Expansion and Settlement in Clash

[Gary Thomas]

But they've taken over civilizations, which I think is good enough for the game.

Not permanently, they haven't. The nearest would be the Turks, and they were not really nomads when they started their expansion.

Cheers

[Mark_Everson]

Plutarck:

I think the basic tone of your post is right, but maybe a bit too detailed in the options. I'd prefer inducements to just be in the form of cash. Its simple to handle, and easy to fine-tune. You can indicate squares or areas and give inducements for up to X people to go there. The cash at low levels would be like propaganda/ info. Higher levels would help to defray the cost of the trip, etc. on up the scale. We can Call it different things in the interface if we like, but I think cash will give a decent simple system in the model. In game terms the cash will both affect the migration decision, and also be at the terminus as an initial stimulus to growth if its larger than the transportation costs.

City-building I'm with you on.

Gary:

A hundred years or so is permanent enough for me. Nothing lasts forever... If you like we can have it so when the "historical" is set to 10 on a 1-10 scale nomads can't be civs. Please don't let too much reality get in the way of the fun!


On nomads, I guess we do need cultural info after all. So one nomad tribe can chase another out, as has generated forced migrations many a time in history. There is also a connection between migrating people as Plutarck is speaking of, and nomunits I think. Since migrating people need to be a package of people that move around, we can use the same game object or at least a related one for civ migrants (that need EG status kept track of anyway) and the nomunits.

[Gary Thomas]

I think that if I am playing the game I do want my civilization to last longer than 100 years.

Cheers

[LDiCesare]

Mark's mechanism of specifying cash incentive is simple enough to model. I agree with it and think it models all the categories of incentives without drowning the player with details.
I think that expansion into uncooperative territory would lead to riots, which could lead to the creation of "rebel" military units, thus modelling invasion without requiring the player to move thousands of armies in order to conquer some space. Think of moving one unit in each of North America square in order to claim it?
I am unsure you can set a cap on migration. People who want to flee will do it, and I can't see why I would like to do that.
How do we model migration of population inside a civ? For instance, Rome didn't grow big just out of Romans. People came form at least all the Latium to grow its population. So we need a population migration model, like each square is attractive to population based on max size sustainable, work available, luxuries, ethnicities there, climate, and whatnot (model that as a value in cash). We consider one's own place, the cost to get from one place to another and, if you're better off moving, there you go. Incidentally, it could help model the migration of people from country to cities or urban regions at modern times, when there was more work available in cities and better transportation than before.

[Mark_Everson]

Hi Laurent:

I am with you on internal migration being eventually desirable. And also agree with all the factors you sited. One disagreement...

Originally posted by LDiCesare
I am unsure you can set a cap on migration. People who want to flee will do it, and I can't see why I would like to do that.

If people from my civ are trying to move away it threatens me as a ruler in my overall economic power, and perhaps also prestige. If they go to a potential future enemy civ there is definitely reason to try to prevent them leaving. Think "Berlin Wall". Making it illegal doesn't make it impossible, but it generally will reduce the flow from what it would otherwise be. I'm not sure how detailed we want to model this, but at least stating a prohibition and assuming state resource enforce it in some way seems desirable to me.

[roquijad]

Great ideas and arguments everywhere...
As I said, my kind of thread!

There're several topics involved here and a it's a little messy. Then, instead of giving more ideas, I'm gonna identify the main topics mentioned already to put some order (am I too naive???)


Topic 1: (unorganized) Nomads
Topic 2: Nomadic Civs
Topic 3: Nomads and Horses
Topic 4: Uncontrolled Migrations
Topic 5: Controlled Migrations
Topic 6: Minor Civs
Topic 7: Troops Recruitment


Topic 1: Nomads
Nomads in Clash should be the equivalent of Sid Meier's barbarians, constantly disturbing normal civs' development through raids. As a plus, our barbarians should have at least, IMO, the following characteristics:
1.1 They don't appear or disappear as in Sid's CIV. They just move constantly, making raids with regularity.
1.2 They can settle. In particular, within your civ. This gives us the potential for challenging scenarios as cultural/religious conflicts arise. (the potential to settle is why I thought nomunits needed EG-type data, Mark)
1.3 They could be "hired" by settled civs as military units.

A first, basic proposal to handle nomads is the nomunit I mentioned in my last post.

Any other fundamental characteristic? Is any of the above unnecesary?


Topic 2: Nomadic Civs
Here the question is if we allow nomads to join together, becoming a more threatening power. That would simulate cases like Attila and the Huns or Gengis Khan and the mongols. If we choose to allow these nomadic civs, then we must decide the modeling depth we want for them. Do nomadic civs have a "tech tree", social classes and internal riots like normal civs do? My position: keep them as simple as possible.


Topic 3: Nomads and Horses
Should nomads play a key role in the game regarding horses as a "weapon"? (of course, given access to horses)


Topic 4: Uncontrolled Migrations
For realism, ideally population should be able to move by themselves w/o player intervention, just looking for better perspectives in life. But it is a big modeling problem, specially in modern times where you have almost the whole world as a possible destination. There're also several variables involved. Africans shouldn't want to go to America if there's slavery there...
I think uncontrolled migrations are great as a game feature, but IMO it's very complicated. I'm not sure we can get a decent model at a reasonable price.

When uncontrolled migrations are a tool of expansion? Europeans migrating freely to the New World helped expand european empires back in the 16th century, but later, when America was free, migrating europeans weren't a tool of expansion. They just "converted" to a new nationality.... how we model that?


Topic 5: Controlled Migrations
As a player you want to influence migrations. Govt control over migrations may take three forms:
5.1 Migration Regulations, describing how easy is to get in or out to/from one civ. This, if we take it, should be something more modern than ancient.
5.2 Colonization Efforts, where the govt invests money to encourage migration in a given direction.
5.3 Forced Migrations, where the ruler picks people from one place and puts them in another.

Just 5.2 and 5.3 are pro-expansionism. Are those two enough tools to expand your civ (by other means than direct military actions against other peoples)?


Topic 6: Minor Civs
A Minor Civ would represent a politically, technologically and economically weak territory. A multitude of small kingdoms or city-states in ancient times and a group of developing countries in modern times. Minor Civs will be there as "scenery" for players, who will be able to trade with them, take advantage of them (protectorates, tributary states, etc) and who knows what else. They could also be a place where a new civ could appear. That would be nice. Not necesarily all (AI) civs start at the beginning of times.
The question is what can they do and what is reserved for "real" civs.


Topic 7: Troops Recruitment
Gary said:

As I said earlier, building units was very rare, raising militia was the commonest way of getting (poor quality) troops. The second commonest was to hire barbarians, nomads for preference. Finally, there were the household troops of the great nobles, who might, sometimes support the High King (Emperor, whatever). Most of the troops that fought against the Huns were Huns...

I strongly agree with you on this, but IMO troops recruitment in general needs a thread of its own. (And I feel we haven't paid enough attention to that part of the game yet.)

[Gary Thomas]

Congratulations roquijad, you just went to the top of my list for posts I most agree with (excluding, naturally, the ones I post myself).

Cheers

[Mark_Everson]

Settlers

I basically agree with the high-level synopsis too. Thanks for doing it Rodrigo. Now on to some specifics. I believe the thing we need first is to enable Expansion.

For now the simplest interface-wise thing we can do for expansion is to basically use settlers per every other game in the genre. So what do our settlers need and how do we build them? Here is my first shot at a simple spec. This could be the sort of default behavior we can use in the first few intro scenarios in the Reworking our Demo Approach thread. Settlers will be a specific instance of Migrants which we'll have later as the game develops.

Settlers are a group of people of one or more EGs (do we need to break it down by social class?) with some level of resources (tools, seed, money, etc.) brought with them. In the long run, the number of people in a settler unit could vary, as well as how much resources they have. This is predicated on a gui for building settlers, and having it be simple enough to not overwhelm people. For now I propose:

Settlers be built in units of 5k people.

Settlers are built just like a military unit. The cost is for sufficient resources to make a settlement on "virgin" territory.

The EG composition of a settler has the same EG proportion as where it is built.

When the settler reaches its destination, hitting "b" or selecting an option on the move menu settles it down. What happens is with the resources and population a new economy is generated that (as best I can manage) takes advantage of local economic sites to make the strongest economy. At low tech levels, most people will be farmers anyway...

There is no limit to how many settlers can be added to a square. When settlers are added to an existing square their resources automatically supplement the farms (farm kapital) etc. in the existing economy.

Refinements and criticisms welcome...

[axi]

I disagree with Mark on Settlers. We don't need them. They are a nuisance and another thing to micromanage. Settling should be handled public works wise, like roads. If the player has knowledge of an uncontested square, he has the option to "plant" a colony there. A colony will be a bundle of people and all types of infraclasses present at the civ, it will be a "flesh out of the flesh" of the city where it's built (the people will be subtracted from the population just like in mil units and the infra will be built) We can thus have various sizes of colonies if we want to or we can plant colony upon colony if we want to strengthen a certain colony. Colonies could even be planted in the midst of barbarian tribes, but they would need protection, or else they would be toast. We can begin with a very simple system and elaborate as much as we want.

This is approximately how Europa Universalis handles this and I like the approach. What is bad about their approach is that there is a chance of colony failure which is really annoying and leads to reloads and micromanagement.

Please consider this option, since migration itself will not eventually be through "units".

[axi]

Though more about it

Let's try to elaborate on what decisions are at stake here:

Things over which we agree:

Settlers (colonies) will be a specific instance of Migrants which we'll have later as the game develops. In the sense that we can in the future have colonies (colonists) built by the private sector as a way to handle spontaneous migration to unowned lands.

Settlers (colonies) be built in units of 5k people. Why not?

The EG composition of a settler (colony) has the same EG proportion as where it is built. Yes and this might be a good way to do some light "ethnic manipulation". It reminds me of Russia and the colonisation of Siberia.

There is no limit to how many settlers (colonies) can be added to a square. Yay!

So my only big difference with Mark is the unit/abstract issue.

Many issues that rise from this:

1. I proposed that all infra that goes with the settlers is proportional to the infra in the square of origin. This means that the cost (and potential) of settlers varies from place to place. People from the biggest city of the civ do not carry with them the same kapital as people from a rural square, since the first carry alot more production capital and less farm kapital than the last. This means that if they are sent in identical squares, they will produce differently and the one who is more suited to the terrain from the beginning will do better. The other has to change his labor allocation and investment, something that will happen in a couple of turns with the current econ model. A smart player would mix colonists from different backgrounds, in order to achieve a better head start. There are countless strategies to be unfolded here and all this because are models are really rich. This could also become a reason for confusion to the novice player, but his problems won't be so dramatic. Also note that historically rulers had alot of direct intervention into the places of colonisation but also into the selection of the colonists. An alternative would be to have a "settler's (or colonist's) backpack" with an amount of infra that depends on tech. I feel that the latter is too arbitrary and rigid in the long run and impractical to code in short term.

2. There is the issue of the class breakdown of colonists. Some would say that historically, only the poor go to the colonies, but they tend to drag along all the other classes (except from the middle class one would say, but they tend to appear quite quickly there too) because they "need" their contributions. Notice the brackets implying that IMO the "need" for judges, policemen, priests, nobles, capitalists and whatnots in a colony is actually imposed by the mother society and it's elites. Heaven denies if we would let our colonists lapse into anarchy and immorality!

3. Another issue is that of time, which is not an issue with the settler solution, but, as with roads, it becomes an issue with the colonies solution. A colony is built instantaneously, or does it take some time? This is important only in the case of a "colonisation race" between rival civs. This could be dealt by pathfinding in the way military units do, which would be done anyway in the settler solution.

4. Would the colonist's voyage cost something more, for the sake of the means of transportation used and the supplies consumed while it lasts? Should the settlers then eat a part of the army supplies? Would this be automatically solved if the settlers were regarded like military units?

5. I proposed that the target square could not be requiredly virgin. Most colonies were founded in places with native populations and some had a dreadful fate exactly because they came into armed conflict with these populations. The problem rises from the existence in the same area of colonists and natives, with different economies. How will the farm and resources sites (which are limited) be partitioned? As we cannot support dual economies now or in the near future, then let's say that in order to colonise a populated square, one has to have control of it, own it. So if you want to colonise an occupied square, take it first and then send the colonists in (like the Israelis do in the West Bank and the Gaza). This also implies that we will have internal colonisation (for the time being controlled by the ruler only). Fast enough we are able to simulate cool stuff like the 1930's forced collectivisation in the Soviet Union. Collectivisation created a redundant rural population which was transferred to the cities to provide labor for the industrialisation. The only issue involved into colonising already occupied territory, is summing up the two groups, which has to be caried out by EG and by economy sector. This is a quite serious part of coding, but it will only be marginally harder than adding the colonists to unoccupied territory, however it could wait for the demo after the one where colonisation is introduced.

6. Could "nomunits" be modelled as settler units, differentiated only by the fact that they can, unlike normal settlers, exploit the land in which they currently are, without having the need to settle?

7. Another issue concerning micromanagement, is what if we want to settle a vast are with a sparse population of settlers. Imagine a map of the 19th century USA where the whole midwest is 30-50 squares. If I want to settle all that area with people from Chicago, will I have to give orders 30-50 times? IMO the setler interface should have a brush tool or window selection tool or whatever lets me sum up a selection of squares that I want to have colonised together from the same origin and compute the minimum number of settlers that I will need, with an option to multiply this with an integer (to send more than one settlers towards the same square) and of course the cost of the whole thing. Otherwise colonisation will be a micromanagement nightmare. Consider that if settlers are used, any such order will create a plethora of units on the map, which will probably devour out computing power and bring confusion to the player. So this is another argument against the settler solution.

From the analysis above, it seems to me now that the only reason NOT to have settler units is the micromanagement. Can this be dealt with realistically? Was I right wrong to ask for an "abstracted" method? However I am 100% certain that I will go nuts if I have to handle dozens of settlers on the map. Automatic movement is a must, if real settlers are used.

Btw, a terminology question. Settlers or colonists? How should we call them?

[Mark_Everson]

Hi Axi:

I had just been thinking of the settlers approach short-term, with something like your colonies for general use longer-term. I think colonies rules per your original suggestion have potential to be a Real challenge to formulate.

I will talk longer-term for a bit. IMO for an automated colonies approach we would need the following:

1. A way of testing the path for hostiles and natural hazards. Teleporting simply won't do for a lot of strategic reasons.

2. If the automated way says "you can't do that" FE due to hostile forces, that is where you would be forced to do it by hand with settler type units or give it up.

3. Ensuring that doing it the settler way and the colony way are virtually identical in final results, otherwise we will give people a reason to micromanage. That includes travel time and any attrition of people or stuff that would be expected to happen along the way. Of course 2 is the exception.

4. I have No idea on a good way to handle peoples sharing economic space. I Desperately want to avoid needing multiple economies. We could use a brainstorm on this one, since peaceful migration Is something that happens historically that I would like Clash to get right.

5. As I think you pointed out nomunits could share code with settlers. Also refugees, if we do them, would Need to be settler-type since they don't know where they're going.

1. I proposed that all infra that goes with the settlers is proportional to the infra in the square of origin.

This may sound great to you, but it sounds like a perfectly awful micromanagement nightmare to me. Are you saying that people from Boston when going West in the States took the same vocations they had in Boston? Practically, those Bostonites would starve almost immediately when they settle down, since the farming percentage of the population is about zilch. But as they expired they could look fashionable with wigs from that wigmakers!

From the analysis above, it seems to me now that the only reason NOT to have settler units is the micromanagement. Can this be dealt with realistically? Was I right wrong to ask for an "abstracted" method? However I am 100% certain that I will go nuts if I have to handle dozens of settlers on the map. Automatic movement is a must, if real settlers are used.

I agree with you here. I think also it is important to have a streamilined approach from a clock-cycles perspective. If potentially thousands of far-moving settler paths had to be recalculated every turn, I think it could be a Major hit to the system. Mil results are worth the clock cycles IMO, but setters, nah! I had pictured something where when the settler is built a path and arrival time are calculated for it. Then it simply appears at the destination on the designated turn. Don't know if we can pull it off...

Just Thinking about the right way to do the colonies that doesn't have any exploits and maintains settler/colony parity has confused my brain. So although we should talk about it, I think any workable colony approach will need to be fairly detailed under-the-hood to avoid exploits, and so is a poor choice for our first shot at this.

[LDiCesare]

I tend to agree with axi. I know that using settlers will be very fast in the coding, so that is a point to consider.
The 2 reasons why settlers are a micromanagement horror is, in my opinion:
1) You have to build a lot of them in order to colonize many squares.
2) Each settler colonizes only one square, thus increasing pathfinding management and increasing the first problem.

To address point 1)
When building settlers, it would be good to be able to say once and for all how many people we want to go. Maybe the RoI can be used to simulate that: Retur on investemnt is high if less than the required number is sent, 0 if all settlers wanted have been sent.
To address point 2)
We could have settler units built, and in the menu instead of "move", have them "colonize" a target square or selection of squares. They needn't be shown, and you don't have to be able to change their orders once they have been sent. The only difference with a regular unit is in the order you give it: Saying "colonize" instead of "move" or "fortify", and maybe selecting more than one square.

An alternative could be to select by a special menu a set of squares as targets for colonization, and when you build settlers they just go there without anyone needing to give orders. Of course, they would have to select the target square or repartition in target squares if they want to split by themselves. That gets rid of the pathfinding part.

[axi]

I should point out one thing: planting colonies is not vastly different than building settlers and it can be coded really quick. There is not long-term vs. short-term solution here. There is a small map vs. big map solution involved. Settlers are a preferrable solution in small maps, where their total number will be small, but they are a real nightmare in big maps. Just like terrain improvements with engineers or trade routes with caravans, or espionage with spies. This kind of logic is outdated in the civ genre and is on the decline. I haven't played civ3 but I bet that workers are a big micromanagement trouble. We should ask Plutarck about it.

The "settler's backpack" is IMO a major p r i c k in our discussion. One can't say that history hasn't seen examples of well equipped colonies (equipped in a way that will optimise production in the target square) and of ill equipped colonies such as the penal colonies or the Clondike. There have been colonies that have been self-sufficient, others that relied heavily on imports and others that virtually starved to death. I insist that i do not want the rigidity of a tech related backpack. Think historically: since the Romans were used to live with aqueducts and arenas, what would they build in the conquered provinciae where they were given land? Aqueducts and arenas of course! Similarily Spaniards would build cathedrals and Americans would build skyskrapers.

So perhaps we should make it proportional to the infrastructure levels of the whole civ, which I hope will be used in other ways too, to make it worthwhile to compute. The infrastructure is not subtracted from the existing one anyway, it is created anew.

P.S. If p r i c k gets censored, how can one have a discussion about roses?

[Mark_Everson]

I don't have any Really good ideas, I'm still fuzzy-headed from this stupid virus but there is one thing I can bring to the party. And Axi, please, be nice to the obscenity police. If you push the limit and someone complains we collectively could get in trouble. That word really wasn't necessary to convey your point, now, was it?

[edit] I realized my retort was "pushing it" every bit as much as Axi!

How about the Default way to settle, whether it be by settler or colony is by Whole Province. If you control land you can just outline a province, place a capital city location (or let the comp pick) and then any people that get to the capital (or are teleported there) automatically are distributed in a reasonable way throughout the province on the next turn. You can use this mechanism for intra-provincial migration too. You can designate settlement projects for one or more provinces at the same time. And of course you can target it by squares, but generally it won't be needed. I am still worried about people appearing all over the place, like in one's own back yard, but I guess a more complex algorithm could take care of that at some point tbd.

Laurent, I think your ideas are reasonable for a start. I don't want to expose people to ROI at the start though, scared away enough already I think. I still think settlers is simpler for the first shot at it. And with settling by province it should streamline things a bit. But we need to see what some others think first!

[Mark_Everson]

One more thought... If we want to do things with a really low-management approach at the start, we can just let people diffuse out on their own as Gary has been advocating. Then the only time you'd use settlers/colonization is if something's really far away, or eventually if you want to shuffle around the ethnic mix. It would be especially low micromanagement. Essentially when a place starts getting crowded the people would build settlers themselves. The challenge is to promote and channel the growth rather than push people yourself. I didn't think of this right away, because the game I really want to play is the one that starts full-o-people, where you have to fight for your existence from day one. That part of the game won't be that important in my kinda game.