Issue: Nations are Broke

[Xylephony]

Can't really find a more eloquent way to put it: nations are broke, and it's really damaging colonization in particular. It's normal for some nations to struggle and even for GPs to go bankrupt and collapse if they're hammered with enough wars, but this isn't like that. I've done two test games, one before the HPM farming efficiency reversion and one after, and in both even most of the GPs are either heavily in debt or just barely floating above the debt line, even when they haven't been the target of major wars. They are so indebted, in fact, that France and Japan only have some 15 ships each, which of course completely cripples their colonial points but also wrecks their ability to project power abroad. Japan has never taken Korea in a single EEM game I've done, and France both in my current test and in my previous test colonized most of Africa only to have to release large portions as primitives again because they were over their colonial points. That in turn fucks the lategame economy, especially as regards cotton and rubber.

I'm not arguing that any of the above shouldn't be possible, but it definitely shouldn't be the norm. Some change that was made at some point is causing serious economic problems when it comes to AI treasuries. I'm not sure if you reduced base tax efficiency so you could add more tax efficiency on the backend for the new tech levels, but if you did that would be my immediate assumption. The AI seems to rarely put much emphasis into commerce tech and rushes the economically punishing military techs, and that could cause them to have no easy way to fund themselves.

Here is a save I made a few decades back in my current game just to give you an idea of the scope of the problem. Tag to a given nation, even a GP, and unless they're a powerhouse like Britain, Germany or the US they're probably in debt with a miniscule navy: Save for EEM Navy Info.zip

[ujtcelvn]

Yeah, I noticed that the AI wasn't building navies for some reason too. Can you test a game after deleting events/Liquidity_Fix.txt? I've been really busy lately and don't have much time to playtest. Maybe the entire tax refund / liquidity reddit came up with was a mistake and the AI doesn't actually pay off debts unless their treasury is huge.

[Xylephony]

Can you test a game after deleting events/Liquidity_Fix.txt?

Sure, I'll start a new game sometime in the next couple days and see if it helps.

[Xylephony]

I played a test through to 1880, and the results are not great.

I'm still getting the impression that something more fundamental is wrong: Bremen went fully bankrupt within 4 months of game start, and I restarted the game entirely and they went bankrupt 15 months later in that game, and Lubeck just a year later, neither of which having ever been at war. Several other cases of completely spontaneous bankruptcy (not from the temporary reduction of income during savegame loads; I minimized the number of times I loaded for this test) were also evident. Even though those were all from minor nations, it seems to be indicative of a trend where the AI is just not getting the money it needs to support itself, OR it's being forced to spend more than it's used to from the start. I don't know if EEM made significant changes to required military spending, but that could explain part of what I'm seeing.

Now the good news is that, through to 1856, most navies were of a relatively normal size. But wars between the GPs in the '50s led to most GPs and future GPs taking on debt, and then the downward spiral set in. By 1870, France was down to 14 ships (slight debt); the Ottomans, 5 (heavy debt); the Russians, 0 (heavy debt, with a significant ongoing socialist rebellion); the Prussians, 14 (moderate debt, also a socialist rebellion); and Japan, 18 (moderate debt). Some nations, like the US and UK, have normally-sized navies despite being engaged with protracted wars, but many nations who earlier had larger navies had by now entirely scrapped them and are indebted.

By 1880, the only European nations with more than 25 ships are Britain and the Netherlands, and most nations, including GPs, are at least moderately indebted. Interestingly, that number includes Britain, who is 200,000 pounds in debt but has ten million pounds in the treasury and a massive positive daily balance. It seems like an AI might only pay off debt at all if that debt is actually having a significant impact on its income, but that's just a guess. Either way it doesn't help resolve this situation at all, because it seems like most of the time by the time a GP goes into debt it reaches a spiral point where it isn't going to get enough funds after that to be able to dig its way back out.

Here's an 1881 save for you to review: EEM 1880.zip

I've also just noticed that industry is simply not as profitable as I remember it being from HPM standard. Bottling works are continually unprofitable, for example, even after the farming efficiency reversion. I'm constantly forced to subsidize them on state capitalism/interventionism even if I only have a single one and beverages are in demand on the market, whereas in HPM they would typically be marginally profitable even in a low-tech nation with planned economy. If that's indicative of a trend, poorer-performing industry could be impacting the world economy more generally if nations are being forced to subsidize industry that is, for whatever reason, less efficient than normal.

[ujtcelvn]

I am certain I did not change productivity of most factories, pop demand for wine, tax efficiency, and navy supply costs. Maybe it's something I did to the defines, I'll reset a bunch of them to HPM values.

[Xylephony]

Maybe it's something I did to the defines, I'll reset a bunch of them to HPM values.

Alright, I'll do a new test with these changes at some point this week and see if that improves it at all. I'll post saves when I have them.

[Xylephony]

New test is in 1884 right now, and if anything it's worse with the defines changes. The Ottomans and Prussia have a bigger navy and that's about it. Most nations are still tremendously indebted. I watched Japan industrialize and then completely de-industrialize down to 0 industry score and just 10 boats, and the same thing happened to Spain, Brazil, Bolivia, and New Zealand, the latter of which is a highly-populous puppet of mine with a major precious goods RGO which should be keeping its economy very stable, but it's not. Nations have still spontaneously gone bankrupt without game load or any wars taking place as well (Lubeck again, as well as Mecklenburg).

Japan in particular seems to focus its research heavily in army and navy and have virtually no commerce techs, which might be why it's having such a problem despite its RGOs. But that can't explain the rest of the world, and the deep lack of industrial profit. Many nations which are in debt, like Austria and Russia, are so because of extremely expensive industrial subsidies, with, again, beverages representing the single greatest subsidizing cost. I can't stress again how weird that is for me to see, because they were the stable good to industrialize off of.

Long story short I feel like industry being underprofitable is still part of the equation here, but that's not the whole story or those nations which fully de-industrialized like Japan would've stabilized, and they haven't. Nations both aren't getting enough funds to support themselves normally and industry is also weak and thus not providing an out to generate a middle-class and higher tax profits.

[ujtcelvn]

I reset everything I can think of to HPM values. There is a rumor that FACTORY_PURCHASE_MIN_FACTOR is actually bugged so I set it to 1 which is rumored to fix the bug, but it was 0.5 in HPM so it's back to 0.5 now. This value is how much a factory will throttle down production if it is missing input goods before it starts firing workers, and the rumor is that there is a bug that makes the factory never return increase production back to normal if the input goods are available, FACTORY_PURCHASE_DRAWDOWN_FACTOR doesn't have a comma after it for some reason, even in vanilla Prices of all goods and production of all goods (except silk and synthetic goods) have been reset to HPM levels And all HPM techs and inventions are the same until 1900, and any inventions I added before 1900 only increase machine parts, coal, iron, sulphur, fertilizer, fuel, rubber, glass, fabric, and cement. Boy I sure hope this fixes it, because if it doesn't, I'm out of ideas.

[ujtcelvn]

Well at least liquor doesn't seem to be unprofitable anymore... Can you open this save file and compare it to how large navies and how much cash reserves the AI is supposed to have by 1885? 49.zip

I think it's working, in older versions, past 1936, the number 1 GP gets like 3x the industrial score of the number 2 GP and the prices of most industrial goods turn red meaning there is crazy amounts of overproduction and deindustrialization. Now, the distribution of industry and the prices of goods are a lot more reasonable. I'm never believing reddit rumors ever again. The rumors might even be true and HPM's economy might have been balanced on everyone's factories being only 50% efficient all the time! 120.zip

[Xylephony]

Sure thing, I will take a look later tonight.

[Xylephony]

Ignoring the 1956 save, the '85 save is definitely much better. There are still some aspects that worry me, including debt still broadly being very high and none of the GPs aside from the UK, US and Netherlands building a navy larger than maybe 35 ships, but 35 ships is better than the 10 or 15 we were seeing before.

I do still worry about the lower GPs, though, and especially Japan. As a nation which is so reliant upon its navy, it still only has 23 ships here, only 6 of them transports, and from the 1956 save it doesn't look like it upsizes its fleet significantly further. For the GPs as a whole, compared to the UK's 150 ships in the 1885 save, all of the GPs together wouldn't be able to contend with them. I'm sure the British would like to imagine that's historically accurate, but not quite.

I don't know that I'd say something else is 'wrong', per say. There's a lot of debt, but there was in HPM too at times. But I do think there are nations that should be building navies like Japan that are just not doing so at the scale they need to to be able to project power, and I'm not entirely sure why. Some of the defines you reverted were related to how many ships a nation should construct, right? Maybe reimplementing that change might help.

If you also removed all inventions that you added that impacted the world economy prior to 1900, HPM's infamous iron and coal shortages might be rearing their head here. I was noticing shortages in both, sometimes as significant as 2000 units or more, even with EEM.

[ujtcelvn]

I ran another test with the old EEM inventions and defines vales (except for FACTORY_PURCHASE_MIN_FACTOR and FACTORY_PURCHASE_DRAWDOWN_FACTOR ) and the results were very similar, with the UK with a massive 178 ships and only Italy, the FSA, the USA, and Russia above 30. Japan still only has 20 ships, and France is still drowning in debt somehow. I think I need to go into even older versions of EEM with even more coal and iron, which I nerfed because of late game overproduction, since late game overproduction might just be a just a meme caused by the FACTORY_PURCHASE defines values. 1885-2.zip

[Xylephony]

Depending on when you reverted those changes, the artisan overproduction from Electricity could've also been impacting it.

[ujtcelvn]

1885 is also a terrible time to compare navies because it looks like the AI will immediately disband all their ironclads and steam frigates the moment they invent battleships.

[Xylephony]

Hmm, that didn't occur to me. I chose 1885 because I thought the income from the Scramble might be important, but I had been consistently checking before '85 so I know the issue was persistent. Could move it to 1900 then.

[ujtcelvn]

I found a reason why deindustralization occurs in the lategame. Quite a few GPs become socialist then, and when they go to war they often stop subsidizing their factories, causing all of their inefficient factories to close and destroying their industrial score. For example, I saw Japan with over 1000 industrial score mobilize and drop down to only about 200 industrial score because they stopped all their subsidies.

[Xylephony]

That is a significant issue, but not for nothing I would've expected to see that in HPM as well, and I don't recall ever seeing crippling de-industrialization on that scale. In the rare cases where there was mass de-industrialization due to occupation or bankruptcy it was almost always quickly brought back up to at least 50% of the prewar industrial score as soon as the war concluded.

[ujtcelvn]

In the rare cases where there was mass de-industrialization due to occupation or bankruptcy it was almost always quickly brought back up to at least 50% of the prewar industrial score as soon as the war concluded.

Oh good, I stopped the test right after I noticed Japan's industry doing a nosedive and messed around with the files some more afterwards. Anyways in the newest update I think the AI being broke is basically resolved now.

[Xylephony]

Cross fingers. I'm in the middle of a DS3 replay right now but I can still spare some time to test here and there, I will run a game with the new version over the next few days and report back.

[Xylephony]

I did a pacifist test as Switzerland, partly so as not to impact the world economy and world militaries too severely, and partly since I had less time to test. I've just finished, and the good news is that you're right, industry is largely fixed. There are problems around the edges like you'd expect, especially as you hypothesized with de-industrialiation (it's taking longer than I anticipated in several cases where nations were fully occupied to recover their industry), but on the whole it's at a good point. Debt is still quite high, but because industry isn't unprofitable nations seem able to handle it and I noticed debt peaking at about 1890 and then gradually sliding down from there. Bottom-line is that I think this is a good economic base to build on top of - there's nothing fundamentally wrong with it at this point and all it needs is gradual tweaking.

Navies seem okay, but not great, and there are several signs that worry me here still. Navy construction through to 1860 seems mostly okay, but by the time Ironclads and Monitors are invented naval sizes drop across the globe. I'm wondering if this might be because nations are trying to build Monitors but they can't, because oil RGOs to make the fuel they need to be constructed largely haven't flipped yet. It might not be a bad idea to specifically restrict discovering monitors until 1870 as an invention limit, and then battleships to 1890, unless you have Sea Power & The Merchant Marine? That makes more sense to me, because it would let those with Sea Power develop the first prototypes more quickly while delaying adoption for other nations slightly, and perhaps increase the naval size of midgame fleets. The shortage could also simply be because of national poverty, though, and it seems like this is true in at least a few cases (Spain and Portugal in particular, despite having colonial empires and thus the need for transport navies, seem to be able to barely manage any ships at all). There are bigger problems with Japan which I will get to.

There were some resource shortages I noticed over the course of my game, but nothing I think can be addressed right now mainly due to some problem children whose issues are fucking with supply and demand: namely Italy, Japan, and especially Russia.

-

One of my main takeaways of this test (and this isn't unique to EEM, either; HPM had this issue too) is the fragility of low-tech regimes, especially those with national values or POP religions which encourage conservatism, even huge ones with massive populations and GP status. In this test, a massive crisis war ripped through Europe in the late 1870s, impacting almost every GP. From the conclusion of that war in about 1879 Russia was shaken, having taken on a lot of debt and war exhaustion. It soldiered on until about 1883, when the first major rebellion came along. Rebels rose virtually constantly after that, breaking Russia's army and coffers until finally the rebels started to topple governments. This is not exaggeration: Russia was in a constant state of collapse from 1885 or so to the mid-1920s. It wasn't until the discovery of Fascism, at which time revanchism in Russia led to a majority of Fascist support in all their states, that they could finally settle on a government with enough popular support to stop the constant risings. Prior to that, they were normally sitting at over 20% war exhaustion at any given time, and cycled through at least 3 governments a decade, usually more; in periods of warfare it was more still, as in 1885 when the first cycle started (Liberal->Socialist->Reactionary->Communist->Reactionary, all in the span of perhaps 2 years). In 1884, before the government first collapsed, Russia had 50% plurality; by 1890 it had 4%. In 1900, 3%; in 1910, 2%; and only by 1930, when fascists were stably in control, did it finally start to crawl back up to 13% (despite ongoing rebellions). But after stabilizing as a Fascist Dictatorship for a while and reconquering many of its cores, Russia was caught in another protracted war where much of its army was destroyed, and the whole cycle began again. In the late 1940s Russia fell first to liberals, then to communists, then to socialists, and in 1950 has an ongoing fascist rebellion.

I bring up plurality in that context mainly just to illustrate how absolutely unceasing the revolutions were, but also because it has direct technology implications. In 1950 Russia has 45% literacy and 38.75 research/day with all Philosophy techs researched. One of the other "problem children", Italy, despite also constantly falling to revolutionaries (I'll get to that) has 99% literacy and 57 research/day, even at 1% plurality. This compared to the seeming research cap at this stage of about 78.5/day. Russia still hasn't researched most commerce techs and a large amount of industrial techs, including several mechanization techs, most critically related to wood and oil output; because of Russia's size, this has a meaningful impact on the world market as a whole. It's even worse when one considers that so much Russian territory is occupied by rebels at any given time that both supply and demand are simultaneously being constrained.

GPs and huge states should be able to collapse. There's nothing wrong with a game where Russia is constantly attacked, never given a chance to rest, and collapses in on itself and never recovers. The problem is, Russia didn't participate in that many wars. They were just constantly in a state of rebellion they could not save themselves from until their POP ideologies began to align with a single party, keeping militancy low enough for them to stabilize after a revolution. But as we see by the 1950 stage that's a false sense of security and can be shattered easily. In some of the saves at the end of this post, you can see Russia has high militancy and ongoing rebellions and can't even pass any reforms, a staple of Autocracy as an NV. I fear that, rather than stabilizing the regime, when it goes through a shock like this it virtually locks a nation into a cycle of constant revolution because the nation in question can't actually pass reforms during these rebellion cycles to save itself. Italy suffered a similar fate, although not exactly the same; its revolutions were not constant like Russia's was, but it was consistently poor enough and subject to sufficiently frequent wars that every 5-6 years the regime would collapse. In its case, though, I think the problem was largely brought on by those conflicts and not as a function of its normal circumstances.

Bottom-line, though, I think there needs to be some sort of stabilization system or mechanic to stop true endless rebellion cycles when they start. I don't know how such a thing would be tracked to identify that a nation is actually in a rebel spiral, but at a certain point of such constant rebellion the nation should simply break entirely, as I see it. All POPs lose all militancy and consciousness and war exhaustion is set low with a temporary (but large) daily reduction to militancy and consciousness, but an authoritarian government takes charge (possibly anarcho-liberals, to represent that the state has completely collapsed and fallen to anarchy?) and the nation either cedes or releases all national cores it has, to be left with only territory that it alone has cores on: a sort of dismantlement not brought about by war, but the complete collapse of the state. As I see it, that is preferable to a nation being effectively nonfunctional for over 30 years ingame. If Russia actually collapsed like that it would stabilize enough for it to be able to recover some plurality, encourage intellectuals again and educate its people, and then gradually go to war to start recovering its territory. Or, if that isn't desirable, something just needs to be done where authoritarian rebellions which take place when a nation has less than 5% plurality have an immediate follow-up event that massively reduces militancy and consciousness, and that party wins the lottery and is just the new, "stable" ruling party.

-

Final notes are on Japan. I watched Japan get absolutely throttled this game. They united under Meiji in the mid-1860s, only to get an event for the first Sino-Chinese war within a few years, a conflict they were locked into for about 5 years. The only economic research Japan had significantly invested into at this time was Metallurgy, which does almost nothing for them because they only had one sulpher and three small coal provinces, so they were poorer than they could've been. Having only united in the 1860s they also had exactly two naval yards in the entire country, one at Matsuyama and one at Osaka, so they could only support a small navy and could not rapidly build one. The unciv event to conscript a navy was actually able to put together a navy 4x the size of Japan's, and even stronger when one considers most of Japan's ships were at 5% strength due to not being able to receive supplies.

After spending so much time mobilized (for nothing, since Japan was never able to put a navy together to fight China) it finally peaced out, heavily indebted, in the early 1870s. Less than a year later, the event to invade Korea fired. At this point Japan was still deeply in debt, had only level 1 naval yards, and had still researched no further economy tehcs. It had one ironclad to its name, and Korea successfully blockaded it for years with its unciv navy, leading to war exhaustion coupled with even further debt from being mobilized for, again, nothing.

Japan also stayed in that war for about 4-5 years. The year by this time is 1882. Russia demands Sakhalin and the Kurils and invades. Japan has still researched no economy techs by this time. They have good military tech, but are crippled by war exhaustion from the war with Korea and have no navy. They invade deep into Russia but the Satsuma Rebellion begins and they crumble; Russia takes the islands, Satsuma gains independence, and Japan falls to a liberal revolt. In 1884 Japan is in debt, with no army, two boats at 5% strength, having lost Sakhalin and Satsuma. It has researched two economy techs since 1867: Freedom of Trade and Experimental Railroad. I wish that were the end of it, but they would go on to have the second event war with Korea just four years later: again with no navy, again a loss after years of being mobilized.

My takeaway here is that Japan, after civilizing, is expected to rapidly be able to engage overseas without having a land route to its targets, but it simply can't. It has to build all its naval yards from scratch, meaning they're rarely ready for the first war it's called to, be it with China or Korea; it's putting so much priority into army and navy research that it can't get together a strong enough economy to actually fund a navy even if it did have the yards it needed to build one; and, even assuming that the yards were there and it chose tech more logically to have a stronger economy and thus better chance of actually building a navy, it's not clear that it would actually be able to. In most of the saves I have Japan's navy is both tiny and at 5% strength from complete inability to buy the goods needed to support it, even when they're sitting at 8th GP. It isn't always going to get the resources it needs to build a clipper fleet even if it wanted to.

My recommendations would be as follows:

1. Japan's historically largest ports should begin as level 1 naval yards from game start: Nagasaki, Yokohama, and Osaka
2. At game start, the Imported Weapons reform should be swapped for the Western Shipyards tech for everyone except Shimazu and the Emperor (since Nagasaki and Osaka would already have shipyards). While not being strictly historical the starting civilization level would remain the same, and this would give Japan the ability to build up its native naval construction capacity through the 1850s and early 1860s, which currently it cannot.
3. When Japan unifies and is civilized, there should be a decision, costing 50,000 pounds, to Centralize the Clan Navies (so it's applicable for both Shogunate and Imperial Japan), giving Japan 1 ironclad, 10 frigates, and 15 clipper transports. This would simulate Japan's purchase of the Kotetsu from France, as well as its historic integration of the navies of its domains. As a side note, when the Ezo Republic spawns it should have an event to gain eight steam frigates, since historically that was the remnants of the Bakufu navy which the rebels fled to Hokkaido with.
4. Tech weights might need to be adjusted further. You were still tinkering with them when I started this test so it's possible that the exact same issue wouldn't happen again, but I'd keep an eye on it.

I think, all together, that would give Japan the ability to start out any foreign war with an at least functional navy and the ability to transport troops rather than wallowing in inefficacy and spending decades mobilized with absolutely nothing to show for it.

[Xylephony]

Here are the saves, they're too large for Github to let me upload them as attachments.

[ujtcelvn]

A way to represent this that won't affect the how HPM really doesn't want Japan to start with ports seems to be railroading some daiymo and the Japanese AI to take the foreign shipyards reform if the Tokugawa caved to or surrendered to Commodore Perry, which was when they ended the sea ban and started developing ships in real life. In my most recent test the Boshin War started in 1862 and the Tokugawa already have 28 ships and Japan and the daimyo who took the reform have around 8 ships each.

[Xylephony]

That's with this hypothetical event in? That sounds very good to me, although I do still think Nagasaki and Osaka should start with ports; both were huge ports and definitely had the facilities to construct clipper equivalents.

[ujtcelvn]

Yes, that's with the event in. Unfortunately there's no way to prevent them from building ships if the ports are added from the start. I tried to make clipper transports, frigates, and man-o-wars not available from the start and have to be unlocked via invention but the game just crashes if I do so. By the way, the Qing DID have a permanent navy with ocean-going ships and DID have a bunch of trading ships all over Southeast Asia so I don't know what HPM was thinking when he removed ports from China. HPM didn't even give a justification for removing them! At least Korea and Japan had wikipedia links, China has nothing. Was he just mad Qing players were conquering the Philippines?

[Xylephony]

Probably. HPM frequently made punitive changes whenever he felt that anyone was abusing a system, and they were often made so rapidly to plug what he perceived as abuse that I doubt he thought them through all the way.

[ujtcelvn]

trying to build Monitors

Perhaps making monitors eat oil instead of fuel would help since there may be oil RGOs available but just not enough fuel factories.

authoritarian rebellions which take place when a nation has less than 5% plurality have an immediate follow-up event that massively reduces militancy and consciousness

There is no way to identify if a nation is actually in a rebel spiral so I added the above effect into the revolution event directly.

I also added a bunch of other stuff, just read the changelog. Unfortunately, the AI is obsessed with disbanding ships whenever a new ship type is invented.

[Xylephony]

Looks good. I'm still in the middle of my DS3 run but I will do a new economic test soonish.

[Xylephony]

I think this is another test partially killed in the cradle. There are absolutely massive iron shortages practically from game start, significant enough that I don't think it's possible to analyze the economy based off of the current version. In 1865, for instance, there was a 2000-unit demand for iron and a 500-unit supply; in 1850 it was even worse. I assume that prior tech weights which gave such heavy prioritization to metallurgy techs were probably as a band-aid for the coal and iron supply worldwide, and removing it to make the AI research more logical outcomes is causing crippling shortages.

I'm not entirely sure what the best fix for this would be. It's possible that the AI weights on metallurgy just have to be there so the world economy doesn't crash from iron shortage. Base mining efficiency could be significantly increased for iron and later efficiency increase techs could be nerfed, but that's not particularly historical, and it interferes with the sense of technological progression you get ingame. There's also the possibility that some major metal mines at, or near, game start aren't presently represented as RGOs and some basic goods could be switched over to provide better mining prospects at start.

Also, on the subject of monitors, the prognosis does not seem very good. Even in 1882 with fairly good world oil supplies I don't see a single nation constructing any. It's possible they're just coded to build only one capital ship type per generation, but I think that their nations just aren't getting good oil supplies and don't find the construction worthwhile. Since monitors never used oil for actual propulsion, why not just create a coal input for their construction and have oil as a maintenance purchase for them, to represent the lubrication oils?

Final note: Japan did really well this game. Built a big navy, unified, beat China and Korea. But the moment it researched steam transports and ironclads it deleted its entire navy to build the new ships (and, side note, since it didn't have many (maybe any) level 3 ports at that time it deleted its whole navy before really being able to build new capital ships...), which means it lost every single means of overseas transport it had, and part of Korea was promptly occupied by rebels and defected. I've seen this a lot in the past, where Japan doesn't fully lose control of Korea but there's a brief window where it can't stop rebels from occupying part of it and defecting, which I now suspect is related to that navy flipping when it doesn't have overseas transport capabilities. I don't know how to address either problem (AI deleting navies before it can build new capital ship types & Japan losing overseas control).

Saves

[ujtcelvn]

There are absolutely massive iron shortages practically from game start

I'll nudge tech weights again, but all I did was increase commerce and industry research for everyone, so metallurgy should have the same relatively high weight. Two tests ago even had an iron surplus from 1850 to 1880....

It's possible they're just coded to build only one capital ship type per generation But the moment it researched steam transports and ironclads it deleted its entire navy to build the new ships

The AI DOESN'T disband all their predreadnoughts when dreadnoughts are invented, they keep their old battleships even to 1956, with countries that build carriers even having mixed carrier - dreadnought - battleship - cruiser fleets. The AI really loves disbanding sail ships after discovering steam ships though. I don't know why.

since it didn't have many (maybe any) level 3 ports

Unfortunately there's no way to check what level a port is in a province, so there's no way to prevent the AI from rushing the ship construction tech before the naval base is built.

[Xylephony]

I imagined not, but it's frustrating to have a relatively high chance of Japan losing Korea just because the AI is stupid and deleted its whole navy. I'll keep an eye on that and try to see how common it is. Worst-case, in instances where Japan loses part of Korea without losing all of it maybe there could be a Vietnam, Algeria or Indonesia-style claim system where they could use a free CB on it to reconquer it.

I'll do another test sometime soon and see if the major shortages remain, but I'd be stunned if they didn't. An over 1k shortage at a point when only 500 units are produced is too big an outlier I think.

[Xylephony]

The AI really loves disbanding sail ships after discovering steam ships though. I don't know why.

Thinking about this more, would it be possible to prevent the AI from discovering Ironclads and Monitors unless they had at least one level 3 naval port?

Also I can confirm the iron shortages are still present and extreme. They were at 450 supply/1750 demand at one point; it's gone much further down on the demand side since then, but I think principally because there was a gigantic war in Germany and a lot of factories closed, and constant iron supply constraint has forced the closure of a lot of factories elsewhere. The demand is still significantly lower than the supply even now, and just a few years ago it was still a 1k difference. Saves.

Also, unrelated but odd behavior in that save: I can't unlock Mission to Civilize or Ironclads/Monitors, which I assume might be related to being a puppet (though there's no such visible restriction on the invention conditions ingame) but it doesn't seem rational in either case. More concerning though, literacy is having 0 effect on any of my POPs' CON gain. I have never seen this before that I can recall, and I have no idea why it would be happening.

[ujtcelvn]

Nope, there is no way to detect what level a naval base is.

Alright, I'll go adjust the coal and iron inventions again.

A bunch of inventions like the ones you mentioned have a requirement of "has_country_flag = existing_country", maybe releasing Australia yourself somehow misses adding that country flag? Yep, you are missing that country flag for some reason. Looks like the system that adds the existing_country flag to countries only gives the flags to the AI, neighboring countries, and GPs for performance reasons, and you have no neighbors so the AI can never give you the flag.

According to the wiki literacy only increases consciousness when plurality is >30% and your plurality is only like 3% so that is working as intended.

[Xylephony]

Looks like the system that adds the existing_country flag to countries only gives the flags to the AI, neighboring countries, and GPs for performance reasons, and you have no neighbors so the AI can never give you the flag.

Is there a way to resolve this for isolated tags like New Zealand and Australia without a significant performance impact? It seems like a potentially severe problem, particularly in Australia's case given that it completely locks you out from colonizing the interior.

According to the wiki literacy only increases consciousness when plurality is >30% and your plurality is only like 3% so that is working as intended.

That would explain it, since I couldn't invent most of the plurality decisions from Ideological Thought due to Britain already having it at game start.

[ujtcelvn]

An AI Australia would give itself the existing_country flag, just choose a different country for a day and save and it will resolve itself. The plurality inventions need the existing_country flag to be invented, nothing to do with Britain having it.

Is there a way to resolve this for isolated tags like New Zealand and Australia without a significant performance impact?

Not really, unless you want DEBUG events and decisions all over your face and run like 30 loops every second.

[Xylephony]

An AI Australia would give itself the existing_country flag, just choose a different country for a day and save and it will resolve itself.

I mean I can edit it into the save easy enough now that I know what it is, it's not a problem for me. I was meaning it for normal players as with the China popup discussion, but if it's not feasible to do it's not. That really blows for people not aware of the issue though.

The plurality inventions need the existing_country flag to be invented, nothing to do with Britain having it.

Oh, I figured it just did that thing where a released puppet starts with all of its overlord's inventions. I didn't even look to see those ones were greyed out.

[ujtcelvn]

I added more coal and iron, let's see if that's enough.

[Xylephony]

Alright, I'll do a new test. Coal had no shortages in the last two tests but I'd bet money that's because the iron supply was so constrained.

[ujtcelvn]

Looks like it's not kicking in early enough. A problem with increasing coal and iron is that sometimes it causes countries such as Japan to deindustrialize. Usually they have >2000 industrial score by 1956 but in this test they 648. In previous tests where I just increased the base amount of coal and iron Japan also deindustralizes, sometimes to like only 200 industrial score. Maybe it's because the USA. Germany, and the UK with even more coal and iron means they can build as many factories as they want and their factories just clobber everyone else to death. Or I just have bad luck.

[Xylephony]

I have another test in 1870 presently and the iron shortage has dropped off, but it was acute from the late 1830s through to the mid-1860s. That would suggest to me that iron's base production is too low, and that Mechanized Mining's big focus on coal can't compensate. And also perhaps that Regenerative Furnaces isn't being prioritized by many nations so the acute shortages aren't often redressed until late in the 1860s.

I'm also still worried about what demand would look like if the shortages at start weren't quite so acute, though. Demand in the 1840s was still regularly spiking into the 1800s, and with a supply of 400-500 that's a big gap. If early-game iron production is shifted up I'm not sure if demand would drop of more slowly or less, if iron-demanding factories were able to remain open due to the increased supply.

[ujtcelvn]

The big improvements in iron mining efficiency doesn't start until 1850 with Regenerative Furnaces. I'm still worried about deindustrialization caused by coal and iron overload though, I'm still not sure if it's not due to bad luck or overproduction. I'll try some more stuff out.

[Xylephony]

I know, that's what I'm saying, that the demand is incredibly high early on and Mechanized Mining's focus is almost entirely on coal so iron production can't really take off until the 1850s. But I'm betting that Regenerative Furnaces is being researched fairly late and that's why the major iron shortage is continuing through the 1860s.

I don't think the wheel needs to be reinvented per say in terms of net supply; iron and coal can still be underproduced relative to demand on average. It's just that the extremity of the underproduction early-game needs to be addressed. Hopefully if it's still underproduced by lategame deindustrialization won't be an issue.

[ujtcelvn]

Can you try the new version? There's around 30% more iron than previously, all unlocking pretty early on. It looked fine in my tests, there's occasionally jumps in demand by about 500 like you mentioned, but they go away in like a year and stabilize back to around the supply. I tried increasing supply further but it caused supply to be above demand most of the time. I've heard rumors that these jumps in demand are just the AI suddenly deciding to upgrade all its forts / ports / railroads / factories but you know how reliable Victoria 2 rumors are...

[Xylephony]

I suspect those rumors are actually correct, and that (economically speaking) the main issue with the previous versions was simply that the supply was so constrained that the spike was gigantic, crippling world access to iron for decades rather than a year or two.

I will try the new version hopefully later tonight and see how it looks.

[Xylephony]

New test looks much better, though at points in 1840 the iron supply actually catches up to and slightly exceeds the demand, so possibly a slight over-correction. Regardless, the economy looks stable enough here for me to do a full-duration test. It will take me a while, though, between finishing my DS3 run and BG3's release. I'll be slower for several weeks at least.

[ujtcelvn]

You're right, there IS too much iron, I think I DID overcorrect it. The newest update reduces iron a bit and also made sure the CSA wasn't banning slavery too early. The old version didn't work properly.

[Xylephony]

Hey, I have been stupidly busy the past few months but just FYI I'm not dead and will try to get back to testing this starting sometime hopefully next month.

[ujtcelvn]

What a coincidence, I have also been stupidly busy these past few months. All I got done was move some RGOs around and cleaned up some decisions.

[Xylephony]

So over the past couple days I had some downtime and decided to poke around while I had a chance, before I get busy again. I'm not far into the 1880s yet so this is far from an exhaustive test, but I did notice some things.

First, generally speaking, the economy is better again than in the last release. No major shortages, no constant bankruptcies. Navies are also better than before, although again that's with a caveat. Many nations, such as Russia and again Japan, struggle to maintain a navy beyond 1860.

That said, there are no major shortages mostly because there is huge overproduction (iron and coal both transcent demand by over 100 units, and would transcend it much further still if not for RGO unemployment), and while bankruptcies are suitably rare, debt is again not. Most European nations, even a few GP-adjacent ones like the Netherlands, are operating on a debt load which places their budget adjacent to bankrupcy and forces them to constantly underfund services. That should be likely for OPMs or extremely low-literacy, sparcely-populated nations like, say, Portugal; I don't think the Dutch should be in there under typical conditions. Also, the issue with smaller nations with more limited economies being wholly incapable of industrializing has been reduced, but is still present.

There is definitely a continuing problem here with balance and economic stability that needs to be addressed. However, as I thought more about it, I've started coming to the conclusion that at the moment we can't actually get at the root of what that is because of Germany. Germany is one of the biggest industrial producers and consumers in the game, and presently due to the P&E sphere changes in Austria's favor even the NGF typically doesn't come into being in a timely way. That leaves a dozen minor powers in Germany that have all of the economic shortfalls that one-province or two-province minors do, especially as regards reduced capacity to industrialize and mobilize armies, which then has direct knock-on effects on industrial demand. I don't think we can draw economic conclusions from games where Germany (or at least the NGF) does not form at roughly the time it should.

Now I stand behind those P&E changes, because they're historical. I don't think the solution is to revert them, I think the solution is to follow what was suggested in #87 and add additional, historical challenges for Austria which will make it extremely difficult for them to fend off the Brother's War, and thus more likely that Prussia will successfully secure hegemony and go on to form Germany. One would then hope that Germany would begin to form more regularly again and the world economy would (hopefully) start to stabilize a bit - and, if it doesn't, at least with Germany formed we can have confidence in the economic results we're seeing.

These are the changes I would recommend around the entire Hungarian/First Italian situation:

1. Move the Hungarian Revolution to a model similar to the CSA, where Hungary is puppeted for ~6 months before war breaks out with Austria. This is historical, as Ferdinand I gave the Hungarians autonomy to try to appease them while Austria was wracked with revolutions and dissent from the Springtime of Nations. A follow-up event would then fire six months after they've been released regarding the Hungarian refusal to ratify the Austrian monarch as King of Hungary (again historical, as they refused to recognize Franz I) and forcing Austria to either declare war to reconquer Hungary or suffer a massive prestige hit and let them go.
2. The same day that the event for Hungarian autonomy fires and Hungary is released, Venice should also be released controlling solely the state of Venetzia and puppeted to Sardinia-Piedmont. Sardinia-Piedmont should then be forced to either declare war on Austria for control of Lombardy or suffer massive prestige and militancy consequences. As a function of this event, all of the Italian states that are not in Austria's sphere should be allied with Sardinia-Piedmont and called into the war for the liberation of Lombardy. This is actually more lenient on Austria than history itself was, as even the Habsburg monarchs of Italy sent troops against Austria in the real First War of Italian Independence, but it gives a clear gameplay benefit if Austria is able to retain its stranglehold on Italian politics and keep the Italian states sphered, and a massive consequence if they cannot.
3. If Austria beats Sardinia-Piedmont then Venice should be annexed back to Austria, though otherwise it should remain a puppet of Sardinia-Piedmont after they win.

I think this would instantly solve the problem with Germany's formation, because Austria would, every single game, be taking a catastrophic beatdown from 1848-1860. First the First War of Italian Independence followed up immediately by the Hungarian revolution, then possibly even in the midst of those ongoing wars Prussia will have researched N&I and be able to strike for hegemony. And, even if Austria is somehow able to hold on against all those threats, they have to be ready for France to immediately press Plombieres in 1860. I think Austria will be consistently too weak to be able to resist Prussia.

I do see two major potential downsides of this modeling: first, that Austria could be utterly crushed and fall so far down that they lose GP. I don't think this issue is as much of a problem, because it almost happened historically and, more importantly, Austria typically isn't as major an industrial power as Germany so it's not as dangerous if they have a decade or two of total chaos before they can get back on their feet, and I do think they can reasonably recover. The second issue, mass revolution in Austria due to years of occupations and ongoing conflicts - that's a bit more of a real concern. I think it can be mitigated by giving Austria events to reduce its war exhaustion and militancy for each of the two rebellions that are successfully defeated, so if it beats either of them it gets a bit of a respite so it's not guaranteed that rebels will break the country. It would need a live test to tweak balance, but I think it'll be okay.

Either way I think this is the best way to solve the Germany issue and get a stable economic baseline to test off of, with proper industrial demand.

[Xylephony]

I tested a bit but it seems inconclusive at the moment. The new war seems to be a bit too easy for Austria to put down. In my tests Sardinia-Piedmont went in to defend Lombardy and Lombardy was released as a neutral non-belligerent, but then San Marco declares independence, Lombardy joins them in their war, and both get crushed. Austria seems to be able to rely entirely on its spherelings to beat both wars while it handles Hungary on its own, so they're not being seriously weakened by the conflict and Prussia is still afraid to initiate the Brother's War.

Of note, the Papal States and Naples were not sphered by Austria but were never called into the war on San Marco's side, which is much of what made that conflict so trivial for Austria's spherelings to put down. Not sure if that's a bug or intentional, but based on what you said in #87 it doesn't sound like it. Might be if there's an MTTH to call them in it's too long, not sure how you're modeling it. I would change it so Lombardy doesn't get called by Venice (it didn't fight on Venice's side so much as support Sardinia-Piedmont's) and is maybe forced to fight on S-P's, but increase the likelihood that other Italian states not sphered by Austria do.

Also, on the subject of S-P, I think the initiation event for the war could use slight change. Right now declining only increases the militancy of those who support Jingoism, but Italian unification was a broadly liberal-nationalist project and it was mostly liberals who were rabid about it. Declining to support Lombardy should increase Liberal militancy even more significantly than Jingoists, while accepting should reduce Liberal militancy (this will also help S-P weather the Springtime of Nations, especially given that it'll be concurrent with the war in most cases) at the cost of increasing Liberal support in the country, maybe by 5%.

Final note, I think it's actually punishing to Sardinia-Piedmont if it wins the war. I tagged over to them to test what would happen in the case of victory, and it seems like nothing. Lombardy is independent but Sardinia-Piedmont can't use the Call to Union without controlling it, and there doesn't seem to be any event to petition for annexation. I assume they'd probably petition to join S-P if S-P was a Great Power and had them in its sphere, but even when I was playing as them S-P ended the war more beat up than Austria was and actually dropped nation rank, with no postwar event to give prestige or WE reduction to help compensate. So that leaves them without any territory gained and with a now-independent Lombardy they probably have to burn a Conquest CB on to annex if they want the land, since the war was probably devastating enough to knock them out of GP contention.

They should definitely get an event that gives a large boost to prestige if they win I think, and maybe there should be a chance (no more than 50%) that Lombardy petitions to be annexed to Sardinia-Piedmont outright. If Lombardy chooses to do so the effects of the Call to Union should perhaps automatically fire, to represent that S-P is transforming into a sort of "north Italian federation" and Italian unity is seen as being at hand, whereas if they decline S-P might be given the choice between accepting their independence for an infamy burn and increased relations with Austria & France (neither wants a united Italy) at the cost of slight liberal militancy, or an immediate Conquest CB at the cost of a more moderate amount of infamy (10 instead of the possible 22 from manual fabricating?).

[ujtcelvn]

Lombardy was released as a neutral non-belligerent THAT is a silly bug, oops, Lombardy shouldn't be released if Austria chooses to suppress the revolt and S-P should be invading if they choose to intervene - the revolts irl was limited to Milan and Venice and over half of the Italian troops stayed loyal to Austria Declining to support Lombardy should increase Liberal militancy even more significantly than Jingoists, while accepting should reduce Liberal militancy (this will also help S-P weather the Springtime of Nations, especially given that it'll be concurrent with the war in most cases) at the cost of increasing Liberal support in the country, maybe by 5%. Done there doesn't seem to be any event to petition for annexation. The original Sardinia-annexing-all-unsphered-or-same-sphered-Italians event should still fire after a while, if Austria decides to release Lombardy without a war The Papal States and Naples were not sphered by Austria but were never called into the war on San Marco's side Yeah this sometimes happens, but I made them more likely to join the war now

The biggest problem now is still Austria's allies fighting the Italians singlehandedly, often adding the liberate Romagna cb on the Papal states...

[Xylephony]

The biggest problem now is still Austria's allies fighting the Italians singlehandedly, often adding the liberate Romagna cb on the Papal states...

Yeah, that is an issue, given that the Pope didn't support the venture at all IRL and was forced into it by circumstance, and then held to it only because his army defected after he outright ordered them to return. Punishing the Pope would be unrealistic given Austria's relationship with the Church and the Pope's disinterest in joining to begin with. More than that, freeing Ravenna would further the purposes of the liberals and a unified Italy outside of Church control, which would run even more counter to Austria's interests.

Historically all the Italian states rebelled against Austria, even ones with Habsburg dynasts, but I can't see it as reasonable to model the same in the mod because that would entirely defeat the benefit of Austria sphering anyone on the peninsula. At the same time, though, with only Venice, the Papal States and Naples supporting the revolt it will definitely be put down easily by Austria's spherelings.

Would there be any way to at least stop the Italian states from being called in on Austria's side? Such as having the Venetian revolt event forcibly break alliances with all of them, so Austria would need to re-ally all its spherelings after the war already started, when they won't be called in? If the Sardinians & Venetians don't need to fight the other Italian states that would be both more historical and a bigger challenge for Austria.

Also, RE rewards for Sardinia-Piedmont proving victorious in the war, I still think a prestige gain of ~15 at minimum is justified, even if they are able to integrate Lombardy later. As I mentioned even when I was controlling them I got wholloped and dropped nation ranks, so succeeding should help give them a bit of a boost to keep them in GP contention.