How is the vicky economy modelled? What are the formulas and underlying assumptions?

LeCare said:

I also disagree with the notion that the desire system is in principle bad. The game is modelling the 19th Century where prices were not fixed all around the world and strictly regulated and controlled. There was no forex, no EU commission and price discrimination was still common. You therefore have to see the world market as auction system, not as normal equilibrium market. Every consumer pays what he wants to pay (first degree price discrimination) and the consumers competing to buy goods therefore drive price up, even without buying anything. Again however the game overdoes it as practically everyone bids above ability.

Click to expand...

If this was the way the game I would agree with you. However, there is no possibility for price discrimination because the market operates under perfect competition theory without transportation costs and there is one world price. Secondly, the problem with what is being called desire is that it is based on perfectly inelastic demand (quantity demanded does not respond to price). Income plays no role in the process. If a pop wants X luxury cost and that is more than is available prices rise even if the pop did not have the income to buy X luxury goods when the luxury round began. The problem here with is called desire it is independent of income. I thought something similar at first until it was explained. Pops still "demand"
the good even if they are broke.

LeCare said:

I also think the shareholder system is not so bad. I mean, to a certain extent this makes sense as factories that earn more also generally pay higher wages (with 8.5 trillion speciality exceptions of course, as always in economics). The good thing about the shareholder system is that it is a cheap way of simulating wages with some realism without having to create a real labour market. That would be madness in the game in my opinion as the 100 trillion chinese craftsmen that emigrate ebverywhere endgame (again trhe ridicolous growth which probably would have made thomas malthus contemplate suicide) will reduce the worlds population into slavery.

Click to expand...

The problem with the shareholder system is that if the factory makes $0 then no workers have income because factories can often break even and stay in business this enhances the d=!d problem that is consumers bid on goods even when they have no income to afford them. So a factory breaking even means no one is getting paid but there wants for goods are still in demand because income is independent of demand in the model, the fatal flaw. PDM fixes this to an extent by setting no min wage to $1 so workers always get paid. A factory that is losing money still pays its workers in the real world. It does not in the game and this can cause further problems with demand. Even U/E workers with no income still bid on goods even though they have no money to pay for them.

LeCare said:

That money comes from gold is a bit simplistic but good enough for 19th century (see gold standard). Fractional reserve banking and fiat currency did exist, but far less than today and gold did still play a big role, as it still does now.

The world market I think is ok. You can speculate a bit too much (buy all of one good, especially oil as US/Russia or Rubber as UK, wait, congrats you own the world...) but especially with oil its not even so unrealistic. The problem again I think is mainly the fact that labour is basically the only variable factor of production in the game. One can therefore produce endless amounts of natural ressources provided one has enough population. That breaks every economy as it basically means aggregate supply (if we assume every human can produce more than it needs to survive) basically tends to infinity, collapsing prices and creating a socialist utopia with the sole problem that the currency doesnt collapse (as there is no deflation) so thereby creating a socialist utopia where people cant buy stuff (wehich of course is an excellent simulation of stalinist russia).

Click to expand...

During the time period even if a nation was not domestically on the gold standard they used gold to trade. So I don't have a problem with this either. I really do not have a big problem with little inflation. However, the problem comes from demand being independent on income. We have little or no inflation because income does not shift demand in any way, hence the d=!d problem. Honestly, I don't think it is unique to luxury goods it just appears most in luxury goods because they are the last round of trading. Incomes usually sufficent to buy life goods so it is everyday goods before the problem shows up, when people run out of income but still affecting price of the good.

Actually, I think labor being the only variable factor is ok. I am not sure if diminishing returns are applied but the caps on labor per RGO would keep what you describe from happening. Honestly, in the 19th century population was low enough to where there were a lot more natural resources than could be exploited by population so increasing population increasing output of natural resources to what seems infinity is not so far off, think the move west in the US. What you miss in the anaylsis is Say's law. The extra income provided from exploiting these resources creates more AD which means prices do not collapse but are more stablized and may not even drop. Look at the industrial revolution, still occuring in the game, AS expanded greatly and rapidly but prices did not drop because the new jobs meant increases in incomes and more aggregate demand.

We also have to take into account the lack of storage when looking at the demand thing - I don't think there's a shopkeeper in the world who, when presented with the choice of 'you can either sell that to me for less, or set it on fire and get nothing', would opt for the latter.

Furthermore, it demonstrably breaks the system - stuff in high demand, with medium supply and low 'actual bought' ends up with rising prices, pricing itself out of the market at a time when it's already barely turning a profit. The price floors and ceilings are NOT, as Patton sometimes assumes, the root of all our problems (but that's not his fault, he's a trained economist and has been taught to blame price controls for all evils ), as you can run an economy on these principals - they're out of fashion in present-day real-life, but have overall been the norm since we invented currency - but combined with the d=!d flaw it becomes deadly. But the d=!d thing is the major issue. Whenever modder experiments get close to eliminating it, the economy performs passably enough that you don't recognise the price controls.

Basically, we're just praying that the Devs clear it in the hardcode.

Please, PLEASE PARADOX, I dont often use caps but we need you to look into this! For the love of god, give us 1.3 patch and fix these economic issues in a game about Economics!!! By balancing the markets it could very well really help with rebel issues and a host of other things.

True the throwing the stuff into the sea is not a good idea but it is less of a problem than the d = !d issue. I could live with throwing stuff into the sea if the model works. Although I have suggested a way to fix this I am not sure the loop and multiple dry runs would take so much processing time, especially on older machines, to make it near impossible for a RTS. It would work in a turn based game. Price controls are just an evil. I was not seeing how bad the model really was earlier.

Since I've got a degree in the subject and a reasonable familiarity with the issue, I'll take a crack at writing a comprehensible, jargon free explanation of the problem.

Price controls have one of two possible and visible results. Either they are relevant, in which case goods will obviously hit either the floor or the ceiling, creating an obvious and easily fixed bug. Or they are irrelevant, and price floats freely between the two. If a good hits the floor price, surpluses will occur. If it hits the ceiling, shortages should occur.

If a good hits the ceiling and surpluses still occur then there is an additional factor at play. This is definitional; a ceiling is located at an arbitrary price point where all supply will be consumed without it being profitable to produce sufficient quantities to fulfill remaining demand. Since this is not the case with luxuries, an additional factor is at play. The demand =! desire (d=!d) theory provides an adequate explanation for this additional factor.

Demand is a function that describes the willingness of a group of people to consume a good given a set (or group) of prices. Desire is people's, well, desire to own an item should price not be a factor. For completeness sake, supply is a function that describes the willingness of a group of people to produce a good given a set of prices.

A major and empirically verified issue with Victoria 2's market for luxury goods, and only luxury goods, is the ability of pops to demand that a luxury item be supplied to them irrespective of their ability to actually pay for said item. In other words, whatever $ amount you put into the demand function for a good's price, the result is still that every pop wishes to consume the good at that price. Furthermore, they do this simultaneously with every good marked luxury in their consumption bundle (the set of goods that an economic actor wishes to, or is required to, consume). Factories then attempt to produce goods to fulfill the desires of every single pop without checking for their ability to pay. This results in rampant overproduction -- the pops that can afford their needs make their purchases, and the rest is left to rot as no one else can afford it. Additionally, factories are simply incapable of producing sufficient supply to even come close to desire. The game engine notices the shortfall in supply, assumes that prices are too low and causes prices to go up. Eventually even the richest pops are incapable of satisfying their demand, leaving the whole production unsold even as ever more factories open to produce more goods that will never sell.

For goods in Victoria marked Everyday or Life, generally supply and demand come together in a rough equilibrium. Because Victoria's market is perfectly competitive (or near enough as makes no difference to this discussion) this equilibrium is achieved with zero economic profit. Economic profit differs from accounting profit; a firm can generate zero economic profit while still generating an accounting profit. For our purposes, the simplest definition of economic profit is money that is left over after all costs of production are paid. In Victoria 2 craftsmen, clerks and capitalists are paid out of economic profit. Since economic profit trends to zero as the competitiveness of the economy increases, the salaries of these three pop types also trend to zero. As they make up a significant portion of the demand for luxury items -- especially in the unmodded game -- this problem tends to reinforce the d=!d problem.

Aeon221 said:

For goods in Victoria marked Everyday or Life, generally supply and demand come together in a rough equilibrium. Because Victoria's market is perfectly competitive (or near enough as makes no difference to this discussion) this equilibrium is achieved with zero economic profit. Economic profit differs from accounting profit; a firm can generate zero economic profit while still generating an accounting profit. For our purposes, the simplest definition of economic profit is money that is left over after all costs of production are paid. In Victoria 2 craftsmen, clerks and capitalists are paid out of economic profit. Since economic profit trends to zero as the competitiveness of the economy increases, the salaries of these three pop types also trend to zero. As they make up a significant portion of the demand for luxury items -- especially in the unmodded game -- this problem tends to reinforce the d=!d problem.

Click to expand...

An excellent and very clear layman-friendly explaination of the economic principals at work Just a couple of qualifiers:

The price floors/ceilings DO give us the capiacity to prevent the profit trending to zero, provided that d=!d is removed. We can make goods fundamentally profitable; that is, if prices are placed at a high enough value compared to their own inputs, we can ensure that they also turn some profits, provided all goods are selling. If a surplus exists, then the factories can be forced to fire excess workers, while avoiding the danger of profits (and therefore wages) trending to zero at 100% demand/supply parity.

This is why the price controls MUST exist in V2, as pricing is not worked out dynamically with reference to input costs. Therefore, input-to-output profit margins must be worked out beforehand and written into the base cost of the goods, so that they can float freely within the floor-to-ceiling zone without becoming more expensive to produce than to sell.


Secondly - d=!d also occurs from countries stockpiling. Military goods will often flip straight into d=!d on game start, as every country in the world begins building a smany military units as it can, simultaneously. Rizzmond has conclusively proven that it only happens on new purchases, which means that somehow countries are introducing 'desire' as well - something I didn't belive was possible, as the stockpile slider should surely attempt to purchase all goods it's set to buy in the trade screen simultaneously, and then simply take the money it needs to do so. I was wrong.

This also means that, should everyday goods be purchased for the stockpile, they also fall into d=!d traps; making it harder to measure and produce repeatable experiments.

I agree with you that Say's law and the very human concept of infinite demand in general will stop the scenario I predicted from occuring in real life (although Im a bit too keynesian to except say's law per se). But that isnt what I meant really. I also agree with you that labour being the only variable factor is quite ok for the time period, especially ideologically, as it was Marx's great time. That also wouldnt be such a problem seen purely from an economic viewpoint

I however am accustomed to see things more from the development economics perspective (we are our own breed, I know) and in this respect the mentioned things of infinite labour annoys me. I hate to have to argue with malthus here, but he also lived in the times so he is relevant here, and seen from that angle the whole thing becomes a bit wierd. The problem is that China (if player controlled) and India (if controlled by the UK) can simply diregard all of the things that averted the malthusian catastrophe in real life (technology/mechanisation) and allowed the industrial revolution in the first place (see lewis dual sector model which is actually based on that). They can simply cram RGO's with workers (there seems to be no dimnishing returns and the cap can be expanded ridicolously much in my opinion), let population explode and then build tons of industry in India/China, driving industry score up to insane levels. That is completely unrealistic even for the time period as it disregards one of the core principles of the game (that idea of uncivilized nations which actually makes so much sense as it creates a periphery) and it allows mad development on the basis of nothing.

I do agree with you that the idea people trying to buy luxury goods without income is ridicolous (there is a reason why its called luxury, it should be a superior good) but I would have to say again that raising prices without buying anything or without even having the money to do it is very well possible and even rational (see dollar auction in game theory) in an auction system as the world market back then kind of was, however this should only be true for necessary goods where demand really is inelastic, as you anyways stated.

But actually, now that you mentioned Say's law, I want to say something else. A lot of people voiced the opinion that the game should be modelled after "economic theories". That sounds like a flight sim that is modelled after physical theories. The problem is only that in economics there is nothing comparable to a physical law. If you want to model the game after economic theories then please first specify what you would like to see:

Classical, Neo-Classical, Marxist (quite close to that actually), Keynesian, Austrian, Chicago/Monetarist, Post-Keynesian etc...

Also in respect to trade what would you like to have:

Mercantilist, Classical/Comparative advantage, Dependency/World System, New Trade theory, etc...

And what model of development would you like:

Nationalist, Lewis Dual-Sector, Triple sector, five steps, Marxist/socialist/developmentalist, Fascist, Harod-Domar/linear growth and so on...

Right, let's once again remind everyone that this thread is here to discuss how the economy works in the game. It is NOT here to discuss real-life economic theory, which is completely inapplicable here due to processor constraints and fundamental design principals.

PLEASE, PLEASE, PLEASE let's not allow our real-world economics qualifications, anthropology degrees, history MAs and Applied Mathematics doctorates etc drag this discussion into an excessively technical discussion of the relative merits of competing real-world economic theories and models that are not only not present in the game, but totally irrelevant to how Vicky 2 attempts to model the economy.

Paradox are NOT going to go back to first principals and re-build the economy from scratch using game theory, Say's law, Walras' auctioneer, Friedmanite Monitarism, keynesian mixed-economics, OR ANY OTHER APPLIED MATHEMATICS/ECONOMICS.

We are looking at the minimum possible change required to create a working, self-propelled economic model, as that would be the most likely thing that could be done with a patch. Removing d=!d is enough, in itself, to provide that change. It may not create a model that corresponds to the ones that you learned in first-year economics, but that's because the economy in V2 is infinitely simpler and follows totally different base assumptions to real-world economics. It is not designed to be plausible economic model, it is designed to be a computer game - and one that does not require amasters degree in an economics-related subject to play.

Reel it in, guys, please

Naselus said:

An excellent and very clear layman-friendly explaination of the economic principals at work Just a couple of qualifiers:

The price floors/ceilings DO give us the capiacity to prevent the profit trending to zero, provided that d=!d is removed. We can make goods fundamentally profitable; that is, if prices are placed at a high enough value compared to their own inputs, we can ensure that they also turn some profits, provided all goods are selling. If a surplus exists, then the factories can be forced to fire excess workers, while avoiding the danger of profits (and therefore wages) trending to zero at 100% demand/supply parity.


This is why the price controls MUST exist in V2, as pricing is not worked out dynamically with reference to input costs. Therefore, input-to-output profit margins must be worked out beforehand and written into the base cost of the goods, so that they can float freely within the floor-to-ceiling zone without becoming more expensive to produce than to sell.

Click to expand...

That's fine, and entirely in line with the goals of price controls, but only comes in when things are working properly, which is slightly out of scope for an explanation of why things aren't working properly. In any event, p=mc can easily occur with current price control settings -- I've seen it happen in game! So including an explanation of how that worsens a bad cycle is in scope. That's why I included a section on how and why p=mc comes into play, and not the above element you explained.

Secondly - d=!d also occurs from countries stockpiling. Military goods will often flip straight into d=!d on game start, as every country in the world begins building a smany military units as it can, simultaneously. Rizzmond has conclusively proven that it only happens on new purchases, which means that somehow countries are introducing 'desire' as well - something I didn't belive was possible, as the stockpile slider should surely attempt to purchase all goods it's set to buy in the trade screen simultaneously, and then simply take the money it needs to do so. I was wrong.

This also means that, should everyday goods be purchased for the stockpile, they also fall into d=!d traps; making it harder to measure and produce repeatable experiments.

Click to expand...

This I'll include. I neglected to, even though I've fallen prey to it myself! It's definitely something that impacts the game, as the AI makes significant use of it for whatever reason.

Just to repeat, I want to have a simple and easily understood explanation for why the current system fails, not how the current system works when everything is running smoothly. Much of this thread is wasted repeated exactly the same thing ad nauseum to players who just don't seem to have the basic failures in mind. I'm hoping having a baseline explanation that can be pointed to will resolve the player confusion issue.

edit: With regard to real world economic modeling, it's time consuming and processor intensive to develop a system that accurately reflects current understanding of how actors operate in a complex system. I can tell you that from first hand modeling experience on numerous programs. The results are obviously superior, but the current system creates a reasonable approximation without excessive processor use. Plus, it's way prettier and far more user friendly than the console or limited gui versions I've created! That's good enough.

Aeon221 said:

T
Just to repeat, I want to have a simple and easily understood explanation for why the current system fails, not how the current system works when everything is running smoothly. Much of this thread is wasted repeated exactly the same thing ad nauseum to players who just don't seem to have the basic failures in mind. I'm hoping having a baseline explanation that can be pointed to will resolve the player confusion issue.

edit: With regard to real world economic modeling, it's time consuming and processor intensive to develop a system that accurately reflects current understanding of how actors operate in a complex system. I can tell you that from first hand modeling experience on numerous programs. The results are obviously superior, but the current system creates a reasonable approximation without excessive processor use. Plus, it's way prettier and far more user friendly than the console or limited gui versions I've created! That's good enough.

Click to expand...

Indeed, and you did an excellent job of doing so; the ironic thing with this thread is that too many people are severely overthinking the issues at hand - Paradox games generally appeal to the highly intelligent, highly-qualified end of the game market, which means when paradox release an economics-based game like V2, they get a lot of economics graduates taking an interest - and then trying to apply high-end economic theories to figure out flaws which are actually based on low-end programming errors. This is most clearly illustrated by Patton's trouble understanding how d=!d could even happen - he's so familiar with basic economic theory that he couldn't comprehend how such a (from his perspecgive) obvious flaw could exist, so we went through a whole page of posts convincing him that it really was this simple.

Unfortunately, this does usually mean that the high-end economics grads begin to provide answers that are, to the layman, an utterly inpenetrable mass of equations using things they are totally unfamiliar with - MCM', QS>QD = -p, w = (O-I)/4 etc. This threatens to dump a useful thread into a terrifyingly technical, specific and - above all - irrelevant discussion that relies on the kind of familiarity with economic theory which 90% of people will not have - including me, and I'm an Economic Anthropologist.

Your post and your aim is fine, my 'reel it in plix' post is mostly aimed at preventing a lengthy disgression into debating which economic theories would be considered most appropriate to model in a theoretical and never-to-be-created ideal V2 world. Let's keep this thread on-topic and about the following:

1. What is ACTUALLY happening in V2.
2. What SHOULD be happening, in V2, within realistic bounds of what is plausible without re-writing the whole system.

90% of the elements of real-world economics do not apply here, so there is little point in trying to analyse the model through real-world economic theory. So please, let's keep it out of the discussion wherever possible and just look at the elements we DO have to work with - you're trained economists, so use that knowledge to analyse the data in front of you and ignore your training in what SHOULD be happening, as many of the base assumptions you automatically adopt are, quite simply, wrong for the model in question

I think if we are going to ask for any real world theories to be applied they should be applied to the micro level only. I certainly would not want to write a code where factories and pops worked in one way when it was a single unit but worked quite differently in the aggregate. To implement something Keynes would mean coding how pops save and invest then allowing for a random shock component. The more complex you make things the more likely they will be harder to get to work right. Being a Chicago school kind of guy I suppose getting the micro right makes the macro right for me since there is a belief the macro can simply be aggregated from the micro.

I think the game fits growth theory well, my specialization back in graduate school my work now is really mostly labor. The problem is growth is not = to development. The models work well for industrialized countries but the game lacks the inefficiencies that causes a difference in developing countries. For instance China is usually the bank loaning everyone money. Lack of well working wide spread banking systems was one of the reasons China lagged. So you have to put those inefficiencies in an artificial way like the unciv penalty. From a growth perspective I like the fact capital, human capital, and technology contribute to growth. They have the factors shown to be determinates from Solow to Romer. It does seem taxes for the uncivs do allow a lot of money for the govts that can be easily put into industry when civilized due to the amounts of resources. Perhaps there should be an RGO efficiency penalty or a strong one for uncivs. Maybe the benefits of free trade should come later than a level 1 tech? That might slow down industrialization for everyone which may not be a bad idea.

I think there are some misconceptions that flawed the micro model. They can be fixed but it is causing problems. If we apply real world theory it should be to fix the micro. If the micro is fixed and income relates to demand then the macro will be fixed. Inflation due to money will work. I don't really think we want to model shocks that cause recessions. In a simple model this is going to be hard to do without some sort of event that destroys capital or something like that.

I would prefer a simple game that might omit things like the importance of banking on development or the importance of the order of liberal reforms you make that works then a sim that tries to add all the features and bogs down. The more complicated the model the harder it is to code and the harder it is to make good AI. I think the bones for the game are good in what it details. If they can just get the demand issue correct and maybe allow for stockpiles in industry then it could work.

An alternative idea I think could work would be to reduce prices for goods bought from government stockpiles or allow governments to give them away for free. It would strain budgets but AI and humans could allow people to buy from the govt. stockpile which would make governments buy the goods pop do not have money for. No theoretical reason for this and theory wise it sucks but it could fix the problem somewhat easily unless we see the same thing happen that did with military goods.

Without seeing the Capi AI files (which are still hidden in the hardcode), it's basically idle speculation whether this is codable within the V2 engine; I suspect that the actual decisions capis make under the current system are basically:

1. Factory makes good that is in high demand. Expand factory.
2. Factory losing money, fire workers.
3. Factory has no workers, loses money (due to maintainence costs), shut factory down.

Now this *should* be fine, provided factories don't get d=!d hits, and base prices have been properly calibrated to ensure that, at base costs, all factories are profitable. That way, when demand = supply, profits will exist and be divided amongst the workers, and when demand > supply, profit should fall - I've found the roughly-ideal profit margin should be 25% when the factory is working properly and prices are all at base.

However, setting Capis to think of 10% profit as 'zero point' and stop expanding when their profits fall to there would be a good addition.

Naselus said:

Right, let's once again remind everyone that this thread is here to discuss how the economy works in the game. It is NOT here to discuss real-life economic theory, which is completely inapplicable here due to processor constraints and fundamental design principals.

PLEASE, PLEASE, PLEASE let's not allow our real-world economics qualifications, anthropology degrees, history MAs and Applied Mathematics doctorates etc drag this discussion into an excessively technical discussion of the relative merits of competing real-world economic theories and models that are not only not present in the game, but totally irrelevant to how Vicky 2 attempts to model the economy.

Paradox are NOT going to go back to first principals and re-build the economy from scratch using game theory, Say's law, Walras' auctioneer, Friedmanite Monitarism, keynesian mixed-economics, OR ANY OTHER APPLIED MATHEMATICS/ECONOMICS.

We are looking at the minimum possible change required to create a working, self-propelled economic model, as that would be the most likely thing that could be done with a patch. Removing d=!d is enough, in itself, to provide that change. It may not create a model that corresponds to the ones that you learned in first-year economics, but that's because the economy in V2 is infinitely simpler and follows totally different base assumptions to real-world economics. It is not designed to be plausible economic model, it is designed to be a computer game - and one that does not require amasters degree in an economics-related subject to play.

Reel it in, guys, please

Click to expand...

Sorry to be picky, but the last part of my post states this what you said exactly. I was referring to a lot of people wanting the game being modelled after real-life laws (see start of thread) and I was trying to state that I think this is impossible. Sorry, for the missunderstanding though.

I'm wondering as kind of arbitrary solution, since I figure that this d=!d thing is hard to kill as non-proffesional coder, wouldn't it just be possible for the time being to code in a fixed price for luxury goods (something ridicolously high) at which the price simply is reset? I mean, it is a ridicolous solution and in every way wrong but I figure its easier to code or?

Also I just think, that's just how we work, that the way the uncivilized nations auto civilize themselves is a bit wierd. I dont think it'll need economic theories, just a bit of expansion of the unciv tag, which is one of the things the game actually simulates quite well with the exception that they shouldn't loan out money or be transformed in Britain's economic powerhouse in just 15 yrs. You just cant tell me that Britain sustaining 30,000+ industry points because of endless indian factories isnt somewhat aso a major flaw of the game.

Naselus said:

Secondly - d=!d also occurs from countries stockpiling. Military goods will often flip straight into d=!d on game start, as every country in the world begins building a smany military units as it can, simultaneously. Rizzmond has conclusively proven that it only happens on new purchases, which means that somehow countries are introducing 'desire' as well - something I didn't belive was possible, as the stockpile slider should surely attempt to purchase all goods it's set to buy in the trade screen simultaneously, and then simply take the money it needs to do so. I was wrong.

This also means that, should everyday goods be purchased for the stockpile, they also fall into d=!d traps; making it harder to measure and produce repeatable experiments.

Click to expand...

I'm still not convinced we can just pin this on AI stockpiling, we don't really have conclusive proof. When build costs for units are removed we lose d=!d for military goods, but that doesn't mean countries aren't still buying military goods for their supply costs. It is bizarre why build costs are affected but not supply costs, and imo I think it is beyond us to speculate why.

Rizzmond said:

I'm still not convinced we can just pin this on AI stockpiling, we don't really have conclusive proof. When build costs for units are removed we lose d=!d for military goods, but that doesn't mean countries aren't still buying military goods for their supply costs. It is bizarre why build costs are affected but not supply costs, and imo I think it is beyond us to speculate why.

Click to expand...

I'm not suggesting it's only down to stockpiles, rizz I'm pointing out that at least some d=!d is definitely stockpile-based. Whether or not any is coming from elsewhere as well, I don't know; the magnification testing we did the other night certainly suggests that the rounding error for small POPs is involved in it as well, while the luxury needs/everyday needs good swap shows that luxuries are behaving differently from everydays.

However, the fact that military goods can d=!d when not in POP needs means that any item that gets bought by the stockpile - which is theoretically anything at all - is potentially vunerable to it from that aspect as well, so it must be noted.

As for why build costs are effected but unit-supply costs aren't - that's actually fairly obvious. Clearly, when a country purchases goods for construction, all demand from the construction order is added to total world demand. But only the amount being purchased in the trade screen is actually bought each day. Here's a test for you:

Start a game as Britain. You'll be top of prestige, and also near-self sufficient anyway. Build, well, anything you like; a fort is usually a good test subject for this since they're fairly expensive. Now, at least one of these goods should be in plentiful supply - it's lumber in PDM, but cement is probably fairly easy for the British government to get hold of too. Then let one single day go by.

Look at the progress made toward completion of the fort.

In my test, the AI needs to buy 348 lumber to build the fort with, and there's about 3000 spare on the market this early as the artisans haven't sorted themselves out yet. But rather than try and buy all 348 in one go, the AI is instead attempting to buy the lumber 55 at a time - presumably to avoid a price spike. Yet because demand is not being added 55 at a time, but instead all at once, I'm bidding on 348 lumber and only putting up the money for 55...

Every country in the world is doing this exact same thing at the start of the game. So the 'actual bought' includes all of those countries bidding on goods they're not going to buy for the next seven days, in addition to all the capitalists buying things properly and all the normal POPs buying needs.

I'd say that's a pretty firm bet on what's going on with this.

I'd like to bring up another fundamental aspect of the model that I suspect is intentional, but flawed: how price in V2 actually responds to supply and demand. Since I don't have a background in econ, please let me know if I'm off the mark here.

As I understand V2, a good's price is determined by the ratio of supply and demand. If the ratio D/S is in a certain range then the price rises and falls at about 1% of base price until it hits a price that corresponds to that range. This leads to situations where the price may start to rise in response to rising demand, because that boosts the D/S ratio, even if demand is much lower than supply.

It does NOT work the way I understand classical microeconomics says it should, that price rises when there's more demand than supply, and falls when supply exceeds demand.

I suspect that PI made it this way to prevent goods prices from pinging their floors and ceilings (or to rise infinitely or fall to 0 if there were no limits)the moment supply and demand are out of whack.

I could easily suggest a better way if my understanding of game economics and real economics is correct.

Dear god....

Please, don't make the game any more complex than it needs to be. True economic modeling would completely sap the interest of a large group of people in exchange for drawing a ton of interest from a handful of people. I majored in business for a year, it was boring... I'd prefer not to play a game based around that.

Caban said:

Dear god....

Please, don't make the game any more complex than it needs to be. True economic modeling would completely sap the interest of a large group of people in exchange for drawing a ton of interest from a handful of people. I majored in business for a year, it was boring... I'd prefer not to play a game based around that.

Click to expand...

I think most people who are concerned with the details of the economic model are really concerned with making it work properly, not making the game more complex than it needs to be.