First let me say that I think the social model is admirable in its own right. I have not yet grasped the implications of the combination with the other models, and most of the discussion in this thread is above me.

However I have spotted some potential problems, more basic than whether religion should be modelled one way or the other:

1. Civ - EG - Nationality
I think there will be problems with the two-tiered model of Civ and Ethnic group. Example: Libyans (an independent EG) are placed between Roman Empire in Tunisia and Egyptian Empire in Egypt. Romans and Egyptians divide the entire Libyan territory between them. The Libyans living in western Lydia will become an EG at the Roman civ level LybRom), while the Lybians living in Eastern Lybia become an EG at the Egyptian civ level (LybEgy). The Lybian civ has effectively been split into two unrelated EGs. This can be handled, but then the Roman Empire expands into the Egyptian part of Lybia: LybEgy is now split further into LybEgy and LybEgyRom. Jugurtha comes out of the Saharan dessert and conquers the southern part of the whole of Lybia. We now have LybRom, LybEgy, LybEgyRom, LybRomJug, LybEgyJug and LybEgyRomJug - and so forth ad nauseam. These mechanisms will take place whenever a square changes hands and will result in a huge amount of infinitesimal EGs over time.

What is needed is a way to combine these very related EGs into one EG at the civ level as before. To this end I propose an OO class clsNation in addition to the proposed clsCiv and clsEG (this was sort of hinted at in the social model since nationality was one of the characteristics of EG, while the implications of this was not described). An objEG would retain its present civ level interpretation as all people in an objCiv that perceive themselves as being of a certain nationality (objNation). If an objCiv expands to include people from another objCiv but with nationality of an objEG already present, these people will be incorporated seamlessly into a joint objEG: The attributes of the new combined objEG is calculated as a weighted (based on population) average of the attributes of the original EG and the newcomers. In the most basic form clsNation need not contain anything more than a name, and probably a dynamic array pointing to all EGs of this nationality. If desired this concept can later be expanded, giving the objNation attributes, representing the proto-ethnic origin of the nation. These attributes might change over time, so as not to be too distant from the constituent EGs. However I am not sure if this would add anything to the gameplay.

The effect of this would be that a nation is divided into fractions along civ-borders, different, but still sufficiently related to be integrated into eachother should the territory they live in change hands. On average each nation would be divided into no more than 2-3 fractions (an independent and one or two civs). Each fraction would change their cultural attributes over time depending on their Governmental and Religious environment, but still they would be very similar (IRL: talking the same language, having the same mythological material etc). If desired it would be possible to model the splitting of a nation into two different nations, if the individual constituents became sufficiently different (maybe by the root-mean-square approach used to calculate religion-culture 'Distance').

The 'How many religions'-issue could be handled in a number of ways:
One nation - one GWR: The first GWR-convert of a nation decides what is possible in the entire nation. This is unrealistic, but simple to model.
One EG - one GWR: The first GWR-convert of an EG decides what is possible for the entire EG of the civ. This is more realistic, but will lead to instant 'mass-conversions' if a civ conquers people EG with a different GWR from another civ. This is a consequence of the all-or-none mechanics of the concept religion. You can average aggressiveness 35 and aggressiveness 65 to get aggressiveness 50 (assuming equal populations), but you cannot average 35% Christianity and 65% Islam. One method overcoming this would be to split the nation the instant two EGs of the same nationality encompass different GWRs, in fact degenerating the model into something like One nation-one GWR.
One EG - any number of GWR: This is very far from the original social model, and the question is: Will it improve gameplay, or just add unnecessary detail. To be really realistic, this would mean probably tracking proportional GWR-allegiance on a province- if not square-basis. Maybe this is not a problem, but is it really necessary?

2. What is a civilisation
I think the origin of this problem is in the very title of the project: Clash of civilisations. The problem is that 'civilisation' in a sense is ambiguous: What is a civ. Is it an entity with a central government - a civ/empire? Rodrigo seems to indicate this in his Social model where there is no clear sundering between civ, government, empire and controlled provinces: They all signify that the player controls a centralised governmental structure with a geographical entity. It is however very clear that an EG is something different, similar to a nation or a tribe. FE 'If Romans (empire I presume) take Egypt (province IP) then Egyptians (EG IP) end up living under Roman rule (governed by the Roman Empire IP) and they wont necessarily become Roman citizens and support the new regime (they remain being in the EG Egyptians IP). Even more, Rome (Empire IP) can invade just part of Egypt (province IP) so you can find Egyptians (EG IP) in roman provinces (Provinces controlled by Roman Empire, but not necessarily inhabited by EG Romans) and in Egyptian civ (Empire IP)'.

However in the thread Who is the player (good question BTW) it was suggested that a civ is something different from an empire, more like a major, more or less homogenous cultural phenomenon. In a way this makes sense, since Roman civ/culture outlasted the Roman civ/empire by centuries - in a way most of the western world - and probably to a degree the near Orient as well - still live in a 'Roman' civ/culture. It is also true that the geographical extent of a civ/empire rarely coincides with the area under a certain civ/culture. I guess a point could be made claiming that the Roman Civ/Empire really consisted of at least two different civ/cultures: the Greek and the Roman. However, if it is done this way it would probably mean that we had to create a special OO class called clsCivCult interacting with the classes clsEG and clsCivEmp. One CivEmp is composed of several EGs, of which one might be the ruling EG; as above one EG is only part of one CivEmp. One CivCult is composed of several Nations, and like with religion there would be several opportunities for how many CivCults are encompassed by one nation, although I would recommend the one nation - one CivCult approach in this case. This objCivCult approach is so similar in game terms to a clsReligion, that a separate modelling of clsCivCult would IMHO be superfluous. (IRL civilisations and religions have been walking hand-in-hand to a degree where the two are almost indistinguishable)

So what are your thoughts on this: Will there be both civ/empire and civ/cult or just the former?

3. Civs - barbarians
This might not be the right place to comment on this, but I am on a streek, so here goes:
Playing demo 4 it was obvious that the 'bad guys' came in two different flavours: a) Rival civs with mobile units and b) barbarians with limited military power (except to conduct a casual raid), primarily there to be incorporated into your empire before the other civs get around to it.

Will this be the same in the game? I realise that computational resources might limit the number of competing AIs available. But still it would be nice if you never really knew if the people of occitania just might rise to be a powerful civ.

4. General advice
This leads to some advice: To avoid confusion, try to define your game-entities very carefully. Preferably give them a unique name or an abbreviation instead of using a common, but ambiguous term like civilisation. Somewhat meaningful abbreviations would also help the reader, and probably the programmer as well. This makes it a lot easier to spot inconsistencies in the models. I prefer EthTol to ethnic tolerance or ET, RelTol to religious tolerance or RT, Aggres to aggressiveness, Trad to traditionalism etc. In this way it is clear when you are talking about a game term, and when you are talking about something general.

5. GWRs - judaism - others

It is true that Judaism may not have the attributes of a GWR in the sense of being universal and spreading to other nations to any major extent. However, some spreading did occur: The Kazar khanate, a Turkish empire in S Russia, E Ukraine, Caucasus, and bordering regions of central Asia, adopted Judaism to a rather large extent. Kagan (king) Bulan was converted in 821 and for the following centuries, this Turkish tribe lived under a religion of traditional rabbinical Judaism, being judged by the Torah. Had the Turkish kingdom along the southern end of the Volga not succumbed to the Scandinavian derived Rus, the history of Judaism might have been very different. I would therefore make an argument that Judaism is a potential GWR. Another potential GWR is Zoroasthrianism.
I think that the splitting of GWRs could in certain scenarios add to gameplay, but I would prefer to keep it simple

1. In general...
Soon after I released the social model, F_Smith joined actively the team and decided to code this model. The "Beast" (you asked about this in the other thread) is what F_Smith has created so we can all participate in the creation process. There's a thread in the forum called "Object Builder" or something like that. In the very first post there's a link to a page where you can see what's coded so far and you can *play* with it and see what happens. The social and govt models are being both coded by F_Smith in this "Beast".

When F_Smith started to code the social model, he suggested some changes to the social model. The model is in essence still the same, but a couple of things were changed. Most notable, ethnic groups are no longer managed at the civ level, but at the mapsquare level. This adds a lot of realism at the cost of using more computational resources, but F_Smith has convinced us so far that the cost wouldn't be too high.
The other change is the concepts of "Primitive Religion" and GRW have kind of disappear and mixed in a simple "religion" thing. This change in particular is not entirely solved in its details. This change was born also to achieve more flexibility.

The other change concerns "similar" religions. In this thread it was proposed by LGJ and others that a religion could suffer division and create branches, like catholicism giving birth to protestantism and the greek church. So far, this is an idea I commited to think about how to implement, but I haven't done anything yet.

Finally, Mark suggested a richer modeling for the very important Nationalism attribute. I haven't started to think about this yet, but some changes will be there sometime.

2. Civ-EG-Nationality
What I originally planned was that if a civ/empire expands in such way that conquers mapsquares with a new nationality, an ethnic group object is created in conquering civ OO-structure. But any other conquest of the same people would only increase the EG population and change its attributes according to some rule like averaging them. A civ would have then as many EGs as nationalities living within it exist and then there's no "dangerous" multiplication of ethnic groups.

In the real implementation, however, since ethnic groups are located (in code terms) in mapsquares, every time a mapsquare changes hands no EG objects are created or deleted from any civ and no averaging is necessary. In this regard F_Smith's approach is much better.

Either way, there's no problem with territories changing hands. It's all under control.
I've to confess, though, that in F_Smith's implementation there's still a part that confuses me. He has, as far as I understand, separate objects for ethnic groups and cultures. So an EG object has a culture object. I don't know what use this division serves. To me an EG *is* a set of cultural attributes. Having them separated may cause some problems when territories change hands... but I'm not sure. I'm maybe simply understanding wrong the implementation.

I disagree with your approach of having a "meta-EG" or something like that (Nation, as you call it) to which all EGs are related someway or another regardless of where they live. You're right when you say lybians will preserve their culture even if splitted in several empires, but as you also mention, after some time it's possible for these to split for good (a nation produces two nations) or to "forget" their national cohesion. I believe is much better keeping it how we have it. For some time lybians will have the same culture (same cultural attributes) regardless of who's controlling them just because variable "Traditionalism" won't admit any sudden change. But if the situation lasts, it is realistic to expect each lybian EG will evolve differently depending on where they live and a reference to a "meta-lybian-nationality" seems strange and "magical". Magical because the only real way for people to preserve their lybian ways is in reference to their tradition and not to some abstract and independent-from-themselves thing that seems to exist in the vacuum.

3. What's a civ?
An excellent point. I should have written a paragraph about it in the model just to make sure implications of EG modeling are understood.

In fact, once we accept we want a "civ" to be able to hold several nationalities as it happened commonly in history, we must deal with the question "what a civ is?". And as you say this is pretty much related to who the player is. My opinion, as you discovered it, is that a civ is a govt and the player is the ruler of that govt. This govt, in time, controls land where many "tribes" (ethnic groups) exist. Usually, only one tribe considers this govt/empire as "its own". That's why I've said a civ (govt) has a nationality. The roman empire was "roman" (or latin, if preferred).

As philosophical this matter might sound, it has important implications. In civ2 this was solved simply not having ethnic groups with a culture and sense of nationalism. There's only "people". As a result, if you invade other civ's posessions, people don't react. Germans can invade France and french people don't give a damn.

Modeling tribes and govts as two different things as opposed to the vague "civ" term in which any of the former concepts may lie easily, is what adds realism and historical accuracy in a number of fields. Since this realism is one of the top priorities in Clash, we need things divided like that. We need tribes/cultures/ethnic_groups and we need govts/empires. Separate.

4. Barbarians
As I said in the other post, barbarians will be nomadic civs with the ability to settle and successfuly become a power.

5. Religions
Apparently, we'll just have "religions". Not GRW, not "Primitive Religions". PRs, btw, are now called Ethnic Religions to avoid any misconception or prejudice. In this perspective I think the problems with Judaism could be solved.

However, what is most important for me is what you've pointed out regarding multiple religions. Most of Clash team members would like to see each ethnic group able to follow more than one religion as opposed to what I originally planned. The best way to do this in order to manage all the implications of it such as govt religious discrimination or religious fervor, is to convert ethnic groups into EG-Cult units. In the US, for example, you'd find these social units:
American-Catholic
American-Protestant
American-Jewish

instead of simply "Americans" having the property of being christians. This would lead us to a huge number of ethnic/cult groups, specially if these social units exist in mapsquares!
This is, btw, partly the approach Manurein (the first social model developer) took, calling each social unit an "agent".

The alternative is to remain within the initial idea of ethnic groups and provide as part of its definition something describing how people is divided among religions, like:
Catholic-30%
Jewish-10%
Protestants-60%

But this implies a mixing of religious attributes is needed to compute several things and in the other side it'd be difficult to apply a discrimination against some religions if ethnic groups have at the same time multiple cults. Things don't look good this way. Mixing religious doesn't sound too bad in this example, but what about averaging christianity with buddhism?

A third alternative, and the one I'm playing in my mind with, is to allow ethnic groups to follow several religions provided they're similar. Just like the example of alternative two above. Making some adjustments to allow govt religious discrimination to act only against "religious blocks" and to keep religous fervor only between these blocks, then it doesn't sound too bad. Averaging religious attributes wouldn't be too bad if religions are alike and ethnic groups could have multiple cults. The branches idea fits perfectly here.

Anyhow, the thing is modeling religions can become a pain in the ass if we push it too far. On the other side, I confess things are not solved in this topic.

Anyway there currently is no way for hybrid religions like Jainism (Islam and Hinduism) nor is there anything for cultures that have what is called plurality, ie people can worship more than one aspect of a religion. The latter i think is ness otherwise we'd be biasing religion from a western/arabic sence which isn't right either. How do i propose this should be modeled. Well that is tricky. First there could be a 0/1 switch for plurality/not and if not, well under no circumstances can it be worshiped by the same person along with another religion. In my suggestion, although i know it is not the best way to do it, is to average the religions out and take the medium for the score. Although i know it could also be extremes, but we haveto choose one or the other i think.

I am not sure I understand.

Do you mean that each square contains a number of objEGs? Each objEG contains a objCulture with the cultural attributes of this particular EG (the reason to do this eludes me - probably an OOD-thing, could you elaborate F_Smith?). I deduce that each objEG must also contain a variable for the number of people of this ethnicity in the square, and depending on how the multiple religion issue is solved, some indication of how many people follow each of a range of possible religions.

This leaves an important question:

Will objEGs of different squares ever act in common?

If not, I fail to see that this models unified action of substantial factions of the population in the Empire. And this was your original idea of the EG-concept wasn't it?

If objEGs do - as I suspect - act in common, then you have to have some criteria for which objEgs will assemble into larger groups. What are they?

It is in this context that I proposed using the Nation. In its simplest form this is really nothing but a name/code that allows objEGs from different origins to assemble into one group acting together - regardless of other cultural attributes, which typically would be averaged in some way. It would probably make sense to put this name/code in an object along with an array pointing to every objEG in the Nation, since this would allow looping through all objEGs of a nation, when necessary. (The addition of 'magical' proto-ethnic characteristics was just an idea (not necessarily a good one, at that), which could diminish the tendency of different objEGs to drift appart as time passes. These characteristics need not be static, but could evolve). I guess it would be possible to imagine other attributes that could have a similar function, the most obvious being language. In fact you could call it language instead if you prefer (I suspect Mark would like this).

This approach has some nice properties:
Historical acuracy. Nations/languages have been very constant through history.
Ease of computation. Basically it just takes comparing two integers to decide if objEGs are 'compatible'.
Handling of simple migration. Basically if a part of an EG migrates from one square to another there will be two possibilities:
There is no objEG of the same Nation there, in which case the people could either a) form an objEG in the new square of the original Nation (if they are numerous) b) join another objEG in the square (the one least distant in cultural attributes terms)
There is an objEG of the same Nation there, in which case the tally of that objEG is just increased.
It might be argued that migration should not take place on an EG basis, but rather as bulk-migration, where every objEG in the origin is reduced proportionally, while any objEG in the target square is increased proportionally. However, thereby you forfeit the opportunity to have discriminated EGs 'vote with their feet' and get the h… out of there.

Hi. Sorry to be just now getting back to you, work is picking up right now. This week and next, my time is kinda limited.

1) Yes, each square has a number of 'ethnic group' objects. Each 'ethnic group' object holds (encapsulates) data describing it -- culture, religion, tendencies and general attributes (number of individuals, demographics info, tech info, etc). Like a 'car' object would hold 'engine', 'color', 'transmision', etc.

2) Yes, in many cases ethnic groups will share the same 'culture'. The 'criteria' for grouping them together is thru their shared 'culture' object.

Again, this is data model speak -- the 'Ethnic Groups' will absolutely have to hold info on what their individual tech levels are.

Like an individual 'car' holds info on what kind of tech it was built with (to an experienced eye).

That has nothing to do with how the player manages tech info, or how the game 'models' technology . . . they're two different things.

Maybe I'm just daft, but...

An EG is an object that represents people in a map square. It should hold data that are particular to the people in that mapsquare alone, like their number, demographic info and maybe possessions, together with some pointers to other objects that they are related to in some way. Why then, do you suggest that the square-based EG should encapsulate data that are not unique to a single square? In your next-to-last post you stated that EGs are grouped together through a shared cultural object. So what should really be stored at the map-level is a pointer to the shared cultural object, which then encapsulates the culture properties. The same goes for tech-level, which I suspect must be a civ-level thing. I would have the EG point to the civ, which then has a property called tech level.
Taking your car-analogy: The car should store data on the make, model, year etc, and a pointer to the owner of the car. The owner object should then hold data that is related to the owner, e.g. sex, nationality, insurance company etc.

And regarding the grouping of EGs. You say that EGs are grouped together by a shared culture object. Let's examine what happens: Take two squares inhabited by people with a Lybian culture. Each square will hold an EG, and a reference to a shared Lybian objCult. If the two squares are occupied by different civs the culture of each EG will evolve, beginning from day one. You cannot change the single Lybian objCult two ways simultaneously, and therefore you have to split the Lybian objCult into two objCults, the moment they are under control of different civs. They thus do not longer have a shared objCult, and cannot be grouped, even if they are united under one civ next turn.
As far as I can see there are three ways out of this dilemma:
a) You could group them according to an attribute that does not change when the square changes owner: This is where my suggested Nation/Language/Whatever comes in. It is simply an attribute, unchanged by controlling civ, that can serve as a common reference to group the EGs by.
b) You could group them together according to some arrithmetic 'nearness' FE the D-principle of the social model. However, whenever there is a need for concerted action on part of grouped EGs in a civ, you would have to check every EG against all other EGs to see which ones should be grouped(constructing a large 2x2 matrix of all EGs in the civ).
c) You could refrain from grouping EGs

And: No I do not think that nation is a political thing. Granted, the nation has become more or less synonymous with the state during the last century and a half, but this is not a phenomenon that has prevailed through history. FE the Latvian nation has only had its own state for 20 years between WWI and WWII and for the last 10 years. None the less it has been a national entity for more than 1000 years, and there are actually still a few people (less than 50 I think) that speak the balto-slavic Latvian language. A nation does not disappear from one day to the next as by conquest. In stead it is slowly assimilated into the surrounding nations, until in the end it disappears. I believe this is accounted for in the social model where people of minority EGs slowly change their allegiance to the dominant EG of the civ.

I'm sorry, I didn't realize you understood the terminology. I try to keep this discussion as non-jargon oriented as possible. No one else here has any exposure to OOA/OOD at all, and most of these models have been designed in a very non-OOD way.

So yes, if you know what a pointer is, the Ethnic Group actually encapsulates a pointer to a 'culture' and 'religion' object. Those objects are stored at the game level.

FYI, it also encapsulates a 'tendencies' object and an 'attributes' object, a 'demographics' object, at this point. There will likely be more, later.

The 'culture' (and religion, for that matter) will split/evolve from time to time, I don't think we should try to get around that. It's easy to do, and absolutely necessary for realism.

If you're interested in the architecture of the data model, please do glance at the threads on the OOA/OOD of the game's models. It would be great if you could look over and critique what's been architected so far.

EG-Techs: Please re-direct discussion to the tech thread. I made a comment about it already.

Know that I see F_Smith's words, I understand what's currently implemented in terms of EG and their culture would help a lot in modeling an approach like the one you're suggesting, Beör (nations). Since EGs and cultures are handled differently, a lybian EG could still be connected to all lybian EGs through this common culture. But I strongly disagree with this!!

Each EG should evolve in its own way given its particular environment. A reference to a pana-culture or protho-culture is a very magical concept. For EGs living appart, the only way to preserve a common culture would be with intensive relationships between them, which in time contradicts the fact that they're appart and living different realities. Cultural attributes of a particular EG should be independent of all other EGs in order to model smooth and realistic divergences in customs and ways of living. Handling culture and EG separately doesn't serve to this goal, so I believe, F_Smith, that you have to integrate them. Otherwise we're saying culture evolves independently of the EG itself as a "thing" of its own. EG' cultural attributes are unique to it in the sense that can change individually with no restriction to what happens to other EGs. The division between EG and culture is artifical and useless, IMO.

I don't think the question was vague. It could be answered with a clear 'Yes' or 'No'. And your answer was - of course - 'Yes'. Had to be, otherwise the original EG-concept would go down the drain.

My next question would be:
How do you decide, which EGs should be grouped? In other words, what makes the German EGs in the province you mentioned German. What identifies a german settlement? F_Smith's answer was 'a common culture object'.
What is your idea on this?

I agree that migration would benefit from its own thread. I think however that we should wait until the framework of the social model is decided upon.

About your next question (germans), the essence of the problem is IMO that you keep seeing a culture as a thing existing on its own. You're asking me really what the "german culture" is like. You want to picture and see the german culture as an independent entity. Otherwise you wouldn't ask me what groups belong to the "german culture". What indeed exist is several groups of people who call themselves germans. The german culture would be, therefore, the aggregation of their cultural attributes. The aggregation is, however, an abstract thing that only helps you see the world in a simplified manner. In this aggregation, of course, you lose info. In a big country like the US, for example, talking about the "american culture" wouldn't let you notice differences from people in the east vs the west, black vs white, etc. The american culture doesn't exist. What exist is several social units in North America with some differences between them in their customs that call themselves "americans". "American culture" is just a simplification the observer does. That's why I don't want to model "cultures", but social units (ethnic groups).

How would the social model handle this potential "knowledge gap"? Is there a way to easily add this stuff in with realistic effects? For example, would a certain EG achieve dominance if it had a higher knowledge level? And if so, is there a way to model the reaction of the other EG's? Also, are there characteristics of individual EG's that would make them more or less likely to research or innovate?

quote: On september 18th we celebrate in Chile our independence day and since friday there've been a lot of activities around the subject and a lot of drinking!

Bolivar the Liberator, huh? What exactly were you drinking my boy? Some strange local liquoir, or just the usual stuff?

Btw, does September have a symbolic meaning for Chile? IIRC, the military coup of 1973 took place in September 11th.

No, no, no. You interpret me all wrong.

From your last post I gathered that at some points in the game it is necessary to have EGs in different squares group so they can act together. Your own example was:

quote: ...a rebellion, like a germanic rebellion in a roman controlled land. In this case, yes, the riots model doesn't look at mapsquares to determine chances of rebellion, but at the province level. According to the riots model, chances of rebelion increase if germans see they have enough population in a province. This can be interpreted as a coordination of germanic settlements in a greater territory than mapsquares.

I know one thing is a difference in the ruling civ, espicially if the civ is of completely differnt governing aspects than the EG would normally be for. Relocation after this has also been discussed. However, what i want to know is what would cause internal splits and convergences in EGs. If you haven't read the tech model or forgot here is an excert of one of my posts pertaining to this matter:

quote: Each EG in each square might exist in squares where a tech is useless to them. FE an EG that is spead over a large penesula where there is a lot of coast and a lot of mountainous regions. Now then because the people at the coast would know how to fish and shipbuild so would the people in the mountains? This isn't right...because the people down below devote more to farming they people in the mountains get better at it? Again i don't think so, espically if they have little contact. Also you are forgetting the fact that when two places have an advanced culture come in, the advanced culture is quickly adapted as far as technology by the old and thus after a few turns you have loads of redundant information that must be caclulated....

Actually, the name 'ethnic group' is perfectly appropriate for the object, at all levels of aggregation. You can have an 'ethnic group' object stored at any level -- civ, mapsquare, province, etc. It'll just be an aggregate at higher levels.

All we're really discussing here is the storage strategy for EG info.

They are attempting to make a single, big data aggregate and break that down when needed. Monolithic design, it's sometimes called. I'm afraid my experience shows me that it's always more efficient to build a complex system out of small parts. 'Component' architecture, as it's called.

They want one big 'EG', broken down as needed. I coded a bunch of small 'EGs', and aggregate them as need be.

I am attempting to explain my use of this 'component' architecture, and why the game program will be better if we use it.

So far I'm not succeeding, I'm afraid. But I enjoy this discussion, it's valuable experience for me that is applicable to my work.

I'm not sure how much programming you've done, but this is actually a common design error from procedural programmers. I'm kind of afraid to mention that, because of the reaction I might get, but there's no other way to put it. I've sen this before, updating legacy Cobol systems. One of our Cobol products, an insurance tool called 'Vantage', is a design nightmare . . . and now they're trying to add EJB-based middle-ware and a servlet-based front-end.

I don't disagree with you that the term EG covers the concept at every level very nicely from a programmers point of view. The problem is that the term was previously associated with another concept: the multisquare EG. And this has really confused things. At least I think we should adopt some prefix notation or such to make matters clear

sqEG: Square based EG - the basic population object
prEG: Province based EG - the aggregate of all sqEGs in a province of similar nationality - similar to the earlier concept EG
ciEG: Civ based EG - the aggregate of all sqEGs in a civ of similar nationality
msEG: any multisquare EG - including EGs crossing civ or province borders

(I rather like prEG )

LGJ

My concerns exactly. How sqEGs are aggregated and msEGs are split up into separate entities is of major importance.

Both + Rodrigo

Axi: Yes, september as a whole is an important time for us. It's usually called the "Nation's Month". We don't celebrate it all days of the month like a carnaval or anything, but usually it's a period to strenghten our culture and customs through various activities. Yes, the ****ing military coup happened in september 11th, 1973.
What did I drink? The usual plus two national drinks: "Pisco" (also present in Peru) and "Chicha".

Beör: I believe the average chilean would choose Spain. Italy would get a very notable 2nd place, I think. BTW, I've been in Europe, so my guesses are based on my experience there...
Of course, the aggregation (average chilean) loses information! The south of Chile, FE, has an important amount of population who immigrated from Germany and has kept german traditions in many ways. I'm sure most of them would prefer Germany instead of Spain. Like any other country in the american continent, Chile is a land of immigrants "mixed" with native americans (the mix level varies a lot from Alaska to Tierra del Fuego, though), so taking the average chilean is just a simplistic way to see it. Why do you ask me this?
BTW, we have a lot more in common than you think, my friend, and distance and years of separation can't break the european/western culture initial immigrants
brought. We just twisted it a little!

Both of you: Bolivar is a major figure in SouthAmerica, but in Chile didn't play such a great role in independence. Our "G. Washington" was Bernardo O'Higgins. See the last name? He came from an irish family... a land of immigrants, remember?

Thanks for the cheering, anyway
----------------

To the social model:

Beör: The game engine will look at the EG's with "german" nationality to compute riot model's equations for that particular thing which is a nationalistic rebellion.

Richard and LGJ: The EG-Tech thing is IMO a key issue that we should solve as a team soon. Please look at the thread I initiated called "Too much shaking goin' on!" because that's IMO just part of a bigger problem.
Anyway, I'll answer very briefly to your questions:

Richard: With the current model, no, you can't have each EG having its own tech level. EG's weren't developed to do that. If we want it that way, though, I see two possibilities. The first is to run the tech model at the EG level. The second is to run it at the civ level as planned and add to each EG modifiers that could reduce or increase civ tech levels in whatever circumstance we want to know what tech knowledge an EG has. The second approach doesn't seem like an easy solution at first sight at least.
On your other question, yes, the idea was to give EG's some characteristics that could make a difference in tech innovation. My idea was to use attributes Asceticism, Traditionalism and Literacy. Tech development, at whatever level we decide to run it, should IMO be influenced by EG's attributes.

LGJ (and Beör): There's no internal splitting within an EG because of the definition of it. An EG is a group of people with the same characteristics. A split never occurs within the EG. Part of the EG could migrate or be conquered by other civ, and in that case the EG is divided in pieces and each one evolves on its own.

I didn't have it like this in the original model just for simplification, but your mountains vs coast example will be 2 different EG's in F_Smith's implementation of the model. As I have said many times, with EG's imlemented in mapsquares a lot more realism is gained at the cost of more resources in the computational side. So in your example you'll find realism. EG's in the mountains can be potentially different than those in the coast.

As for convergence, cultural characteristics aren't IMO the key for convergence. Even if two EG's are very similar in their customs, the issue of nationality still persists, so it becomes the barrier to converge into one single EG. There're a lot of examples of this. Think, FE, in Sweden and Finland. Almost same culture, yet they identify themselves as different peoples (different EG's). So, what the model does is to allow EG's with a low Nationalism level to be absorved by the "dominant" EG. That's the only criteria for convergence. In essence this say that an EG with low sense of identity and where its members hardly see they're a social unit of its own, are willing to join other "tribe" and take their ways as own.

About your kurdish/israeli example, the first thing to note is EG's in different civs are not "forced" to evolve differently. If they live a life in different civs but these civs are alike in the parameters concerning social change, then they won't diverge much. The change in culture is given by the reality the EG lives in, so divergences should be measured in that perspective.

Second, Traditionalism. A case like the hebrews could be modeled with a rather high Traditionalism. Even living in very different places, tradition plays the role of avoiding culture change. It can even be set to 100%, where no change at all is allowed.

-----

manurein
Clash of Civilizations
Social Model
Paris, France
May 99 posted September 20, 2000 16:15

Hi all
I designed a social model for this game some times ago.
I think that this model answers a few questions Böer is raising.
You can see it here : ../../Forum21/HTML/000151.html

See the mixing and the migration sub-models.
This model was based on the assumption that cultures tend to converge when they are in neighbouring places for some time.

A few remarks : first, I have left the Clash team for lack of time, but i still try to follow it, and i really want to play it : keep the pressure guys!

Concerning the procedural vs OO design, it's obvious that OO is far superior in such a project. This said, i think that you make a mistake when you try to design directly OO models; IMHO you are missing a step : OO design concerns the implementation of the model; what you are doing now is deciding what you models will do ; this is the specification phase, and OO has nothing to do with specification. First decide what your models must do, give the equations they will implement, etc. When your model is ready, then implement it, which means first designing it then coding and testing it. Clash is a very complex project, far more complex than most commercial games. If you mix the specification and the design phase, you overcomplicate the job needlessly.

Also, you should try to design each model as a "black box" : the other models dont need to know "how" they do, they need to know "what" they do.
What I mean is, all a model needs to know about the other models is the datas that the other models can give him for integrating in their computations. They dont need to know how these datas are produced.
If a model does what it is intended to do, then it can compute ant value concerning what it does, even if it does not use it internally.
Then, each model designer should design his model knowing what the other models are intended to do. Then ask the other model designers to supply him with the datas he needs for his computations. And supply the others with the datas they ask.

Thats all guys, good luck!

-----

Beör
Chieftain
Copenhagen, Denmark
Aug 2000 posted September 21, 2000 08:39

F_Smith
I sort of agree

If the lower limit of EG size is 1000 people, the havock caused by migration will not be as large as I fortold. You could probably keep it in the 20.000-30.000 range, which I find high but acceptable. I read the other day that there is between 5000 and 6000 languages being spoken in the world, and people speaking these languages will be subdivided into many groups. But mind you many of these languages (I believe it was a little more than half) are on the edge of extinction, being spoken by less than 10000 people. We will have to test this. I say that if more than these 30.000 cultures are created we have to consider an alternative.
Actually there is another thing about several of these aggregation techniques (common object, identical attributes, similar attributes) that I don't like and that I think points to a problem with the method:

YOU WILL NOT KNOW WHAT TO CALL THE AGGREGATE.

If you use the nationality approach you will call the aggegate 'Germans' or 'Germans in XX province' or 'Germans in YY civ'.

With the other methods it will be:
'All persons that were originally Ggerman, but are now something else characterised by their cultural attributes a,b,c,d,e,f,g,h,i'
or
'All persons with cultural attributes a,b,c,d,e,f,g,h,i'
or
'All persons with cultural attributes close to a,b,c,d,e,f,g,h,i'

Grouping by nationality will preserve the flavor, while still allowing EGs to have differing cultural attributes. The aggregates will be larger and more heterogenous, but isn't that closer to reality?

-----

F_Smith
Prince
Austin, Tx 78728
May 99 posted September 21, 2000 09:48

Richard:
I think we're discussing that in the other thread, so I won't repeat here.

* * *

Boer:

Actually, the player can easily filter/aggregate by each and every EG attribute.

Since the player won't have god-like control, the only reason the player will need to know what civ-level aggregates are would be for whatever game models require. So the aggregate problem is a matter of the 'civ' level modelling, not 'square' level 'ethnic groups'. Even if we store 'ethnic group' info at the civ level, we'll have the same 'aggregate' issue.

I will have the filter available for a player to see how many people of a single ethnicity, how many people of a single culture, etc, are in their civ.

* * *

Manurein:

Hi. Welcome back.

[This message has been edited by F_Smith (edited September 21, 2000).]

-----

Beör
Chieftain
Copenhagen, Denmark
Aug 2000 posted September 21, 2000 17:31

manurein
Read your old model - it was not manureout (OK you probably heard that one before!).

There were several things I liked (although I am glad we got rid if the race thing ). And you are right: Many of your thoughts on grouping people in 'agents' and 'meta-agents' are valuable in the present discussion.

I would like to emphasize one thing from your model that I think is worth discussing (and I am sorry: this will relate to scale theory): The use of nominal scale (cathegories) instead of ordinal scales with so many steps that they are in fact continuous has several advantages:

1. You now what to call your groups: the Green Race, the Blue Race.

2. By using nominal scales you don't imply that one category is superior to another: the Blues are not better than the Greens, they are just different.

3. You don't imply that they are placed on a continuum. That somewhere between the Blue and the Green there is a Blue-Green. And between the Blue-Green and the Blue there is a Blue-Green-Blue.

The weakness of nominal scales are just as obvious:

1. You have to find a meaningful name for all categories.

2. You cannot perform arithmetic on nominal scales. Thus there is no average or median. The mixing of people gets diffcult, because you cannot estimate the 'typical' value of the merged people, other than choosing the value of the largest number of people.

So when we want to aggregate people it is necessary to use at least an ordinal scale, where arithmetic operations are possible. But this leads to the problems of no names, lacking identity and establishing proximity (particularly when you have to do this in several dimensions).

Fortunately F_Smith has opened my eyes:

F_Smith

Of course you are right EGs can be aggregated by any of the attributes, and the player will rarely if ever have to do that. I was going of on a tangent because of the discussion I had with Rodrigo in the social thread.

I guess the filter you mention will translate values on the 0-100 scale into another scale with much broader intervals (FE 0-20, 20-40. 40-60, 60-80, 80-100 or maybe even wider) which we could then give a name and use this name to describe the people to the player: '15% of your people are very agressive' or 'On average your population is moderately nationalistic'. Beneath this we would still have the actual numbers thus allowing aggregation of EGs either on the detailed scale with the disadvantages I have outlined or on the less detailed scale.

The other kind of aggregation is the one performed by actions/events in the game, and here we can decide ourselves which method we find most appropriate in the circumstances

Ingeneous - why didn't I think of that .

Whisper mode

F_smith: That's Beör - not Boer. B E Ö R

Funny how you can identify with a silly alias acquired about a month ago.
Whisper mode off
[This message has been edited by Beör (edited September 21, 2000).]

-----

BRIEF SUMMARY
People is grouped in Ethnic Groups (EG) and characterized through a set of cultural attributes. Each EG has also some other variables defining its status like if it suffers slavery or not.

EG's have one or more religions. The number is still TBD.

Cultural attributes are subject to change along the game. The way they change depends on the reality the EG lives in.

Religions work as agents of cultural change also. When an EG follows a religion, it tries to take the attitudes the religion dictates or encourages. To this end, religions are defined as a set of "desirable values for cultural attributes". These values are fixed for the whole game. Just the intensity with which people follow the religion changes.

Religions spread through the world through contact between civs, which is measured in trade routes.

List of Cultural Attributes
Nationalism
Ethnic Tolerance
Religious Tolerance
Individualism
Asceticism
Traditionalism
Aggressiveness
Land Connection

List of other variables defining each EG
Religiously Discriminated (dummy: the EG can be or not discriminated because of its religion)
Ethnically Discriminated (dummy: the EG can be or not discriminated because of its nationality)
Enslaved (dummy: the EG is enslaved or not)
Nationality

GLOSSARY (MOST IMPORTANT TERMS)
Ethnic Group (EG): A group of people with the same customs and cultural profile.

Primitive or Ethnic Religion: The religion an EG is assumed to have at the beginning of the game.

Great Religion of the World (GRW): Religions like Islam that are allowed to spread globally and eventually overshadow all Primitive Religions.

Religion's moral code or doctrine: The set of cultural attributes values the religion encourages.

Minorities: All EG's that are discriminated because of religion or nationality.

Majorities: All EG's not discriminated.

CURRENT STATUS
The model is being coded by F_Smith in the "Beast". You can check it in http://home.austin.rr.com/lostmerchant/Clash/Editor/ClashEditor.html

The number of religions per EG is still unsolved.

The following cultural attributes should be included in a future update: Literacy, Natural Affinity and Corruption.

The handling of the attribute Nationalism should be refined.

Religious converting is missing. Spreading is the passive way by which a religion gets more followers (already covered), while converting is the active attempt by preachers to get more followers.

It has been suggested that religions should have branches, like christianity producing catholicism and protestantism. Still unsolved.

The model initially was supposed to manage migrations. It doesn't right now and this issue may end up being handled somewhere else.

There won't be 5-6 GRW. I said there'll be 5-6 groups of GRW. There will be as many GRW as you want.

Hinduism: I don't understand why you don't want to model it as a GRW. If the religion itself (its ideas, its philosophy, etc) makes reference to an ethnicity, then it's an Ethnic Religion. If not, it's a GRW. I don't see the problem. If you say it doesn't spread as a GRW, then you're not understanding what a GRW is. A GRW has the ability to spread beyond ethnicities. It's not its destiny to spread vastly. It just can if the religion is attractive.

I keep suspecting that you merge the real world meanings for "ethnic religion" and GRW with the definitions we have here for the game. For whatever religion you consider (hinduism, celtic cult, etc), what you should do is to see in what category it fits better given the definition of the model for ER and GRW. If it doesn't fit correctly (or mostly correctly) in any of the two categories, then let me know. Use the following as definitions:

Ethnic Religion: A cult that is practiced only (or almost only) within a particular ethnicity. Example: ancient egyptian cult. Its capability to spread beyond the original ethnicity is so small that we consider it null. This capability is so small because cult ideas and practices tend to refer explicitly to an ethnicity in many ways "God XX protects the YY tribe", FE.

GRW: A cult that can be practiced beyond ethnicities. It's capable to spread beyond the original ethnicity. This capability exists because the religion has null or almost null references to a particular ethnicity. "Alah is good to all who follow him", FE. Actual spreading into another EG depends on affinity between culture and religion's philosophy/doctrine/ideas.
-----------

Celtic: There was a point in time where Celtic religion would be by the definition a GWR, but once Greeko-Roman and even moreso Christiranity religions moved in, it lost much of its hold and what was still left, mostly in Ireland and Scottland for the most part was of generally the same enthinicity or at most 2 very similar enthnicities. At this point it would be considered an ER because it no longer had the power to spread like it did before.

Hinduism is IMO a GRW. The fact that it hasn't spread beyond India and the nearby territories has IMO more to do with the affinity part I talked about in my previous post.

The important thing, though, IMO is having the right overall flavor. We won't make a big mistake if we put celtic religion in the ER category and Hinduism in the GRW category. As a result in game terms, celtic cult will exist only in celtic peoples and could be overshadowed by a GRW like it happened with christianity. On the other side, handling Hinduism as a GRW allows the religion to spread given the right conditions (FE if the Indian civ is able to export its culture through conquest) and wouldn't disappear with the arrival of another GRW.

You should, if possible, have a advocate of each GRW decide its stats.
This would help avoid being insulting, and a follower of a religion could better determine its stats.