DESIGN: Modeling population

[Keygen]

Yes, simplicity is surely a key to any game, especially to new players, I agree. But keeping the game simple, in this particular genre, will not allow it to evolve much. The solution in conceptual level is very simple. Scale complexity depending in difficulty level. The higher level you play the more complex micromanagement you'll have to deal with. And if you find that model not what you'd wish for then we can make an additional add-on/expansion wich will add more advanced features.

[Mihai]

You can steal the idea from "Clash of Civilization"
Population is in every square and when grow above a certain limit (I guess that is related with what kind of land is ), a part of that population migrate to an adiacent square. Their idea of "region" can be used in CtP as "city and the surronding rural area".
That would mean that the rural population is more vulnerable to enemy armies - maybe even siple pass of an enemy army can result in population decrease. Only the population in the city will benefit from some facilities (city improvements) as city wall. Some city improvements will allow the population of city proper to grow (i.e. aqueduct).

It is quite complex, but it could be interesting...

[Gen.Dragolen]

Keygen,

Having different levels of difficulty with matching complexity would make it a very challenging project to code. That is why most games rely on haing the same code model with the AI getting different bonuses or the player penalties to make the difficulty harder.

The core of it is how much micro-management is reasonable and still have the game play enjoyable. The population model in 'General' is fairly simple, yet easy to learn to manage. For me it was always a problem of running out of bodies when multiple opponents would attack each turn.

Life is complicated and everything is inter-related in some way. To model this for CtP, you have to expect a more complicated design in order to simplify the game play. To that end, demographics come into the core of the model.

To put in my 2 cent's worth for a more complicated model, look at what the ancient Egyptians were facing:
- the Nile river valley flooded every year for several months
- the flood plains were incredibly productive for agriculture
- the flood plains were also prone to diseases born by parasites and insects
- the Temples and the Pharaoh levied taxes on the villages that were usually paid with labour
- a small portion of the population were skilled artizans that often lived at work sites like the Pyramids at Giza for years at a time, and travelled to other project sites
- trade took place along the river
- average life expectancy was around 40 years, often shortened by diseases
- hard construction labour would shorten life expectancy
- basic medical care looked after injured labourers on the big projects
- economy was based on human power, including slaves

So for the average Egyptians they would have a life span extended by:
- plentiful food supply, rich in protien from cattle and plentiful from multiple grain crops in the richest soils
- the medical treatment they could receive at the Temples including follow up care for the injured

Their lifespan would be shortened by:
- numerous diseases that spread easily and quickly because of the river
- hard labour either in daily living or on construction work

So you have natural environment contributing, technology, social organization, and food supply moderating the population levels. So right there, that's 4 variables that we would need in the model, and they are all subject to change from weather, politics and natural disasters. Now that's 7 variables. For a programmer to work with that many variables could be a daunting task, especially if you want to have them only apply for a certain game difficulty level.

I think that the more complicated the modelling, with the more controls the player has on the variables we can control (society, technology, improvements, tax rate, etc...) the easier it will be to manage since it will reflect more of what we do each day in our own lives.

(Not a bad bargain for only 2 cents, eh ?)


D.

[Gilgamensch]

If you call this 2 cents, I would like to get a full euro/dollar

But managing all this would be to much MM (micromanagement) for me.

But similar things have been done in other games. But they weren't this kind of strategy-games.

But we might find a way to get different levels.

[Gen.Dragolen]

Gilgamesh,

If the micromanagement is too much for the average player who just wants to concentrate on building or warfare, then that's where the advisors should be acting for you.

Since you, as the leader of the civilization, cannot be everywhere at once, having satraps to do the leg work would be nice. Not to mention get rid of those annoying mayors/governors. What I would like to see is an advisor who comes up with reports based on how often and on what conditions you set in the preferences.

If the advisors can look after things like connecting all cities with roads, then it can ask how long you want it to take and how many workers it would be allowed to use. This could work for population management as well:
- have the workers able to be assigned to an advisor for lets say agriculture who you then task to make agricultural tile improvements for a given city or a group of cities
- have the advisors make decisions to affect happiness and productivity and have them make reports on the impacts
- difficulty can be set by modifying efficiencies of the government, technology, and terrain types, and any other population details you care to add the to model

And yes, Gil, when I'm selling ideas, my 2 cents worth is often more like a full euro's worth.

The nice part about emphasizing details in the modelling of the population and all, is that if you pursue a certain individualistic policy, then you should have a radically different out come from anyone else playing the same way. Part of what always annoyed me was that eventually the game broke down into a race to a certain technology or unit and then curb-stomp the neighbourhood...

Gets kinda boring after a few dozen games. What I always hoped for was if I took a certain path, and went a different route at a certain point when history showed what happened already, and then see if I could have made a difference.

Is anyone working on the model for the programmers yet ?


D.

[Gilgamensch]

If the micromanagement is too much for the average player who just wants to concentrate on building or warfare, then that's where the advisors should be acting for you.

Since you, as the leader of the civilization, cannot be everywhere at once, having satraps to do the leg work would be nice. Not to mention get rid of those annoying mayors/governors. What I would like to see is an advisor who comes up with reports based on how often and on what conditions you set in the preferences.

You are saying that we shall use advisors, but don't like governours They shall do the same job.

Before you bash me: I see your difference, you would like to see the advisors acting more on a global perspective, which would be better, me thinks. But the problem there would be (for example) unit-production: It is also a question of where to build them. And this would be depending on the big picture, which you need to decide upon. Like you have a high production city on the edge of your empire, most likely the AI/governour would use it for the most expensive unit (catapult) instead of a more faster unit (knight). And I must admit: I don't see any way how to teach the AI this.

If units would be constructed virtually and you could place them in non-besieged cities, it wouldn't be a problem. (like the 5 Star series of SSI) You could just 'say':

I want 5 catapults and 5 knights and 5 legions.

The advisor could then show up with:

5 catapults will take 22 turns
5 knights will take 25 turns
5 legions will take 12 turns

(figures a pulled out of the air )

And with setting priorities, you could alter it.

This brings me to another idea, but we (the programmers) would have to alter quite a bit:

Instead of having cities producing units, why not change it all complete.

Like PW, we could set a percentage for military production. In a submenu you could choose to 'buy' whichever unit you want. Production time would depend on the cost of the unit divided by the average production of your cities.

The unit placement can only be done in cities with capitol or barracks.

[Gen.Dragolen]

Gil,

"... the problem there would be (for example) unit-production.."

Now why would I ever trash you for bringing excellent points to the discussion ?

I can see what you mean about teaching the AI to build units in an somewhat strategic manner is a big challenge. But you set out a couple of good ideas to help it along: the need for barracks and city status.

Since the scale of the map makes some of the situations we are used to thinking about on a smaller local scale seem ridiculous on a larger scale, for example besieging a city, we have to make some choices about how to represent them. SHould the city have a 'Besieged' status ?

I recall reading about how many months some castles and fortified cities could hold out when a large force showed up to lay siege. Depending on how much grain and potable water was available, the siege could last for a few months to a few years.

Again, the level of technology, the government type, the improvements and the morale of the populace usually decided the out come up until an army showed up with a siege train. Then the level of fortification became the deciding factor: catapults, ramps, under-mining towers or walls, etc... vs. the walls, the terrain, and the number and morale of the defenders.

Add in the effect on agriculture of an invading army (and any forces sent in relief of the siege) and the country side will be stripped bare of food. In the Dark Ages, the French coastal areas from Normandy to Belgium was usually in a state of famine. Not from the Viking raids pillaging the country side, but from the French armies sent to deal with the Vikings eating all the food while in the field.

So how do we integrate all of this detail and still make it managable ? Thank the mathematicians for the algorithm.

If we have the populace's morale/health is factored in to production, we can have unit production governed by many of the efficiencies we would have to manage with keeping the people fed and happy. Troop quality and quantity are going to be limitted by how many people are there to begin with: if you have a small population base, then pulling large numbers of the fittest men to form a combat unit is going to have a negative impact on the efficiency of future production.

What this leads to is a limit on how many units of a given type that can be produced in a given city before there is a lack of bodies to put in the barracks. With the longer turns in the ancient eras, this is not such a big deal, but as the turn period shortens to a year, then the impact is on the society is going to be huge.

Just look at the changes we saw from WWI and WW II in western societies. Women gained the right to vote, Income Tax became a primary revenue source for most governments, and armies began to have hundreds of thousands to millions of people in them. Casualties will weigh heavy on any society if war lasts very long.

This is one of the reasons I wanted you to look at General to see just how tricky it can be to manage a population like that.

So this brings us to what to do with the existing CtP model. If we are going to have some automated management, just how much do we want to let it handle. Like I said, there should be some serious divergence from the norm if you get to make some small changes to the model once you get to having a society that has existed for a couple of thousand years.

Any suggestions about which variables we want to have the player control and what should be fully automatic ?


D.

[Mihai]

I don't know if this would be the place for that...
Cities under siege - they shouldn't be able to gather all the resurces from their area. That was somehow done in Civ2 with the imposibility of harvest a square ocuped by and enemy unit. I say to go further and make imposible to harvest a square if there is an enemy unit in it or between it and city. Maybe squares adiacent to enemy unit give just half as usual. This way a city surrounded buy 4 enemy units can be really under siege, harvesting nothing. The city will consume the food in granary and then starve.

[Matthew Hayden]

Perhaps this could work

I can't to be discontinuous, but I'd like to suggest another production principle.

1 - Every unit uses one pop-point.

2 - Every pop-point ALWAYS represents one-thousand people.

3 - Advanced farms astronomically increase food output of a tile (farming would not double the effectiveness of tile, it would increase dozens of times compared to an uncultivated tile).

4 - Public works is a luxury a player has little acces to in any large amount until, say, the civ has at least one major city (or some historically appropriate condition).



In point 4 I refer to public works being set at a maximun of just 10%, or only actually being enabled by an advance like engineering, and then you are limited by the simple lack of production that comes in, since the basic mine would only boost tile output by about ten or fifteen (can't remember what exactly it is offhand).

I'm trying to figure this out as I go along, and it strikes me as much simpler than what I've read already, at least in my mind, but of course I'm not the one who's trying to implement the changes to the source-code.

Best of luck however it goes!

[Ekmek]

Re: Perhaps this could work

Originally posted by Matthew Hayden


2 - Every pop-point ALWAYS represents one-thousand people.

I can live with this simplicity. But I'd also like to see with units that you see how large it is. Like if it loses strength in battle, I would like to see it down to 700 troops. Maybe later we could calculate the amount of casualties from a battle and a war. Then as a nice touch maybe as a war progresses war discontent is linked to casualties and not just duration. Of course TV could increase discontent...

All of this could be seen in a military advisor tab tracking a war from declaration to peacetreaty (and being able to name the war)...

just random thoughts hit me on the possibilities for the pop modelling...