Request: please rework Africa to have a shred of hope

For updates and discussion of Terra Invicta, a grand-strategy alien invasion simulator
Post Reply
Matman
Posts: 2
Joined: Mon Nov 14, 2022 1:51 pm

Request: please rework Africa to have a shred of hope

Post by Matman »

First: loving the game, totally addicted. Incredibly ambitious project.

My biggest gripe is the way African nations are currently set up at game start. They are all on the brink of collapse. I get that it should be a major disadvantage to start in poor nations with low resources, but as a South African, I was looking forward to at least trying. You guys would surely have read Artemis by Andy Weir. Why not a space center in Kenya? It's got a lot of advantages: near the equator, good infrastructure, Nairobi is on an elevated plateau..... let's try.

What I found is that the game models all African countries (with the possible exception of Egypt) as "about to implode". That the entire continent is 2-3 years and one shock news item away from all-out war and societal collapse. That sucks. Stings a little. In the game, high inequality inevitably leads to anarchy which inevitably leads to civil war, and the cycle continues into totalitarianism and a never-ending series of coups. Countries start with an insane percentage of spoils (corruption does exist, but not 30-60% of government spending, lads).

In the real world, there are some dynamic economies, young populations, imperfect but relatively stable democracies that have existed for decades and where the leadership has passed to the opposition and back again (a crucial test of democracy). It's a big continent! Not everywhere is a cesspool of violence.

What's worse, others have already mentioned that low investment points already make it nigh impossible to reverse a negative trend. Pumping 100% into welfare (or anything else) has a negligible result in poor nations, due to the few investment points, regardless of population size. So even if a player focuses entirely on the continent and puts investment points into the right places, more often than not it's a losing battle anyway. And anything taken by the AI, or not taken at all, just collapses that much quicker.

So, my first request would be to set African countries on a better path at game start, one that isn't so completely pessimistic.

My second request is for societal collapse, if it happens, to happen more slowly.

My third, more complicated request, is to create the potential for a country to ameliorate its situation without player intervention. The AI never seems to develop countries, and once at the bottom, countries are stuck there unless the player invades, unifies, and pumps direct investment every year. Even a coup should be able to result in an improvement. The cycle can be broken. Where are the Thomas Sankara's of the world?
User avatar
johnnylump
Site Admin
Posts: 1208
Joined: Wed Nov 11, 2015 4:12 am

Re: Request: please rework Africa to have a shred of hope

Post by johnnylump »

Thank you for your thoughts.

Just so it's absolutely clear, we did nothing specifically to single out African nations as worse than other places just because they are in Africa. The design goes to great lengths to avoid any kind of essentialism in any nation or culture.

The game has a model of internal politics and international relations that is the same for all nations. We plug numbers from real-world indices where they are available into that model. Outcomes are the product of that and some RNG to cover circumstances outside the model.

GDP, inequality, democracy, and education are all straight from various indices. Unrest is rooted in a violence index that's adjusted upward when insurgencies are active. Cohesion and Miltech are eyeballed. The starting priority templates are eyeballed, too, although I've generally assumed the factions will assume control of most nations relatively quickly and correct things. More on that below.

(I suppose there's a grad-school-level debate about the pro-western bias in those indices; I'll vaguely concede that's a possibility but I think a lot of social science researchers are smart enough to be cognizant of that possibility as well and try to control for it. I also don't think there's any better data sets out there that we could use.)

The game also models various observed phenomena like the resource curse (nations with lots of natural resources have trouble maintaining a democracy) and that anocracies suffer much more instability than democracies and authoritarian states.

Nations near the equator get more boost from the boost priority than those at high absolute latitudes. It had been my hope that would make African nations like Kenya attractive versus the big northern hemisphere empires, but that doesn't seem to be the case. Small, stable nations are generally less expensive to Direct Invest in than big ones.

Because, honestly, I don't know the outcomes of every interaction of the model; the system is too complex so I can only watch what it does like players do, and try to make some sense of the feedback I get. I'd note here that my home nation of the USA falls into mass civil violence pretty frequently in the game, as well.

The game doesn't define IPs as government spending. It's fungible economic surplus from public and private sources -- abstracted production that isn't committed to keeping the nation together. IPs currently scale in a cube root to GDP, meaning smaller nations have more IPs per unit relative to big empires.

One of the underlying themes in the game is that humanity can do amazing things if it can get its act together. Spoils is one of the main obstacles to that -- it represents people enriching themselves as the expense of common good, a universal problem that some countries are better able to reduce than others through strong institutions. The factions are all a kind of willpower to overcome that (except the Initiative, I guess), so the 'story' in all that is course-correcting these nations to utilize their resources for what they define as the general good. The minimum required spoils value is rooted in a value the model calculates, called "corruption" internally. It's based on democracy score (strong institutions) and education. 50% spoils doesn't mean half the economy is being extracted and exploited for the elites; it means half of what could be reasonably directed other things is instead going for that.

The easy handwave here when the model produces worse-than-expected outcomes is that the collective shock at the alien arrival scrambles things for the worse. I don't think that's a huge leap of logic.

ALL THAT SAID, I'll generally concede your main point. I just wanted to be clear how the game gets there -- buncha numbers go into a model common to all nations, and corruption and unrest and coups and all that are what come out. Tweaking it is possible but also complex as changing a number changes behavior across the system and we have to watch a bunch of playthroughs to see what the ultimate effect is. Based on player feedback, the corruption calculation is probably a bit too harsh. I'd very much like to see the global south contain more viable options for the player, but I don't want to overcorrect and pretend that they don't have a tall hill to climb.

Anyhow, thanks again for your thoughts; I've been thinking about how to approach this and will continue to do so.
2alexey
Posts: 27
Joined: Tue Nov 15, 2022 3:52 am

Re: Request: please rework Africa to have a shred of hope

Post by 2alexey »

Current Africa is not a viable start/main focus, but it is absolutely is a huge bonus if you fix it.
Low economy is in many ways a blessing, since holding states in Africa is extremely cheap and thus mission control is very good to be built there, and there are nations that have reasonable investment points, which you can also stack by proper advisors/getting more direct investments.

While the game's model would probably benefit from reducing feedback strength of discontent -> poor economy ->discontent ,unfortunately, since faction actions like coup and invasion tend to damage countries, that reduce economy, and reduced economy leads to states being more succeptable to takeovers African suffers from double feedback cycle. In my game, since
I took over most major powers as resistance (North America, Pan-Asia, Greater India and Eurasian Union) Humanity first, that controlled EU and UK would endlessly invade African countries held by servants/protectorate, and then I assume lose those countries to revolution/agents actions, only to invade again.

Also, unfortunately agents stirring up decent can only be effectively countered by other agents lovering it, which I presume is the reason we often see some countries suffer endless revolutions, as "national government" can't counter dissent on its own, which is a "game abstraction".

I assume a semi-decent approach to Africa would be to take one state that is relatively decent GDP and unification base, improve it for 1-2-3 years and then unify other states into it, continuing to improve it's economy. That would really help in long run, since it would potentially give x0.25 global science points, or so a huge boost to midgame.
Malaan
Posts: 1
Joined: Thu Nov 17, 2022 9:17 am

Re: Request: please rework Africa to have a shred of hope

Post by Malaan »

In my first games I thought Africa would be a waste of time. Tried and tried, but didn't work. In my present game I changed my strategy and it was my second focus and I made it prosper. You can unify the entire continent which might even be a bit too strong and in terms of military strategy it is very easy to hold if you can blockade and you have Egypt as a chokepoint to the near east. You can even go for Israel with is rugged and core economy and have an even greater defensive point. I find it quite easy to take hold of it and being "sneaky" about it. Just grabbed the empty control points and worked my way up from there.

One thing that bothers me a bit about the unification process in general is that you know which country you need to start the process. Maybe a bit of randomisation in the future?
Ian_W
Posts: 42
Joined: Tue Nov 08, 2022 8:10 pm

Re: Request: please rework Africa to have a shred of hope

Post by Ian_W »

I think part of the problem is the relative costs of controlling small, poor countries as versus rich, immediately useful ones.

If Africa was cheap to grab, and you could effectively stabilise it so it was useful for the move into space (and Im thinking of the costs of direct investment as well as the opportunity cost of Stabilise Nation missions), and the AI wasn't very interested in it, then a player could spend most of the 2020s fixing up africa as the NPC powers kept attacking each other in Europe, North America and so on.
neilwilkes
Posts: 120
Joined: Sat Oct 01, 2022 10:44 am

Re: Request: please rework Africa to have a shred of hope

Post by neilwilkes »

Ian_W wrote: Fri Nov 18, 2022 5:28 am I think part of the problem is the relative costs of controlling small, poor countries as versus rich, immediately useful ones.

If Africa was cheap to grab, and you could effectively stabilise it so it was useful for the move into space (and Im thinking of the costs of direct investment as well as the opportunity cost of Stabilise Nation missions), and the AI wasn't very interested in it, then a player could spend most of the 2020s fixing up africa as the NPC powers kept attacking each other in Europe, North America and so on.
But that is not how African nations tend to work in reality (and yes, this is only a game).
I strongly suggest you read the book 'It's our turn to eat' (https://www.goodreads.com/book/show/629 ... n-to-eat) . I met the author, and heard the tapes he recorded when the others did not know they were being recorded and to call it an eye opener would be an understatement. Instead of doing his job, he was told to 'just drive around in your little car and keep your nose out of things'.
Fixing the corruption - and also the discontent - is very, very difficult simply because of the 'It's our turn now' attitudes that seem to prevail.
Ian_W
Posts: 42
Joined: Tue Nov 08, 2022 8:10 pm

Re: Request: please rework Africa to have a shred of hope

Post by Ian_W »

neilwilkes wrote: Fri Nov 18, 2022 4:05 pm
Fixing the corruption - and also the discontent - is very, very difficult simply because of the 'It's our turn now' attitudes that seem to prevail.
Which the game reflects well by elites being unhappy if Spoils are not high enough, and this making coups and loss of control very easy unless the country is Spoiled, and if the country is Spoiled then it's essentially kept in it's current condition.

Deep seated corruption is a wicked problem - it was around 190 years between the coup against Governor Bligh and the Wood Royal Commission, and the New South Wales was corrupt for all of it (and Wood didn't fix the NSW Police, either).

This vicious cycle can be opposed in game, but you're going to need to use a lot of time and actions in Stabilise Nation missions, CP in Direct Investment, and so on.

Edit : Just thought of a pretty justifiable thing ... Space Program expenditure is treated like 50% Education and 25% Unity, representing most of a space program is actually training engineers of various sorts, and it provides a definite focus for national pride and national unity. This should feed into the things that help break the vicious cycle. Add in making the cost of Stabilise and Protect missions relative to national GDP and we get a major balance between the First and Third World (I also suspect that something equivalent to the Scout rocket, which at the end of the day was a small payload put on an anti aircraft missile put on a anti ship missile put on an existing ABM missile, which is the 'small payload into orbit' would cost a lot less than the cost of equipping a full Army corps).

An ideal situation, I think, is the secret cabals of AI (and player !) squabbling over the countries that already have development and space programs, and while a determined group builds something in what were seen as the waste places of the world.
neilwilkes
Posts: 120
Joined: Sat Oct 01, 2022 10:44 am

Re: Request: please rework Africa to have a shred of hope

Post by neilwilkes »

Some really good points here, Ian.
I like the Space Programme thought - but something is bothering me and I need some time to work out quite what it is.
More later
Matman
Posts: 2
Joined: Mon Nov 14, 2022 1:51 pm

Re: Request: please rework Africa to have a shred of hope

Post by Matman »

wow, super impressed by the dev interaction and the discussion. This is an amazing community.

I was wrong to say that African nations are modeled as being "on the brink of collapse". I should have said they are in the process of collapsing. That's the issue. I would bet a fair chunk of cash that if the game could be run without aliens or factions, that the starting conditions and modeled effects would be enough to bring most African countries, and many SEA/South American countries, into the death spiral of coup, inequality, cohesion, unrest. For countries with many decades of stability (despite very real problems) this implies a flaw in the game model.

Alexey, you've mentioned the pro-cyclical effects, I agree. They seem almost inescapable, these small poor countries in-game don't seem to have enough IPs to really reverse negative trends. Unless perhaps that is all they are doing for the first ten years. I tried again in a new game, started with very high popular support in Egypt, Ethiopia and Tanzania. It seemed too good an opportunity. Long story short, I needed all my counselors on pretty permanent damage control just to keep an even keel.

I see a few potential solutions, but I understand it can be hard to test without creating big ramifications elsewhere.

1. reduce starting spoils. If spoils represent "people enriching themselves at the expense of common good" rather than pure corruption, then it really makes no sense to start with 30-60% of total IPs. There are plenty of greedy self-aggrandizers diverting necessary resources for their own ends all over the world. Africa is not unique in that respect. But if spoils do represent corruption, then even in corrupt regimes, it's hard to see 30-60% being diverted from "fungible economic surplus from public and private sources". The big court case in South Africa a few years back, involving the then-president Zuma, was over R2m, about $300k. That's a tuesday lunch for the elites in many rich countries. I don't mean to gloss over the very real and very problematic issue of corruption, but many countries still function. Arguably, capital flight is the real issue there. Surplus being siphoned out of the country, into richer nations. Then you'd be modelling neo-colonialism. While this might be realistic... damn man, that's dark! Hopefully too dark.

2. Reduce the effects of poverty and anocracy on unrest. Yes, studies do show a rough correlation, but to put that into a model creates inevitable causation, which is dangerous. It's one thing to say many countries with high unrest are poor, but to model a world where poverty itself leads to instability is quite another. If the effect is small, it can be interesting background noise, but right now both effects dominate their respective countries. In reality tThere are many poor countries which are very peaceful, some where the culture itself is peaceful. Similarly with anocracy, yes studies do show anocracies as being less stable, but anocracies are also a strange lump of a category. By some metrics Russia is considered an anocracy, so are Somalia, Madagascar, Rwanda, Singapore. These countries have nothing in common. An anocracy is just "democracy but with flaws", it's a very developmental definition that creates a spectrum, at the end of which is the ideal of western democracy. Any other system of government, besides an autocracy, would fall under the label of anocracy. Successful or otherwise. (the game necessarily needs many abstractions, and this is one of them. Quite reasonable. But you could still consider reducing the negative effects, to average out the nuances of the real world).

2b. Consider the strength of the effect of inequality on cohesion as well. If the IPs of the country cannot produce enough unity to create a sense of national identity, then something may be wrong. We've seen "nationalizing moments" in many countries with very colonial borders, so-called "state-nations". It doesn't seem to cost the world. Maybe welfare and unity could be proportional to population size? That would be very interesting, but I accept that this would likely have too big of a ripple effect.

3. I really like Ian's suggestion that a space programme could have positive impacts on the economy and education. If implemented correctly, it would imply investment, including foreign private investment, local spending, a local multiplier effect, local jobs, and technical skills in demand. If this is modeled as an absolute rather than a relative effect, as it should be, it could be relatively negligible for big, high-population, high-GDP countries, while being quite beneficial to small poor countries. The UN office in Nairobi is thought to provide a pretty chunky boost to the local economy.

4. Create a few more starting federations. The EU is such a great place to farm MC because it begins in a federation. That's realistic, and the EU should be very connected at game start. But there are already some regional partnerships in Africa. The SADC, UEMOA, the EAC... and of course the AU. None remotely so integrated as the EU, but not too far from the Eurasion Economic Union. One approach might be to start with the tech unlocked and countries allied but not in federation, adding one more step relative to the EU.

5. Reduce the cost of direct investment. Perhaps make it proportional to both gdp and population size? GiveDirectly and Grassroots Economics both show the potential of local multiplier effects. Even a small amount of direct investment in low-population, low-gdp-per-capita countries can lead to a big increase in the effective demand for goods and services.

6. Model some counter-cyclical potential. In the game, a country stuck on high unrest and low cohesion tends to stay there. In reality, you might have a strong or weak civil society, religious institutions, national figures. History and culture play such a big part. Without modelling any of that, even a revolution or coup can be bad or it can be good. A revolution can be a unifying event, a coup can mark the end of a deeply unpopular regime. Separatist movements might create national unrest but an increase in local cohesion. Not all sudden changes in government are created equal. I accept that this would be impossible to properly model, but having a small percentage chance of significantly increasing stability and lowering unrest, for every sudden change in government, could be enough. (a big enough improvement to counter the downward cycle, not just slow it down)

edit: 7. Maybe a steeper gradient in the relationship between GDP and control point cost? And perhaps a reduction in CP cost for autocracies and anocracies? That would be interesting. It would reflect the lower amount of effort necessary to control/coerce/convince a small group of elites, vs a vibrant democracy where thousands of voices are constantly competing. It would also mean that the player could more easily grab "secondary" countries and develop them. Hopefully the AI would too. That's a whole other issue, the AI's propensity to loot rather than build. Can be realistic, but so dark!

All of this is a gut feeling though. What I would really love to do is to run the game without aliens or factions, and output country stats and events, month on month, into a database. Then run some analysis on which starting conditions led to collapse, how fast, and (hopefully) why.
Ian_W
Posts: 42
Joined: Tue Nov 08, 2022 8:10 pm

Re: Request: please rework Africa to have a shred of hope

Post by Ian_W »

"2. Reduce the effects of poverty and anocracy on unrest."

I think the way TI models poverty and unrest is wrong.

People being poor does not seem to trigger social unrest, if people see things are, on the whole, getting better. But if things start getting worse, then troubles start.

I'd like to point to the very stable Suharto government in Indonesia, which survived just fine despite widespread poverty from 1968, but did not survive the economic hit from the Asian Financial Crisis of the late 1990s.

Here's a bunch of other examples from Africa.

https://www.cfr.org/backgrounder/africas-leaders-life

Now, TI *is* very dark with it's background. We've got the combination of global warming, alien shenanigans and secret cabals quarrelling. Johnnylump *is* right about the Usa usually falling into a spiral of civil disorder in TI, unless active measures are taken.

And yeah, spoils *is* neocolonialism and capital flight.

But it'd be nice if more strategies were viable, and the meta strategies were nerfed (or at lease the AI was trained to recognise and contest them).
Post Reply