Expansion and Settlement in Clash

[LDiCesare]

What about using the smallest X possible? i.e. the smallest number we can use to compute something? That would be the marginal value. I don't know much about Y, but I'd tend to leave with all that I have, thus Y could be the ratio of existing resources/population, multiplied by X. That is probably the maximum value anyway. The problem is this Y depends on the starting location, not on the target, but I can't think of ways around it except to decide of a fixed value, which would be "what you can carry with you", but even that varies with tech (I Can carry more with a truck than with a horse, with a horse than on foot).
I don't think people save up and go. They sell and go. Probably, they save, sell and go, but I think the savings part is little compared to the selling off your house and business part.
How do you plan to compute the number of migrants? The Goths migrated en masse in Europe, mainly because they were being harrassed, but they didn't want to fight their way through the Roman Empire (they had to because of the Romans and Byzantines reaction mostly).
Do you have a formula for the cost of the travel in?

[Mark_Everson]

Hi Laurent:

Just to be clear, my approach was meant to be a very first pass at colonization by diffusion with just demo 7 in mind. I should have stated explicitly that at this point I think we want diffusion to only work for adjacent squares, and that the settlers will just teleport there. Eventually diffusion can be a lot more complicated, but let's not worry about that now.

Originally posted by LDiCesare
What about using the smallest X possible? i.e. the smallest number we can use to compute something? That would be the marginal value.

It would be simpler but inaccurate to use the marginal value to determine colonization attractiveness. Many locations will be great for only a few people, but once there are many people per site, attractiveness will plummet. That is why you need to know How Many colonists there are to make the correct assessment. (And also what capital they bring with them) I will shortly be sending out the code for this, and you can take a look at the economic happiness functions and the effect that population has upon them. For example, a small number of colonists can find broken terrain fairly attractive, whereas a large colony there would starve.

I don't know much about Y, but I'd tend to leave with all that I have, thus Y could be the ratio of existing resources/population, multiplied by X. That is probably the maximum value anyway. The problem is this Y depends on the starting location, not on the target, but I can't think of ways around it except to decide of a fixed value, which would be "what you can carry with you", but even that varies with tech (I Can carry more with a truck than with a horse, with a horse than on foot).
I don't think people save up and go. They sell and go. Probably, they save, sell and go, but I think the savings part is little compared to the selling off your house and business part.

I don't know the best way to do this short-term. Longer-term we could get fairly complicated. I suggest for now that we just assume all capital is portable, and that any settlers take their fraction of the capital with them. This is obviously very crude and will need to be changed at some point, but I think it will get the right flavor for now.

How do you plan to compute the number of migrants? The Goths migrated en masse in Europe, mainly because they were being harrassed, but they didn't want to fight their way through the Roman Empire (they had to because of the Romans and Byzantines reaction mostly).
Do you have a formula for the cost of the travel in?

For now we are just handling economic-driven migration by these methods. The number of migrants should be small, at most 10% of the originating square. Perhaps 5% is even better.

For the player to do "spot colonization" I'd say for now let's use a settler unit. I'm not sure that Gary has it set up to be able to build a settler yet, what do you say Gary? I can do the economy side of it fairly easily. Eventually most of the "spot colonization" should be doable without settlers, but the advantage of the settler is that we have that now whereas we don't have the interface for the player to influence migration available yet. Since I'm assuming all migrations are at most a diagonal move away, there is no travel cost factored in yet on my proposal.

[Simon Loverix]

Whatever way migrations will be handled, escorted settlers are especially fine for nomadic invasions. The nomad's advantages would be speed, no support required and a greater part of the unit (most males, no?) consists out of better (cavalry) military elements, as opposed to settler elements. The disadvantages: no accumulation of infrastructure, wealth or knowledge.

This approach would then work both for nomadic migration and governemental colonization/deportation. Other migration is diffused, only with bigger numbers.

This way

[LDiCesare]

What about sea squares?

Playing demo7 to test it, I discovered that sea squares can be colonized. They will be settled as of the current code. This seems a bit weird. I think coastal squares could be colonized, considering people live on the coast and fish there, but it feels a bit strange. Although it allows people to cross a 1-2 squares channel in addition to colonizing the sea. How is sea fishing modelled in the game? Probably putting 0 farm/resource to sea squares would prevent people from getting there.
Opinions? Idea about sea exploitation?

[Mark_Everson]

Re: What about sea squares?

Originally posted by LDiCesare
Playing demo7 to test it, I discovered that sea squares can be colonized. They will be settled as of the current code. This seems a bit weird.

OOps! Have to fix that...

How is sea fishing modelled in the game?
Opinions? Idea about sea exploitation?

I haven't gotten very far in modeling fishing. The basic idea is that fishing would be done using fish resources that are out at sea and can be gathered. There would be an infrastructure class in any coastal square that would represent a fishing fleet. The fishing fleet infra would allow harvesting of fish resources offshore. At a guess, a given amount of fleet would have so many harvesting points that would be used to move to the resource and harvest it. For far-off resources, less would be able to be harvested since more time would be spent moving to it.

How exactly to do this with the existing code is a TBD. I'll put this in the econ thread too, since any further discussion of this should happen there.

[gonzo_for_civ]

Hmm.. It's been a while since I last posted here. I was just playing and I realized that you can colonize sea squares. I don't know if this has been mentioned because 5 pages is a lot to read

[Mark_Everson]

Hi Gonzo!

Thanks for trying to help, but if you look in the post immediately above yours, you'll see the issue had already come up .

[gonzo_for_civ]

Doh!! Well, I'm off to play some clash. Like I said, I'm too lazy to read 5 whole pages <>

[Mark_Everson]

From the features list, a point relevant to this thread:

F166 (Richard) I have discovered an abusive strategy. It allows one small army to plunder an entire enemy civ. What I'll do is march one warrior into the outskirts of their civ, claiming the square. Then I have the warrior pick up all the people and move to another square. I then pick up all the people in the second square. Since the first square is empty, they will send about a thousand people to colonize it. I then march my warrior back to the first square and pick up those new colonists, leaving the empty second square which will also be colonized again.

This can continue until their population is too low to colonize anymore. At the end of this process, their entire civilization had about 4000 people and my lone warrior was escorting over 11,000 of their people. They had a big task force on their capital, but those armies just sat back and watch me steal all of their people.

For the short term we can just tell players "Don't do that!". But we need a better fix long-term. Moving with settlers should slow down a unit, but that's not enough to shut down this strategy. Perhaps military units can only escort a certain number of hostile people, and after that threshold is reached the people would be presumed to be able to escape? I don't want to make things too realistic, but presumably some simple and sensible rules should be able to stop this. Ideas?

Often, they would colonize a square that my armies were actually standing on. That seems wrong.

We can just prohibit this.

[LDiCesare]

Perhaps military units can only escort a certain number of hostile people, and after that threshold is reached the people would be presumed to be able to escape

Sounds good. Another option is to allow unwilling people to riot against their "shepherds". The code for militia could be used here too, so people would revolt if you tried to escort them with not enough guards. It could be fun if you attack an escort, get enough of them killed so the escorted settlers revolt and get rid of their chains.

[Mark_Everson]

Originally posted by LDiCesare
Another option is to allow unwilling people to riot against their "shepherds". The code for militia could be used here too, so people would revolt if you tried to escort them with not enough guards. It could be fun if you attack an escort, get enough of them killed so the escorted settlers revolt and get rid of their chains.

Sounds reasonable to me! My only objection is that this sort of thing might promote micromanagement, and be a bit hard for the AI to handle. But I'm not sure we can gauge that before actually trying it. Hopefully we'll get some opinions from others as to whether this will work or not.

[ogj20]

I think the root of the problem here is the way conquest is handled. With the system as it is, it is possible for a warrior band of 500 people to march halfway across the world, converting tens of thousands of opposing people with each step (admittedly 'unarmed' people (pitchforks anyone?)) and not loose a single warrior in the process.

First off let me say that I understand that no-one thinks this is the way it should be eventually. Things like the idea of militia springing up to defend against invaders, the riot model causing unrest in captured lands, and ideas about troops suffering attrition if they are in out-of-supply squares will reduce this.

But, It still seems silly that e.g. 500 man russian army could spend a single turn (even if that is ~4years or whatever) and convert potentially hundreds of thousands of people flag waving americans into book-brandishing russians of an american ethnicity! Who here thinks that as soon as the warrior unit left the square would just revert straight back to being american? Or that the warriors would suffer significant casualties trying to enforce the change of ownership whilst they were there?

I suggest that owned squares should have a loyalty/security value. If everything is normal, It should be 1.

Each turn it would be modified according to:

1) decreased by an amount proportional to the sum of the enemy (Civ's they are at war with) population in the neighbouring 8 squares

2) Increased by an amount proportional to the sum of the neighbouring friendly/allied pop

There should be a multiplier factor for the population of a neighbouring city/fortified town whatever, and a much larger multiplier for influence of the capital of a province.

3) Decreased by an amount proportional to the quantity of enemy military in neighbouring squares - danger of invasion

and conversely

4) increased by an amount proportionl to the amount of their military in neighbouring squares - Border patrols.

Then if an enemy unit actually entered the square it would cause a fairly massive drop, proportional to the enemy military unit's power compared to the population of the square (i.e. 500 warriors terrify 1000 poulation, worry 10000 and get laughed at by the local gun clubs of a square with a million pop)

Military units could either just sit in a square, holed up, worrying people, or they could spend a turn actively going out and supressing resistance, and 'educating' the populace in realpolitik with their boot heels. Actively conquoring a square would run the risk of loosing troops though, again based on the size of the local population compared to the strength of the army unit.

When the loyality/security of a square dropped below some value (zero?) it would become contested land, dropping out of the control of it's owning player, when it dropped below (-0.5?) it would look around for a new owner, based on the nearest Civs, their power and the social factors from the social model. Note that though having troops in the square at the time would give a Civ a big advantage here, i don't think it should necessarily guarantee they got the ownership of the square, I'd like it to be possible for invaded lands to try and join a thirdparty Civ if it was very attractive to them and they felt it could offer them defense.

This would mean that to capture populous enemy lands you'd have to send a reasonably large force, and spend some time on it, but also that it might be possible for squares to be intimidated into switching sides if there were masses of unopposed troops stationed just across the border.

it would make isolated, almost surrounded squares easier to capture than secure squares in the centre of enemy provinces.

Of course if a province capital is captured all the squares in the province should suffer a heafty loss of security, meaning that all of them on the border will probably surrender (switch sides), those near the border with still standing provinces will join them for security, and the rest will run around like headless chickens, waiting to either be caputured by the enemy, reinforced by their owner, or engulfed by the wave of changes of allegiance that will ripple out from the newly caputured capital and the borders.

This would also provide a method for provinces to dynamically change their borders, favouring the stronger better defended provinces.

And as to why this would solve the population collecting 'feature'. You should only be allowed to pick up population from a square you control, and you no longer get control of a square immediately, so if you want to grab all the pop you have to hang around long enough to 'convince' the people to obey, making yourself a sitting duck.

It also neatly removes the micromanagement involved in recapturing squares in the path an invading army leaves behind them, and allows a way for Peaceful Civ to expand territorially, by slowly intimidating and outnumbering their neighbours until they switch sides, but never actually attacking.

It would of course also naturally lead to a need to garison borders, which seems sensible.

Anyway, what do people think? Do you agree, do you disagree but think there are apects of this which could be interesting to develop further, or do you think it's completely the wrong way to go?

[Mark_Everson]

Hi ogj20:

I have to run in a few minutes, but I thought I could give you some quick feedback. As you say we're aware of the current problems, and have some plans to fix them.

I think your specific solutions to our current issues are quite reasonable, and I like the way your proposal takes care of most of the issues we have cropping up in D7.1. I personally would like to see your idea fleshed out with more details if you're game. Please check out the Govt/Social/Riots models in a little detail first, so you know how to put some of these things into the game terms and values we're using.

Originally posted by ogj20
Military units could either just sit in a square, holed up, worrying people, or they could spend a turn actively going out and supressing resistance, and 'educating' the populace in realpolitik with their boot heels. Actively conquoring a square would run the risk of loosing troops though, again based on the size of the local population compared to the strength of the army unit.

This seems to have danger of micromanagement (MM)

This would also provide a method for provinces to dynamically change their borders, favouring the stronger better defended provinces.

Not sure I like this one, if the player sets borders they should stay that way

Anyway, what do people think? Do you agree, do you disagree but think there are apects of this which could be interesting to develop further, or do you think it's completely the wrong way to go?

Aside from where I've noted I think the ideas are pretty good, and worth pursuing. Maybe we could even get them in soon.(We'd need some small changes to the Govt/Social model to indicate affinity of EGs for different civs).

What do others think?

[LDiCesare]

The idea looks good overall, except I ant to point out a few things:
Conquistadors did conquer huge lands with pitifully small armies.
I don't want to have to say to an army how it handles the population everytime it goes in a square (MM), so I want it to be automatic: I'd like to use diplomatic info (at peace/cease-fire vs. at war) regarding the owning civ in order to know if the troops try to conquer or not. Being at war, they automatically try to conquer (and meet resistance), but that will not cause a change of ownership unless the army remains for some turns.
The system must be careful about huge armies: they'd conquer everything, but as soon as they leave, the population should switch back to their previous civ.

[ogj20]

To stop huge armies conquering everything in sight, you could put a cap on how far the loyalty value will move in a turn, meaning that even HUGE armies will have to spend a bit of time convincing the people to switch sides, just a lot less than small ones.

As I recall a large part of the reason the conquistadors were so succesful in conquering large huge lands with small armies was partially due to a significant technology edge (powerful horses, gunpowder etc.), which would be dealt with because they'd be quite powerful even if numerically small units, and I'd evisaged the power of a unit being what decided how good it was at convincing the conquered square to change side, rather than its pure manpower.

In addition it's my understanding that an even more significant factor was the effect of the diseases they brought with them running ahead of them and devastating the population ahead of them, leaving them with in fact relatively few people to conquer, an effect which the disease model will presumably create for us.

In addition, how 'conquered' were the people they conquered, or more importantly how 'assimilated' were they. If the spanish had just gone home and let up the military pressure, I think it's fairly likely that the most of the 'conquered' people would have reverted back to their previous civilizations failry quickly, rather than continuing as a far flung colony of the spanish.

I've also thought that perhaps the loyalty value of a square could be used to modify things like the tax revenue you get from it, making shaky borderland territory less productive than your well defended heartlands?

I agree that some thought will need to be put in to avoid encouraging excesive MM, since players may be tempted to split TF's up to hold down large swathes of territory, though I think if the scaling of attrition and the effect military has on loyalty were set right, this could fairly easily be made unnecessary or unattractive.

I agree that the system should be automatic, and that units should only try and conquer if you are at war with that Civ, and also if they are unlikely to suffer crippling loses in the process, other wise they should just camp out and glare at the opposition, which presumably leads you off into the seige model somewhat.

Also I agree that there will need to be links with the work in the Gov/social./Riots models, both for the effects on loyalty of the various elements, and for the process of deciding which Civ to join when a decision is needed, I'll finish looking at them and see if I can come up with a more fleshed out idea.