navy and take all regions except 1. Then the enemy combatant would only have 10 supply in total no? But assuming the divisions I trapped in Korea use less than 10 supplies they don´t get weakened since they are living off the land?

I mean, I would need to see this in action, but it seems like the EUIV system where as long as you keep your troop numbers low you can keep rolling all the moment you want no?

Or maybe I am understanding wrong, I need to see this working.

 

[RickInVA]

But all of those things ARE the same thing.

If your tanks don't have enough fuel you siphon what fuel you can into a few tanks and fight with just those.
If you are out of ammo you move what you can to a few tanks and fight with just those.
If your tanks break down / lose their treads / strip their gears then you what parts you can to repair the others and fight with just those.
If you are bombed flat and most of your tanks are destroyed then you gather what is left and you fight with just those.

In each case and every case the result of mishaps is the unit has to fight with less than a full compliment of tanks.

Hence the new supply system works exactly as intended.

However, if I am only out of fuel all my tanks are operational once I get enough fuel.
All my tanks are operational when I get more ammo, etc., etc.

Real life requires me to only replace the missing article, not everything that goes into producing the whole item.

If I need to replace the roof on my house I don't also replace the foundation, the plumbing, the windows, etc., I just replace the roof.

 

[Alexander Seil]

For me, and I see I am not alone, when you make things "simple" you break immersion for a lot of us. We want to worry about these things, we want to micro them. Things like fuel consumption are things the real world commanders, planners, and leaders had to worry about. I want to have to worry about them too. I want to have to make the hard decisions about who gets the scarce fuel. I'm sorry I won't get a chance to.

In a flow-based production system, making a separate fuel flow that needs to be converted from oil flow would mean that instawin is achieved by any gamey exploit that cuts off oil supplies for longer than 2 weeks. It then becomes a pretty stark choice, either make a whole warehouse management mini-game in the style of WitP to prevent stockpiling exploits, or drop the whole thing and make abstract "replacement chits" that represent everything. Naturally, a lot of people will argue in favor of the warehouse management mini-game, but no AI anyone can currently develop will be able to handle that.

However, if I am only out of fuel all my tanks are operational once I get enough fuel.
All my tanks are operational when I get more ammo, etc., etc.

Again, the hypotheticals. I'd be really curious to know of a case where a large unit had everything except the one thing that isn't directly simulated. Difficulties with getting fuel go hand in hand with difficulties in getting spare parts.

Real life requires me to only replace the missing article, not everything that goes into producing the whole item.

Real life also requires you to not know that with probability 0.75, a war will break out over Danzig in 3 years, 9 months.

 

[Acaios]

British India is a separate country, and I doubt that Italian Army in East Africa could supply themselves "off the land" with anything at all.

I can assure you that italian navy had their problem to send supplies in north africa. They never ever tried to force suez or gibraltar to send supplies in ethiopia.

All the armchair generals-cum-mathematicians on this forum agree. Wouldn't it be great if the game modeled a province-by-province supply network that received supply requests from units and then routed supplies optimally given location of depots and infrastructure, taking into account future requirements due to planned offensives and redeployments? And the AI knew how to optimally place depots to avoid getting blasted into dust by clever humans? And all of that ran on your 10 year old laptop at two weeks a second?

Fortunately for all of us, those silly amateurs at Paradox don't know how to make such a simple thing work, otherwise they wouldn't be making video games, but rather cornering the market in strategy and operations consulting.

It's not a difficult task to program a "find shortest available route". AI do it for your every time you move an army more than 2 provinces away.
Again, it's not a difficult task to program an AI capable of choosing the right spot for supplies center. Way easier than making it decide the right troops to produce. IRL army use a system of equation and some statistic to decide the distance between supply points and the ammount of the stockpiles. Simplified version of those equations are used even in civil transport (amazon) to choose the location of warehouses.

In the end, it's not a problem of mine if you can't afford to spend 500$ every 3 or 4 years to buy an up to date hardware. If you can't really afford that, you probably shouldn't be playing a game but looking for a better job instead. Software Houses should not be limited in their job just because you don't want to put away your commodore 64.

 

[Heatth]

I don´t think so; "supplies" seems to be just a "flag" to determine combat ability/effectiveness and also influence attrition.

I know. But you still need them. Again, in my example, is Korea doomed to have a small army just because it is a small country? It can import resources and factories, so why not supply?

Sure, and maybe I am just assuming, but if you cut off a large pocket with a large force, then it will wither as the "local supply" will not support the numbers. If you cut off a large pocket with a small force, then it will nevert wither, but its a small force so you would just destroy it. Once again, positve thinking

True but at last in large pockets you will still have some units on supply, unlike in HOI3, where everything was doomed. Furthermore, haven't you thought of the middle ground? Where the army just happen to be about the size the pocket can support? You can also disband some divisions so that they fit into the new restrains.

The new system didn't removed encirclements, but it made so they make a lot more sense. Large pockets may be weaker than if they had access to the full supply chain, but they are not completely hopeless. Also, as a side effect, the new system prevents the absurd exploit of encircling the capital and causing the whole country to go out of supply.

 

[BarrosRodrigues]

The question is, superior for what?

Maybe for Human understanding, even for those with a few feathers short of a whole duck?

In a flow-based production system, making a separate fuel flow that needs to be converted from oil flow would mean that instawin is achieved by any gamey exploit that cuts off oil supplies for longer than 2 weeks. It then becomes a pretty stark choice, either make a whole warehouse management mini-game in the style of WitP to prevent stockpiling exploits, or drop the whole thing and make abstract "replacement chits" that represent everything. Naturally, a lot of people will argue in favor of the warehouse management mini-game, but no AI anyone can currently develop will be able to handle that.
.

How about simply not allowing an encircled capital to cut off supply to the rest of the country? The moment the old capital is encircled the capital simply moves to another province. If the Black Ice team did it for HOI III the devs can do it for sure.

 

[Alexander Seil]

I can assure you that italian navy had their problem to send supplies in north africa. They never ever tried to force suez or gibraltar to send supplies in ethiopia.

And? They enter the war, Suez is closed, they lose East Africa.

It's not a difficult task to program a "find shortest available route". AI do it for your every time you move an army more than 2 provinces away.

Shortest available route to what? A "representational" supply system so many are clamoring for implicitly involves simulating a throughput-capped railway grid with finite rolling stock. All of this stock has to return to the rail yards over the course of every "cycle," meaning that this is similar to solving for optimal flow in an electrical grid. Now, this is all great, and approximately solvable with simplex, once it is determined which part of the grid is optimal to use...which makes it computationally prohibitive in a game, and puts AI at a tremendous disadvantage, as adding or changing any elements of the grid amounts to solving this problem for every combination of provinces where you could add or change something. Humans can do this with heuristics and experience, an AI has to brute force it all the way, and it's already a nightmare even if the frontline isn't moving (and there is no need to plan ahead, which makes it stupidly hard even for humans).

Again, it's not a difficult task to program an AI capable of choosing the right spot for supplies center. Way easier than making it decide the right troops to produce.

See above.

IRL army use a system of equation and some statistic to decide the distance between supply points and the ammount of the stockpiles. Simplified version of those equations are used even in civil transport (amazon) to choose the location of warehouses.

Real life institutions ultimately have humans making these calls, with the aid of some mathematical models. And Amazon can pretty much bank on static demand patterns, making it pretty simple to solve. It doesn't have a magic algorithm that predicts an optimal capacity expansion path for 10 years that they intend to (or should) follow because the math says so.

In the end, it's not a problem of mine if you can't afford to spend 500$ every 3 or 4 years to buy an up to date hardware. If you can't really afford that, you probably shouldn't be playing a game but looking for a better job instead. Software Houses should not be limited in their job just because you don't want to put away your commodore 64.

I think Paradox is interested in making sales to anyone who's willing to pay, not just to the PC master race or to people with access to institutional computing clusters.

EDIT: Just for lulz, search for "traveling salesman problem" in the Hearts of Iron 3 forum. You're not the first one to propose this brilliantly simple solution of "just add depots" for HoI3's supply problems.

 

[keynes2.0]

Then they drive to the front and engage the enemy, being knocked out within a few days. They only fill their petrol tanks once in their lifetime.

So they are sending crews out in the tanks without any training?

 

[Nicolas I]

If I can choose between a simple, working abstracted system and a broken, complex opaque system I know which I will choose.

That is both manicheism and a false dilemma fallacy. The presentation of a false choice often reflects a deliberate attempt to eliminate several options that may occupy the middle ground on an issue.

Most people here don't argue that the new HOI4 system is all that bad and don't say HOI3 system was sound. They argue that HOI4 could have been better with fuel/supplies included within the new supply system, as it appeared it the DD#6 under fuel use and supply use in the division planner.

EDIT: please read manichAeism. Thanks to Aries666 for correcting this small mistake.

 

[Heatth]

That is both manicheism and a false dilemma fallacy. The presentation of a false choice often reflects a deliberate attempt to eliminate several options that may occupy the middle ground on an issue.

Most people here don't argue that the new HOI4 system is all that bad and don't say HOI3 system was sound. They argue that HOI4 could have been better with fuel/supplies included within the new supply system, as it appeared it the DD#6 under fuel use and supply use in the division planner.

"Could be better" is a poor criticism if you don't specify how. People compare to what they are familiar with, which, in this case, is often HOI3.

Also, using DD#6 as an example is not of much use. The game was in pre-alpha back then. It is true there are fuel use statistics back then, but the whole thing was a very early work in progress. We don't know if the supply system of back them was even remotely functional. There is little point brining that up, the game change during development and that is simply an example of that. The devs clearly think the current system is better and we can't argue because we don't know how it was at the time.

(btw, "supply use" is still an statistic, as you can see in DD#30)

EDIT:Furthermore, Aries666 wasn't presenting a false choice. The person he was replying to was complaining about change. They were saying the HOI4 system was inferior to the previous system. It was a binary comparison, so Aries wasn't making a fallacy. At most he was oversimplifying the issue by dismissing the previous system as broken.

 

[Aries666]

That is both manicheism and a false dilemma fallacy. The presentation of a false choice often reflects a deliberate attempt to eliminate several options that may occupy the middle ground on an issue.

Most people here don't argue that the new HOI4 system is all that bad and don't say HOI3 system was sound. They argue that HOI4 could have been better with fuel/supplies included within the new supply system, as it appeared it the DD#6 under fuel use and supply use in the division planner.

When a person constructs an uneccessarily wordy response, incorrectly uses terms whilst also incorrectly spelling them I just laugh as that the only fitting response.

 

[Acaios]

And? They enter the war, Suez is closed, they lose East Africa.

they fought for over an year with supplies found in ethiopia.
Are you really trying to say that a system where the capital resupply all troops on the map is realistic? do you really think that the USPACFLT is supplied from washington dc?
It doesn't even exist a central supply depot since 1800, when napoleon invented the military logistic, as it would be a way too easy target for an attack.

Shortest available route to what? A "representational" supply system so many are clamoring for implicitly involves simulating a throughput-capped railway grid with finite rolling stock. All of this stock has to return to the rail yards over the course of every "cycle," meaning that this is similar to solving for optimal flow in an electrical grid. Now, this is all great, and approximately solvable with simplex, once it is determined which part of the grid is optimal to use...which makes it computationally prohibitive in a game, and puts AI at a tremendous disadvantage, as adding or changing any elements of the grid amounts to solving this problem for every combination of provinces where you could add or change something. Humans can do this with heuristics and experience, an AI has to brute force it all the way, and it's already a nightmare even if the frontline isn't moving (and there is no need to plan ahead, which makes it stupidly hard even for humans).

See above.

You're making thing more complex than they are.
army A have to choose between 3 depot. Depot1 is 500km away. Depot2 is 100km away, but supplies should pass an already "used" province. Depot3 is 150km away with no infrastructure usage between them. The army will now resupply from depot 3. The end.

Real life institutions ultimately have humans making these calls, with the aid of some mathematical models. And Amazon can pretty much bank on static demand patterns, making it pretty simple to solve. It doesn't have a magic algorithm that predicts an optimal capacity expansion path for 10 years that they intend to (or should) follow because the math says so.

Military logistic is a science, not a dice game.
Among all possible solution, there's one and only one that best suit the current situation. Math CAN fing it out and like all other sciences it DO allow us to predict the future expansion of a logistical network once given the right input, not only in the next 10 years, but in the next 1000 years with the only limitation that statistical data you have to use can change in a such long timeframe.

I think Paradox is interested in making sales to anyone who's willing to pay, not just to the PC master race or to people with access to institutional computing clusters.

actually, pentium IV with geforce 8 series and 2 gb og ram is not an "institutional computing cluster", it's more like a pretty average 2006 desktop pc.
HoIIV will cost more than the value of a pc like this.
don't even try such an argument btw.
I have a tape drive somewhere in my basement, so why not return to sell software on magnetic tapes? After all, a lot of people can still use them.
If you have such an obsolete pc and you pretend to find software houses programming for your hardware, you really should fill ashamed to slow down human progress.

EDIT: Just for lulz, search for "traveling salesman problem" in the Hearts of Iron 3 forum. You're not the first one to propose this brilliantly simple solution of "just add depots" for HoI3's supply problems.

ofc is not, since usually people who approach grand strategy game actually... like strategy! (don't u say)
Logistical warfare is one of the mayor point since the german unification war and their railroads. There's no strategy in clicking the "upgrade infrastructure" button and wait till the end of the production.

 

[Axe99]

When a person constructs an uneccessarily wordy response, incorrectly uses terms whilst also incorrectly spelling them I just laugh as that the only fitting response.

However, grammar and semantic issues aside, his underlying point was a very sound one. There are very easy ways to take the good things in the supply system as proposed (of which there are many) and alter a few other things to come up with a supply system that will present a simple but sensibly abstract system, compared with the simplified (but not abstracted, as at the macro level the proposed system is not a reasonable approximation of how supply, fuel or equipment replenishment worked). Of course, I'm sure the proposed system is a fun one, but it reminds me a bit of when I was studying first-year economics. When I stopped trying to make sense of it, and just tried to understand it as an internal system with little relationship to the real world, it was far easier to work with. I fear that the current system will be like that to many people familiar with WW2-era logistics - it will actually be more confusing than HoI3's system initially, because it works in a counter-intuitive fashion.

 

[Aries666]

However, grammar and semantic issues aside, his underlying point was a very sound one. There are very easy ways to take the good things in the supply system as proposed (of which there are many) and alter a few other things to come up with a supply system that will present a simple but sensibly abstract system, compared with the simplified (but not abstracted, as at the macro level the proposed system is not a reasonable approximation of how supply, fuel or equipment replenishment worked). Of course, I'm sure the proposed system is a fun one, but it reminds me a bit of when I was studying first-year economics. When I stopped trying to make sense of it, and just tried to understand it as an internal system with little relationship to the real world, it was far easier to work with. I fear that the current system will be like that to many people familiar with WW2-era logistics - it will actually be more confusing than HoI3's system initially, because it works in a counter-intuitive fashion.

It didn't make any claims to his point, just that I don't take kindly to someone attempting some intellectual one-upmanship and totally failing.

As to the discussion at hand. I can empathise with peoples desire to have a system that works both from a gameplay perspective and from a realism perspective. However, I would prefer something that works as a game as opposed to something that works in the inner imagings of my head. To that end HOI3 (our best previous reference point) was a system that failed in gameplay terms and was barely realistic anyway. In HOI4 we are presented with a more streamlined system that should enhance gameplay and will be slightly less realistic. For me this is perfectly acceptable because when I play a game I accept that it is exactly that and don't try to rationalise or approximate it to real life.