Population Model v. 0.1

[Builder]

I've just read the thread in Civ3 section(?).

I'm willing to join!!!

[The Joker]

Builder:

Welcome on board!

Right now most of the design discussion is about the population and economy models of our game. A fast overview of the two appropriate threads should show you that we are making massive changes to these two aspects of the game compared to Civ2.

Many people have some favorite aspects of the game that they like to develop. What are your favorite game aspects? If they are ones that we are not discussing at the moment then you could try finding one of the old threads on the subject and post some ideas or comments, or you could create a new thread with some ideas that you would like to implement. We will of cause respond to whatever you do. You could also read through the economy or population threads, and comment on the models there.

Whatever you chose welcome on the team!

------------------
"It is only when we have lost everything
that we are free to do anything."
- Fight Club

[Guildmaster]

Ok so how exactly are we going to track population?
What kinds of formulas are we going to use?

------------------
Goes down smooth when I get a clean hit of that skunky, funky, smelly green sh.t
- Cypress Hill

[The Joker]

About the middle of this thread I have posted an age group submodel, which could be used to handle pop changes. It includes a birth rate and a mortality rate, and also includes people in different age groups.

It is pretty simple. I think even I could program it (but that would mean that it would be dosbased, which is not good). I also think it wouldn't be too dificult to link it to the econ model, once that has been developed a bit further.

------------------
"Never underestimate the power of denial."
- American Beauty

[heardie]

http://iln.go.com/html_p/c/341134/52068/341166.asp

How's that for good reading? I don't get it, because I haven;t learnt Calculus or Differential Equations but I like the bit the says the dude estimated the US poplulation of 1950 to within 1%, considering he made that prediction in the 19th century

[The Joker]

I think I have learned something similar in school. It might be a good way to describe a population in the modern world, but in ancient times I think it might be a bit more complex. Often the population actually dropped. The food supply (which was far from constant) pretty much determined the population growth back then.

We might be able to use that function for population, but I actually prefer a more "real" way of doing it, where we actually determine how many kids people have, and find the mortality of the people. This could model pop growth in a more "realistic" way than anything else.

------------------
"Never underestimate the power of denial."
- American Beauty

[heardie]

Herers are few more links on the topic
Maybe they will help http://www.cquest.toronto.edu/zoo/bi...w/appendix.htm
http://web.city.ac.uk/mathematics/X2...pulation1.html
http://www.okstate.edu/artsci/zoo_ho...ati/growth.htm
http://www.arcytech.org/java/population/facts_math.html [ Big one ]
http://www.math.duke.edu/education/c.../contents.html
http://www.jump.net/~otherwise/popul...opulation.html
http://www.ent.orst.edu/jepsonp/2000lecture5/

[The Joker]

I think all these links are saying virtually the same. There are ways to mathematically portray changes in population.

But these are all using "virtual" variables such as births per year. How do you know how many births are per year without counting them (which we can't, since we're talking about virtual populations). Therefore I think we need to use other variables - kids per family and # of families.

This makes calculating the population is a bit more complex. But it also offers more versatility to the model. So if a lot of young people die in a war, for instance, there will be fewer potential parents and therefore less kids.

------------------
"I chose not to choose life. I chose something else."
- Trainspotting

[ElmoTheElk]

You're right about that. And it's done (partitially) in Civ2 this way also. (i.e. # of kids in a family).

Only far more expanded

[amjayee]

I really like mathematics and find it interesting, but I think the same as Joker. As I have said earlier, the functions presented on those pages are good for estimating long-term changes. They do not work in our game, since we need short-term changes in population; there is births, then there are deaths of old people, and infant mortality. But we need also deaths caused by famines, wars, and diseases. So I think it would really be best to have people get children, then kill some of the people every turn in various ways. In long term, the population changes will follow roughly some kind of exponential curve, but it also stores some data about the history of that population; wars, famines and great pandemics can be seen from the graph without knowing they existed.

[The Joker]

Elmo:

Except that the births per family in Civ2 wasn't used for anything at all. I really loved the demographics of Civ2, but they had absolutely no effect on the game. That is to be different in GGS.


Amjayee:

This is excactly like what I think. We need to figure out short term pop changes, and then store it all, so you can get a graph showing the pop over a long timeframe.

------------------
"I chose not to choose life. I chose something else."
- Trainspotting

[S. Kroeze]

I hope that the following quote and especially the figures will be of some benefit while making a mathematical model of population growth. In my opinion it shows that shrinkage of population during several centuries actually did occur.

quote: When the population did increase, growth occurred at a very slow rate; annually the rate of increase was always below the 1 %. To illustrate my point I will give estimates of the population of France over a lap of time, France being the country we are best informed about. Of course they are just estimates, but they are at least intelligent estimates! ~400 BC: ~3 mill ~~ 1 AD: ~5 mill ~200 AD: ~6,5 mill ~400 AD: ~5 mill ~600 AD: ~4,5 mill ~800 AD: ~5 mill 1200 AD: 10,5 mill 1300 AD: 16 mill 1400 AD: 11 mill 1500 AD: 15 mill 1550 AD: 16 mill 1600 AD: 18,5 mill 1650 AD: 21 mill 1700 AD: 22 mill 1750 AD: 24 mill 1800 AD: 29 mill And some figures for Europe as a whole(estimates too of course, even more uncertain): ~400 BC: ~20 mill ~200 AD: ~36 mill ~600 AD: ~26 mill 1000 AD: ~36 mill 1300 AD: ~79 mill 1500 AD: ~81 mill 1700 AD: 140 mill Some conclusions can be drawn from these cold figures: - population growth before the Industrial Revolution was very slow - the availability or lack of food was decisive - the level of agricultural technique determined the magnitude of the average harvest, limiting population size - as soon as this limit was exceeded, disaster was inescapable and population shrinkage set in (later Roman empire, fourteenth century) - after a disaster it took several centuries to reach the ancient level again - a well-organized government could relieve suffering, but couldn't avert disaster

I have quoted a post made by myself in a now forgotten thread. I hope a disease model will be linked to your population model; otherwise it might turn out rather unrealistic.

Regards!

[The Joker]

I remember you posted that list, S.Kroeze.

And we should make sure that the population will only rise, if there is food available. What will propably be most dificult is to make sure that when the population rises to above the max level it will suddenly drop way below it, because of disease etc.

But with a good disease model, and with resistance to disease linked to the overall health of people (and therefore to the food per capita) it should be possible to have an epidemic suddenly strike, and kill a lot of people immediately.

------------------
"I chose not to choose life. I chose something else."
- Trainspotting

[Guildmaster]

hehehe

I like where this is going...
Insert disease model. Remember, of course that each disease is tracked by
1) The incubation period of the disease
2) How easily it is spread from person to person
3) How many people who get infected are going to die
4) How hard it is to cure the disease

What we could do is add a "threshold" to the health aspect of the population model. So if a sample of the population has a health of 6, that would mean that the disease has to have an infection rate of at least 6 to spread to that sample. 5 or less and it wouldn't spread there. Maybe a few insignificant cases to weed out the weak and unfit, 1% at most maybe depending on how close to the threshold the disease actually comes. 1% infection for 1 less, half for each point lesser. So a health of 6 vs. infectiosness of 3 means .25% (1% for 5, .5% for 4, .25% for 3, .125% for 2, etc). infected, anything 1% or less isn't enough to spread the disease.
Of course if the infection rate is equal to the health level, you get normal infection occur. If it's greater than the health level then you get it double up to a maximum of 99% of the population infected.

------------------
Goes down smooth when I get a clean hit of that skunky, funky, smelly green sh.t
- Cypress Hill

[amjayee]

Guildmaster: That sounded well. Also I think that no single disease could kill the whole population on certain area; there are always some that are immune, or strong enough. Perhaps the health level and the properties of the disease would decide the max amount of people that can get infected (and of which some max level would die, depending on the disease stats).

S.Kroeze: Interesting post, again. Wow! I didn't know that France had more population than Finland has now already 1 AD! We have after all roughly the same area.