The computer game series Civilization  created by Sid Meier manages to find a spot near the top of every list of top computer games of all time. As a moderate game enthusiast, I agree with such a placement. It is an addictive game that remains enjoyable long after being dated by better software technology. There are three versions created in the last 15 years with a fourth due in 2005, but previews show it will not likely address many of the challenges to the realism that I address here.
The title reveals the massiveness of its scope. You start as an ancient tribe just inventing the wheel and the alphabet with continuous progression in time to the modern world and beyond. Libertarians and anarchists who have played have probably grumbled at the portrayal of government. 'Anarchy' in the game is the period of chaos, revolution, and poor productivity experienced when switching between supposedly more stable and permanent forms of government. Despotism is the original form of government at the dawn of civilization, and different government types are 'discovered' through research, and democracy is considered the most advanced and generally advantageous form to have in the game.
Having played the game, I thought it would be good to have a more realistic and thus libertarian version of the game. However, the game requires a government to be controlled by the player in order for there to be a game to actually play. Trying to imagine an anarchist version, I could only think that being able to control a single individual unable to control the choices of even your own anarchist society, I gave up the thought of reconciling the political flaws of the game in a manner to retain the entertainment value.
The series has tried to make entertaining scenarios where a player can play real historical wars and events. Some scenarios it could do moderately well such as the World Wars. However, George W. Bush's current War on Iraq could not even remotely be recreated within the design framework of the Civilization series. Not only does it fail completely to accurately model human nature in government, it equally fails to model 4th generation war .
An attempt to make a Civilization game model of the Iraq war would have one enormous problem. It could correctly model the three week conquest of Iraq , but that is it. In the Civilization game, that would be the end of war, end of story. Iraq with its oil reserves would then be a part of the American Empire. Destroyed infrastructure would have to be replaced, but citizens would go from pawns of one state to pawns of another without any further opposition or guerilla war. Judging by Bush administration pre-war statements, it seems they might have based their lack of post-invasion strategy on playing Civilization.
But what would be the best way to model the fundamental nature of real civilization to replicate all these historical events? While seeking this answer, I discovered that real civilization with its major complexities could in fact be modeled into a playable game.
The existing series starts with nomads settling down starting their first non-nomadic village. This dawn of land-bound agriculture should not begin with despotism as Sid Meier has it, but the opposite: anarchism. In a world of wide open spaces and nomadic lifestyle, those who want to leave can easily do so, and chasing them down to enslave them would be an impossible task, so any government must be wholly by consent at the start of the game. What 'you the player' as a people are is not a state, but a people/culture/tribe. (I could add nation, but the modern meaning of nation has changed.) The existing games blur this distinction, just as state propaganda would not have us comprehend the difference.
However, it is true that this 'people'--either yours or other peoples in the game--could become a state, and likely many will. It happened historically, and the game would be inaccurate if perpetual blissful anarchy was the norm. This possibility starts after agricultural former nomads become fixed to the land, forgetting the nomadic way of life, and space constraints first make escape less likely to succeed. Two societies cross, often in competition for relatively scarce resources. Language barriers naturally exist, but were not modeled in the Civilization games. These societies can either spend the effort to learn to practice peaceful trade, learning your neighbor tribe's language, going to war, or some combination. Whenever one formerly anarchist tribe conquers another one, a state is born. But if there is no complete genocide, and the loser is made to pay tribute, tax, or enslaved, then this new combined dual society has castes and/or classes.
The existing games would make such a conquest the end of game for the conquered, and the conquered take on the culture and identity of the victor. Instead, this should not be the end of game for the conquered. In reality, the conquered people live on, and may fight against the state, either within its laws by trying to change them, outside the law of the state for their own higher law, or some combination. Eventually they could seek to emigrate somewhere where they can have more control over maintaining their identity.
One particularly interesting historical people that any realistic Civilization game should be able to model is the Israelite/Jewish people's history. Start from their entry into Egypt per the Torah record. Joseph was offered a high position in government by the Pharaoh for his management skills, and in return, the Israelites became a privileged caste, whereas the rest of the country became indentured servants. But as a minority, eventually the tables turned and the influence of the masses turned their privilege to slavery, from which they eventually made their Exodus. After the Exodus they are again a free society, though eventually enslaving themselves to a king. Part of Israel is then scattered by Assyria to such a degree that they lose their identity. (In game terms, this loss of self-awareness as a people would be the 'end game-failure' condition, if this were the whole people.) Then there is the Roman Diaspora in AD 70, but the Jews continue to maintain their cultural identity in multiple nations.
This is what should be modeled: A game player controls a united people whom are not bound by force. The people could stay free, become enslaved in a state, become enslavers in the state, have their identity fade away through assimilation, hostile propaganda, or even genocide by enemies. If the game player does enslave another people, then at most he can control most of their production. He cannot control their will. (Though he may try and have some success with propaganda, bread, circuses, and buying allegiance through granting privilege to leaders of the enslaved people.) Likewise if he is enslaved, he controls his people's will to resist, the fighting spirit, the counter propaganda, and even attempts to seek friendly cultures to make an alliance against his enslavers.
The factor that unites a people should have a name. Perhaps the best term for this, get ready . . . is religion. This may seem odd and incompatible with the reality of multi-cultural cooperation, but there is a type of religion that makes the reality instead of the Plato's cave type perception. Are the American Baptists and American Mormons part of the same, or hostile religions? They seem to be able to go to war together to kill the Assyrian Orthodox Christian women and children in Iraq and be proud of it as justified collateral damage. In truth, their religion is not Baptist or Mormon, but American.
Every state is a religion. Generally we call this worship of state civic religion, with its flags, anthems, pledges of allegiance, rituals, and holidays. (Notice the etymological origin from holy day.) States often allow freedom of religion, but by propaganda, forced schooling/brainwashing, and threats to dissenters, it instills the civic religion into the core of every approved religion. You no longer have to believe that the state or king is God, only the instrument of God to be obeyed and the legitimate maker of laws of right and wrong.
In the existing series, a player is the head of state who must choose how to allocate workers to agriculture, manufacturing, trade, science, entertainment and such fields. This could still be done similarly in this essay's revised game concept as a popular leader of a people, and it can be assumed that your people as a people consent to your general wisdom. (Exceptions nevertheless exist for alienating some people, and you could experience an outflow of people feeling unappreciated.) If whole cities are entirely of your people, then control of city dynamics could be similar to the existing games. Some of these cities could be in free anarchies, while others could be controlled and regulated by hostile states demanding tribute. Individual cities could be mixed cultures, and the trade and interaction of your people could vary by the strengths, weaknesses, peaceful or hostile state of other cultures to yours.
So everyone would want to know, 'How do you model choices between capitalism and socialism?' Unfortunately, those terms are too ambiguous. I can't imagine a capitalism/socialism meter where you choose between 0-100% to represent the two. Instead, there needs to be meters to choose between tradeoff in voluntary cooperation vs. voluntary competition. There should be at least two of these: One internal one for how your own people are treated, and external meters for each of the other peoples that could change based on relative mutual regard between each people. Too much cooperation makes some people inefficient free riders, but too much internal competition gives no reason for your people to be loyal to you if they could be better off joining some other society.
If a player did create a state, then the game should accurately model the historical difficulty of undoing it. The creation of a state creates massive privilege among the conquering people. If a player created a state and then wants to abandon it, realism requires that at most he lead a break off culture (a counter-culture) because the portion of his people who obtained privilege would not follow him in this abandonment. So the game should not only allow for the various mergers between peoples, but even a single culture could split in two.
This game concept will have opponents. If a player (even a computer AI) decides to create a state and subjugate or destroy opposing cultures, then genocide is the inevitable approach against those who refuse subjugation. Players who likewise try to refuse subjugation will have to survive genocide attempted upon them. The opponents of this game will claim horrification at the possibility of modeling the Holocaust. Nevertheless, if the game provides an accurate model of this reality, it provides a means to observe causes of many possible historical or theoretical holocausts. Is not this a good way to remember the Holocaust to not repeat it in the real world?
I have no idea if Sid Meier and Co. are too ideologically invested in the inaccuracies of their series to desire these corrections or not. This concept is different enough from the Civilization series that a competitor could make such a game described here without violating Civilization copyrights if necessary. Still, this is only a rough outline of the structure of this game concept. Feel free to add ideas that could be useful to model the real nature of civilization, hopefully to present to a video game company to see if they would create such a game.
 This seems the obvious reason they did not return to their homeland when the famine ended. Why leave a privileged life for one of work? See Gen. 45:18-20 and compare Gen. 47:20-21.