Well, the alternative would be to get rid of the no-sleeping limitation completely, allow the player to attempt to rest around enemies, and depending on the distance between the enemies and the player, have a 90% and up chance to be interrupted, drawing all enemies within a certain percentage of the original "no-sleeping" distance to the player at once. Then, the player would not be alerted to enemies, would have to use their common sense when resting, and would have decent repercussions for attempting to nap next to a horde of angry NPCs or monsters.
Though it hasn't changed much at all, I'm going to toss this back out again in the hope that it can inspire even more ideas. Hooray for spoiler tags so it's not so much of a giganto-post.
Ideas and Improvements Regarding Large Cities in TESV:1. HOUSING Spoiler 1a. One thing that bummed me about the Imperial City's design (and really, the design of quite a few cities in Oblivion) was that is was relatively flat. Each housing district rose no higher than two stories. In my opinion, that needs to change. Large cities need to literally be built up a bit. I was envisioning staircases that rose up to a second level of housing, which was then connected by criss-crossed balconies and bridges. So when you stood on the road in the Elven Gardens District and slowly looked up, you'd see the entrance to Luther Broad's Boarding House, the second story containing his renter rooms, and then you'd see the balcony of the entrance to the home/inn stacked above his, with walkways going across the open air above the lane and connecting with other homes stacked above homes. Imagine Skingrad-style buildings, but maybe one story taller, and with an apartment stack feel to it; each floor being its own residence.
1b. While I liked having districts roughly split between upper-class, middle-class, and lower-class in Oblivion, I'd like to see a more ethnic flair in TESV. For example (assuming Skyrim), let's assume that the city of Winterhold is about 65% Nord. In regard to the Nordic population, they have roughly divided themselves into upper-, middle-, and lower-class segments of the city. The upper-class homes are of course immaculate, while the lower-class homes have visible damage, and aren't surrounded by the best of aesthetically-pleasing flora and city atmosphere. But for the rest of the minorities in the cities, are they going to just disperse themselves evenly among the rich and poor? I don't believe so, which is why I suggest ethnic cultural pockets in cities. For an example, the Dunmer tend to stick with other Dunmer, and have ended up forming their own "little Morrowind," just as in real life, we have "little Havana" or "little Italy." These clusters of homes might be adorned with cultural references, perhaps relics of Great House paraphernalia, perhaps we might find quite a few homes has a small shrine to their ancestors, etc, etc. Add two, three, or four of these little culture pockets into a large city aside its predominant population, and you make for some very interesting things.
1c. Outside the physical walls of the cities, suburbs should exist. Little immediate outlying communities that are still part of the city in question, but don't have the fortune to live within the walls. Or the reverse could be true, that their suburbs are nicer than the average inner-city dwelling. Further, tension or dissonance should exist between walled-in city-dwellers and outside-the-wall city dwellers depending on those circumstances. Do the walled in citizenry hold a grudge against the cozy backyard-bearing upper-middle-class cottages outside the walls as compared to their cramped and hectic in-city living? Or do the outside-the-walls slums hold resentment against the splendor of the inner cities?
1d. Along with suburbs outside the walls, farming land should exist also. A city must have a visible and sensible means of supporting itself. Unless a city is a renowned economic hub that could believably get all of its edible needs through trade, agriculture must exist in each city. That means: wheat farms, rice fields, potato/vegetable gardens, orchards for fruit, livestock pens, and the works. Furthermore, what type of edible material a city has available should tangibly depend on its food supply. If the city is up in rugged and mountainous terrain, then it's likely their only native food-source will be livestock, and meat will be in abundance in the stores. Everything else would be imported, meaning it would be more expensive.
1e. Real estate is something valuable, and in large cities, multiple options of real estate make sense. However, large and populated cities should not all go the route of "Oh, there's this unoccupied house for sale here. By and large, houses for sale should be occupied, and the ones selling should be the ones currently occupying. In other words, there should not be a convenient empty house for sale in each city, which the player must purchase from the count/ruler. Each city should have at least two or three potential houses that the player could acquire. And not all acquisitions should be as easy as "hand over X amount of money and become a homeowner." Some sellers might require a certain amount of reputation before they hand over the keys. Others might require some loyalty to a specific organization or cause. Still others might not be planning to sell at all, and they must be coaxed via some form of quest, whether that be subtle manipulation or not-so-subtle forcing of the hand. The real estate options in each city should not have the obvious distinction of "lower-class house," "middle-class house," and "upper-class house" (and whatever in between, for cities with more than 3 options). Some cities may only have lower-end options or higher-end options, and the differences in quality between the options might be far more subtle. Other cities may have houses that are pretty much the same quality, and the only choice is the type of layout the player prefers. Some cities' sellers may not take kindly to the player attempting to become a property mogul, and they may refuse to sell if the player has too much real estate in the area already. Other sellers may not care. Finally, most importantly of all, the player should be able to view the houses before deciding to buy, instead of just forking over a huge amount of money just to see the inside of the house for the first time.
1f. Taverns/Inns as dispersed amongst cities should be a bit different in nature, both stylistically and in naming. Tavern, for example, would be a better suited name to a liquor-serving bar that also happens to have some food for sale and maybe a room or two or three if needed (perhaps reserved for those too intoxicated for their own good). Inn, on the other hand, implies an establishment with a dozen or two rooms that just happens to serve some refreshments and maybe a meal. Bed-and-breakfast would of course mean a place with a small number of rooms and a plethora of food and a publican that goes out of their way to provide service. There should be strides to differentiate between these different styles of room-renting, and those styles should be enforced by the type of patrons to that establishment, the dialogue and services provided by the publicans, the actual size and capacity of the buildings themselves, etc, etc. Also, for all taverns and inns, the player should have the option to book a room for an extended period of time, potentially receiving a discount for a longer booking period (so to cut down on the need to repurchase the same room every single night even though you plan on staying there for a month). The player should be able to choose between open rooms, perhaps paying more for a more spacious room with better storage. Also, the player should not be the only one renting rooms. Random NPCs, potentially on the same setup as random encounters, should also book rooms at inns/taverns, and they should bear the hallmarks of travelers. Perhaps they are not native to the province, and they are a source of extra-provincial rumors. Perhaps they are on a trade-route within the province, and they rest at inns along every stop in their journey. Finally, there should be the off-chance that the tavern/inn might be full that night, if enough random NPCs or scheduled NPCs happen to be staying there at once.
2. COMMERCE Spoiler 2a. Don't have monopolies in large cities. Have around 2 to 3 distinct merchants for every service needed. Have 2 or 3 booksellers. Have 2 or three armorers, or perhaps even 2 for each type. You get the picture. Just don't have monopolies.
2b. [This idea is assuming the return of Merchants having a finite amount of gold that replenishes over time i.e. Morrowind] Give large-city merchants large amounts of available capital. I'm thinking between 15,000 and 50,000. Now, hear me out before you start screaming "unbalanced!" The replenishment for the merchants should not happen daily, as in MW. It should happen weekly. Actual physical trade caravans should actually physically visit the market of a large city and actually physically replenish their gold supplies. Also, said weekly trade caravans should trigger the respawning of merchant items (instead of those items respawning in 24 hours like Morrowind, or 3 days IIRC like Oblivion). Which leads me to the next point?
2c. Trade caravans should make weekly visits to a large city, and perhaps they should have a schedule of different large cities to rotate between. When they visit a large city, the player should be able to trade with them as well. There should be inter-provincial trade caravans and extra-provincial trade caravans. Inter-provincial caravans would carry unusually good equipment/items, but those equipment/items would not be exotic to the province. The extra-provincial trade caravans would only visit border cities, and their wares might include exotics not really found within the province. For example, a Redguard trader crossing the border to Skyrim and visiting the city of Falkreath might be able to sell you some Redguard-styled weaponry.
2d. These trade caravans should have multiple merchants and a plethora of guards, but they should be able to be encountered outside cities. More importantly, they should be able to be knocked off for a very nice profit. Perhaps in gold, or perhaps in jewels, equipment, items, etc. But knocking off a trade caravan should have serious consequences. Say you knock off the trade caravan that supplies the merchants of Winterhold every week with fresh gold. By robbing the caravan, you've loaded yourself up with some really nice things. But the merchants in the city of Winterhold will no longer be able to replenish their gold stores, meaning you will have a much harder time selling your wares in that area for about 3-4 weeks, until a new caravan respawns.
2e. Merchants and thieves do not mix. If the merchant has any reason to suspect the player promotes, is involved with, or is engaged in thievery, then the merchant should be able to judge the amount of his suspicion against the desire to make a profit on the player. If the suspicion outweighs the profit, then the merchant should refuse service to the player and demand the player leave. A lot of people dislike this kind of thing, usually saying, "How is anyone going to know I'm a thief?!" Well, they're not going to know (unless you're a terrible thief). However, they are going to be able to infer. And that inference can be made by the following: the people you associate with, the places you associate with, and the number of times things have gone notoriously missing with you in the area at the time (in other words, your reputation/disposition with thievery organizations and thieves, as well as relevant faction reputation modifications upon completion of thievery quests. Note this is vastly different than a blanket fame/infamy system). If your illicit reputation gets too high, and you're dealing with a merchant with a high enough responsibility value, they should refuse you until their disposition is high enough to override that refusal. Also, if you steal from a merchant and attempt to sell back the stolen item to that merchant, the same should apply. If the player gets too much of a bad reputation, they will have to seek out less reputable merchants to meet their needs [see 5e].
2f. Merchants should be mostly specialized, save for a few general traders who balance out their stock-versatility with less quality and less bartering clout. Those specialized merchants should have obvious rivalry amongst those selling the same things. If there are three apothecaries within the city of Dawnstar, then how much I sell and interact with each merchant should affect disposition amongst all merchants. Say we have merchant A, B, and C. Also say that I favor merchant A, selling all my available wares to him and only buying elsewhere when he doesn't have an ingredient in stock. Therefore, my disposition with merchant A will gradually increase as my store-loyalty is further proven, netting me better deals via mercantile. However, Merchant B and C's dispositions should gradually dwindle, meaning worse deals; after all, I am favoring and providing extensive business to their competitor.
EDIT: It has been brought up by Rhemaius that merchants whom you are neglecting may potentially try to earn your business by offering you even greater deals. To factor this in, the merchant's personality attribute would be the deciding factor of the threshold of neglect they'd be willing to put up with before giving up on you. For example, an ill-tempered Orc will get angry and give up on you fairly quickly, whereas an eager-to-please Bosmer would be willing to try far harder before reaching the same point.
2g. While it is unrealistic for every single item of merchant stock to be strewn about all on the ground floor of the shop, right in the open for any opportunistic thief to snag, it is just as equally unrealistic for there to be a small handful of merchant stock on the ground floor, with the more valuable stuff concealed within a container placed far below the floor, inaccessible to the player save for console commands. Instead, merchants can do the following: Have a small selection of lower quality items (for functionality) and higher-quality items (for the shelf wow-factor) actually on the shop's ground floor. Include some in-store storage for another segment of the store's stock, such as lock-boxes or chests, but be smart about their placement. Put them in secure locations, such as a locked basemant or a locked upstairs chamber. Eerything else (for merchants that have far more stock than they could plausibly lock up, or merchants who are security-conscious) should be placed in external warehouses, which are under guard by hired rotational mercenaries (perhaps the Fighter's Guild). That way, the merchants can still have a ton of stock, can still display their wares, can still have believable storage for the wares they don't care to display, and the player still has a fair chance (accompanied by reasonable risk involved with that chance) to rob the merchant blind and to steal anything a merchant has in stock. And whereas Morrowind severely punished players from stealing items set to respawn from merchants (stealing three potions from a vendor that were meant to respawn would result in those potions never respawning again), stealing items from a merchant should double the time in which the merchant will get that item back in their inventory via caravans (see 2b.).
2h. Gold (or bartering with loot) should not be a merchant's only form of accepted payment. Letters of credit (see 6i.) should be accepted by all merchants within the same city the letter of credit was issued in. [NOTE: This idea assumes the return of gold having weight]. If a city-state has enough of an alliance or economic partnership with another city-state, then letters of credit should be interchangeable between those two cities. However, cities that are none-too-trusting towards each other (or cities that are outright rivals) should not accept each other's letters of credit.
2i. The majority of the stock for food merchants should be dependent upon the agricultural resources available to that city (see 1d.) If the player disrupts the resources for that city by harvesting crops for themselves or killing off livestock, then food merchants in that city should be short of respawning supplies for whatever the player has taken. The supplies should reappear in the amount of time it would take new (or offset) crops to come to fruition or new livestock to be bought or raised, which I think is fair at roughly one in-game month.
3. RELIGION Spoiler 3a. The dominant religion of the province should hold dominance in its cities, of course. Chapels of the provincial faith should be prominent. Around those domiant chapels should live a selection of the middle-class and lower-class of the provincial race. Living alongside them should be missionaries and volunteer-like NPCs who strive in their little immediate chapel-community to keep people fed, clothed, and saved by the gods. Beggars should find refuge in such places.
3b. But in the cultural pockets I mentioned, (see 1b), other religions and abbeys/chapels /temples should exist and flourish. They should by no means hold the physical prominence of the dominant provincial religion, but they should be placed right in the middle of their respective culture pockets. Their religions should be just as in-depth and followable as the dominant provincial religion. Dunmer, as mentioned before, might have public and private shrines to their ancestors, or perhaps less conspicuous shrines to their "good" daedra, Azura, Boethiah, and Mephala. Altmer might have temples devoted to Auriel, etc, etc.
3c. Religion should be a point of contention between different cultures, especially between pantheons that are incredibly similar such as the Nords' pantheon and the Imperial Nine Divines. There should be much argument between the two; perhaps street-preachers could set up shop in the plazas, and two could compete with each-other in oratory to try and gain followers. Those who have been aided and helped by the dominant religion (i.e. beggars and the poor) should have everything wonderful to say about the dominant pantheon, and absolutely nothing good to say about the others. In fact, if you are a member of an opposing religious faction, you should be held by these groups as someone either to be pitied, preached to, distrusted, or driven off.
3d. Funerals should be performed by the appropriate faith when an NPC dies (either via the player or by other causes), rather than just lying dead in their homes or out in the street. AI packages would redirect the appropriate funerary service officials, as well as friends and family of the deceased. Some generic lines would be said, briefly summarizing that faith's views about death and afterlife. For more traditional funerary practices, a gravestone would appear in the graveyard, and either the gravestone would already be named for that NPC or a simple script would rename it. The graveyards would be designed with enough plots for everyone in the city. For less traditional funerary practices, there might not be need of a gravestone at all... C'mon. You all know you want to partake in a Bosmeri cannibal-funeral.
4. LAW ENFORCEMENT Spoiler 4a. First point to get out of the way: The Imperial Legion. Since we have no idea whether the Empire will still exist in TES:V, or whether provinces will become rival countries with a shaky, tenuous peace-agreement on the face of things, the Legion is questionable. If the Empire is still shakily stable, somewhat, then the Legion may be a serious force of contention and power in the next province (after all, Ocato / new "leader" may order legions to hold control of all provinces at any cost). If the Empire is no more / tenuous on-the-surface peace agreement, there are likely no Legions, or if there are, they are there to monitor Imperial interests in the province and Imperial interests alone. Or perhaps there will be all-out civil war as the focus of the next game (god I hope not), in which case, Legions might be the Imperial-supporting faction you can choose.
4b. Let's assume this scenario, then, regarding the Legion (assuming Skyrim). The Empire has dissolved, but both Skyrim and Cyrodiil have forged a relatively strong political alliance with the other, while tenuous on-the-surface-peace exists between them and everyone else. So Legions will likely still be stationed in Skyrim, as a token gesture. They won't have day-to-day interaction with the actual cities, and they'll mostly stay confined to Imperial forts scattered throughout the area.
4c. Let's make law a bit subjective, shall we? After all, from the PGE v3 regarding Skyrim, Solitude, Winterhold, Whiterun, Dragonstar, and the like are all their own independent (and sometimes rival) city-states. While huge and major infractions (such as killing a hero beloved throughout all of Skyrim) should be unanimously punished regardless of area, things like stealing from local shops, killing a civilian, etc, might not only be found out by other city-states, they're also not likely to care. Even something that could seem major, such as killing the King of Winterhold, might be met with mixed reaction throughout the area. Quite a few regions might hold you accountable for the murder, but the city-state of Solitude might not, seeing as how Winterhold was its ancient rival. Solitude might even praise you for your action. The same thing applies to Skyrim guards in general and the Imperial Legion, if they're included. They might not know, and they might not even care.
4d. Guards should not be robots. Firstly, as has been mentioned numerous times before, they should not be able to lock on like a homing missile to exact locations, even if you're sneaking and/or chameleon'd or invisible. But on a less-obvious note, they should not be emotional robots as well. They should not all fit the stereotype of the "good, never-flinching, incorruptible guard." Sure, in Oblivion, we had guys like Audens Avidius, guys like Ulrich Leland. But those were two out of many, they had quests related to them. I'm talking about something much broader. The guards should not all have uniform responsibility settings. It should be varied. In fact, there should be a few guards who have startlingly low responsibility settings. These would be corrupt guards, which would open up whole new avenues for lawless characters and their interactions with the law. Say you get busted for attempting to pick a lock. The guard who finds you happens to be fairly self-interested and a bit shady, but you don't know this. So you attempt to bribe him a little, to have him waive the fine. You're taking a risk; he might be incredibly devoted to his job, in which case trying to bribe him to get off will only add a higher bounty to your fine or a longer jail term. But if he's low enough to take it, you've improved your speechcraft skill and found a much cheaper way out of the situation without having all your stolen items confiscated. And to help this along a bit, I think guards should all have proper names; that way, you can remember the corrupt ones, and remember that you might be able to get away with slipping them some coin.
4e. There should be a distinction between palace/castle guards and regular guard. The palace guards should be of a higher echelon, men and women hand-picked for their assignment, either by the nobles themselves, or by a palace/castle Captain. Their skills and equipment should be of a higher quality, and for the equipment that means both in stats AND in looks (as in unique meshes and textures). While any city symbols present on the regular guard armor should be present on the palace armor as well, it should be a higher class of workmanship. However, the palace guard should be significantly smaller, limited to door-guards, a few patrols, and one or two personal bodyguards. They should have a small barracks within the castle itself
4f. The regular guard should have a complex all to their own, one that is separate from the palace/castle. It should be fully decked out with trainers, smiths, healers, etc; everything that a guard force might need in case of a major problem. The city's prison should be located in that complex as well; let's face it, what noble is going to want their city's criminal element technically sharing the same abode as them? The guard's quarters should make sense in regard to rank (which would mean that the guard itself needs its own faction with a hierarchy system, regardless of whether or not the player can join it). The lower rank-and-files have multiple beds to a room, and have to share quarters and space. The officer level guardsmen have quarters to themselves, but they are on the smaller side. The captain of the guard has the largest room; perhaps even multiple rooms. There should be a great amount of pride that the guardsmen have for their position, and there should be a slight amount of contention or jealousy between normal guardsmen and palace guardsmen. The regular guard force should be much larger than the palace guardsmen, but their equipment, even for the officers, should be of an obviously lesser quality. For the rank-and-file, they shouldn't even have a full suit of guard-style armor yet, instead having to substitute with iron pieces when necessary.
5. CRIMINALITY Spoiler 5a. Walking the city at night should be a dicey proposition. There might be pickpockets, outright robbers, con-artists, even potentially murderers. The night-time should be a time when the law-abiding citizenry shrinks into their little quarters of safety, and the darker underside of the city peeks out its head. This means that RadiantAI will actually have to be improved upon a bit, to rationally allow for this kind of stuff. So say you're walking around in a less-than-appealing neighborhood. Some Wood-Elf, whose fingers have more of a mind of their own than what's good for them, starts sneaking up behind you. [Not sure how an NPC actually sneaking undetected might be pulled off; the generic sneak animation obviously isn't going to work to hide them from you if their stealth-check is passed. Perhaps they could (cringe) do it like WoW does it, by just becoming invisible until the stealth-check is failed] He attempts to pick your pocket and, depending on his skill or need, pulls free as many items as he thinks he can get away with, starting with the most valuable. At this point, the game calculates whether he succeeds without being detected or fails. If he fails, a top-screen message alerts the player, who can then lawfully attack the thief. Or if he succeeds, the player won't be any the wiser until he discovers that the item happens to be missing. Quest-items and equipped items would be exempt, of course.
5b. The night-guard cannot be in the same condition and the same caliber as the day-guard. Whereas in the day, the guard rotation guarantees there will be a guard in any area no matter what, there should be significant holes in the rotation during the night due to understaffing and the less-desirous conditions of night-work. This way, criminal activities will have brief yet logical windows to operate. While I'm not fond of having something like a permanent location of a Black Market in the game, some form of illicit trade should exist that's different than the "fence" system of Oblivion. Say, during these holes in the night-guard rotation, a sort of secretive criminal co-op is hastily undergone. And depending on the night of the week, it's held at different locations and different times throughout the city, to lessen the chance of someone catching on. Discovering the locations and times of these black markets should be a matter of your reputation with the criminal element of that city, and perhaps your reputation and rank in the Thieves' Guild (if there's not rivalry between the guild and freelancers in cities).
5c. The Thieves' Guild should be difficult to find in such cities. If they are fairly neutral or even cooperative with the freelancers in their area, then finding them should only be possible via winning over the trust of these late-night criminal elements. And winning trust from people who are said to have no trust should be a rather hard thing to do. Or, if the Thieves' Guild and these freelancers do not get along well, then finding a recruiter to the guild should be counter-intuitive i.e. you'd expect to find the guild around such freelancers, but they're not there, and those guys aren't going to help you in discovering how to get in contact with them.
5d. Night-time is also when a different element of shadiness should come forth. Occasionally, a suspicious-looking (and never-before-seen) NPC should be discovered in the area (perhaps like a Fallout-3-esque random event). He should look and feel sinister, in both his physical appearance and his choice of attire. Attempted conversation should result in brutishly nasty and short words along the line of "Get away from me" or "I don't want to talk to you." If the player manages to follow this individual undetected, the NPC will lead the player to some location (such as a house or back-alley, or perhaps even outside the city), where that NPC will engage in an assassination of another NPC (naturally, the victim NPC will have to be a freshly-spawned, nameless background NPC). If the player remains undetected through the assassination, then nothing happens. If the assassin detects the player after completing the assassination, then they will attack the player.
5e. Certain clubs or taverns should only be open during the late hours of the night, and getting into them should be a matter of having a good disposition with the right people or faction. Inside, the player will find commonplace use (and ability to purchase) both moon sugar and the more refined skooma, along with the more acceptable alcoholic beverages. The patrons of these late-night establishments should have high aggression levels, so that if something about the player rubs them the wrong direction, they'll get up (perhaps in two's or three's) and begin to pummel the player. Perhaps scripting should be enabled so that they'll leave the player alone after the player's health has dropped to a certain percentage, and they'll yield w/ dialogue if their own health has dropped to a certain percentage. Highly illicit work/quests could be found here, ranging from smuggling materials to extralegal mercenary work to theft and perhaps piracy. Further, these late-night clubs or taverns could also act as a supplier of equipment and general merchandise to characters that've lost the trust of the legitimate day-time merchants [see 2e.]
5f. In these late-night taverns, gambling should be a constant presence, something the player should be able to engage in. While I'm not a fan of mini-games, the ability to gamble seems a worthy inclusion. When it comes down to it, betting in the arena did nothing for me, and the "game of shells" in the Winged Guar in Tribunal was similarly limited. I'm looking for two or three different games, and they should be along the level of complexity of pazaak in KotOR. It should be something the player could honestly invest money in, something that requires a bit of skill, as well as a bit of luck (which, by the way, could add whole new elements behind the Luck attribute). It should also be something by which the player could win or lose quite a bit of money in. It should also have economy-friendly limits, so that the player could not bottomlessly milk gambling to amass an immediate fortune. Opponent NPCs should be able to gamble only they money that they possess. Overall, there could be one or two serious games, as well as a few chance games (i.e. shells). Furthermore, it might be possible that a dealer/NPC might be cheating, adding a whole new complexity to things.
5g. Late-night player excursions may also be the perfect opportunity for factions hostile enough toward the player to send agents to attempt to end the player's life. For example, if you've irritated the Order of the Hour enough, they may decide to send someone to quietly finish you off during the dead hours between dusk and dawn. It should also be part of the AI to detect the radius of guards. And since guard patrols should be heavily limited at night, it'll prevent those laughable assassination attempts where the NPC draws and rushes the player in broad daylight and right in front of a troupe of soldiers.
6. INFRASTRUCTURE Spoiler 6a. Let's start with things that Oblivion largely had in its cities that gave off a good sense of believability and life: Stables outside every city gate, guard rotations and housing, castle dungeons, the Imperial Prison and the legion compound within, the black horse courier, a sewer system, and most importantly the Office of Imperial Commerce. Why is the last one the most important? Because it's the best example of one who's nature of existence is not tied into some form of game mechanic or quest. Sure, you go there to buy the IC house, but that doesn't necessarily relate to the name of the building, now does it? Including pieces of infrastructure, even if they are things that don't directly affect the player through any form of game-mechanic or quest, adds a huge element of believability to cities.
6b. Now let's look at things that can be added in, starting with the one thing that NO large city would ever function without: a Census and Tax administration. While you, the player, are not going to be paying any taxes, that doesn't mean the citizenry aren't going to be. By having an official building, and by creating some NPCs to man it, you add yet another layer of depth to the world, even if, in reality, it doesn't actually get used for collecting taxes. But tax agents could wander the city every so often, taking their dues. Naturally, quests and subplots could be tied into such a building, but even if none are, having a working tax building is a must-have for any large city.
6c. Now ask yourself, who decides how to build and add onto these cities in the first place? Okay, it's likely the king/duke/count/ruler-person that makes those kinds of calls. But who is actually paid and contracted out to do any building? The answer: Building commissions, or their equivalents. Companies comparable to the Atrius Building Commision or the Vanech Building Commission in the books A Dance in Fire and The Argonian Account. Or better yet, companies comparable to the East Empire Company in Morrowind. Companies such as these need to exist in the world, and they should have at least token offices in large cities.
6d. My biggest complaint about the Imperial City was its road design. It gave the appearance of a large paved road designed for horses or carriages, with pedestrian sidewalks off on the sides. Yet how could they expect horses and carriages to get past the pedestrian-designed steps carved into these roads at frequent intervals? Bad city planning. Design of walkways and roadways should lend believability to their purpose. After all, if those roads were not designed for horses or carriages, then what was the point in having pedestrian sidewalks at all? What was the point in having the roads so big? Why didn't they flatten and shrink the roads, respectively, to allow for better use of urban land for residences? More believable and more efficient city layouts are a necessity.
6e. Have Missions or Embassies, and have them where they make sense; assuming Skyrim, there could be a High Rock mission, a Hammerfell mission, and a Morrowind Mission, each within large cities respectively close to their corresponding borders. These would interject some political feel into the game, as well as some cultural addition (and of course, great potential hooks for quests). I don't really see the existence of a Cyrodiil mission, because that's what the Legion's for.
6f. Actually have the rank hierarchy for the Skyrim guard / Imperial Legion matter. Higher-up soldiers get noticeably better accommodations, noticeably more respect from lower soldiers (at least to their faces), and represent an actual chain of command.
6g. Libraries could exist as competition to bookstores. While most bookstores (in Oblivion at least) held multiple copies of generic books, libraries would be a repository for more unique, harder-to-find books (not skill-books). If, for example, a quest required the player to gain some information from a specific book, and that book was far too expensive at the local bookstores (or if they didn't have it at all), the player could go to a library, pay a modest 24-hour entry fee (like renting a room at an inn), and read the book at their leisure. It would not be a "borrowing" system, like traditional libraries; instead, it would be a "pay to get in, and read while you're there" type of thing. Books could not leave the premises (though a thief could still steal, numerous guards w/ high detection rates would be around, and if caught once, the player would be banned for a period of time). While this might seem rather pointless (the player could theoretically read those books for free at bookstores w/out paying at all, this inclusion would also mean that most booksellers have either limited selection, or they hide most of their rarer, harder-to-find wares in the back. This could also allow book-collecting quests, where the library could offer rewards for the player to find various titles throughout the world.
6h. What is a city without a place to store its wealth? City-state treasuries or banks are a must-have for any large city, and while they are likely to serve as the guardians for the wealth and capital of the rich and noble, any enterprising bank wouldn't turn down an investment from a seasoned adventurer with a fair amount of loot or cash. [NOTE: This idea assumes the return of gold having weight]. Treasuries would allow the player to remove the burden of his/her coins in a safe environment, would allow the player to rent secure storage containers or even rooms for prized possessions, would reward the player a tiny amount of interest for investing capital into their funds, and they would also provide the player with letters of credit (for use as a weightless substitute to actual gold) for a small percentage fee. However, since such treasuries would be devised to accommodate the fairly wealthy, a basic down payment would be required to open a holding account with them (perhaps around 1,000 to 5,000 septims?).
6i. A public square or forum is a must for a large city, as well as a public garden. The Arboretum in the Imperial City did a fairly decent job of portraying a garden, but not too much of a public gathering place. Stone benches should line the square, and tables/chairs set up categorically should be commonplace. Statues and architectural elements should either serve as the centerpiece for a square or ring around it (or perhaps a water-feature in the place of statues). Vendor stalls should line the outskirts of a public square selling food (sweetrolls, anyone?), broadsides should be handed out (and they should range from advertizing leaflets, guild recruiting, philosophical treatises, announcements for the next play, documents heavily influenced by ruling bodies, to documents comparable to The Common Tongue), beggars should be heartily at work, minstrels and bards should be competing (or collaborating), and doomsday prophets and religious zealots alike could compete for an open ear. The garden(s) could be either together with or separate from the square, but they should be a decent sampling of the region's local flora and fauna. Specific sections could be set aside as first-come-first-serve alchemist's resources, but the majority of the garden should be off-limits to harvesting. Every hustle-and-bustle grey city needs its little slice of nature for people to safely get away.
6j. Most cities are not built all at once. Most are instead built over many generations and time periods, continuously adding onto various sections and parts to make the whole of the city as the player would experience it. Therefore, it makes little to no sense to have every city use the same texture for the stonework, woodwork, and what have you. There should be parts of cities that look extremely old (though extremely old should not necessarily mean extremely shoddy). Further, there should be parts that look incredibly new, as well as transition architecture for the in-between building stages.
6k. An Arena should exist, of course, but stylistically different from the one that appeared in Cyrodiil. For one, the player should not be confined to a small little box to watch the action from. They should have to find seating amidst the crowd, and access to closer seating requires more money. Furthermore, most people view the arena as nothing more than a blood-pit. Really, the arena is a platform for all forms of entertainment, and it can serve as a stage as well as a killing pit. Imagine if weekly or monthly plays could be performed there! We could see men and women acting out A Hypothetical Treachery. The Horror of Castle Xyr could be dramatically done by moonlight. Same with Legend of the Krately House. And there's no denying it: You all know you want to see the full performance of the Lusty Argonian Maid. Tickets for the plays, of course, would cost a bit more than tickets for the bloodbath. And think of the numerous Crassius-Cuiro-esque plot-hooks that could be forged from this lighter side to the arena.
7. RULERS Spoiler 7a. If you are the ruler of a city, no matter what your title and power in respect to the whole scheme of things, are you really going to sit around all day and hold court, personally listening to any and all who decide to come up to you? No, you've likely got more pressing issues to deal with than that. That's why you're going to have stewards to deal with all that mundane work, who will then contact you if something major enough should come up. In other words, you can't just walk off the street into a castle/palace, take a few steps, and then be able to wave and chat up the person who has near-absolute political control over the immediate (and perhaps even less-than-immediate) area.
7b. Assuming this, the throne rooms for the respective rulers now need to be located away from the main entrance of a castle/palace. Instead, there should be a large secured foyer that is frequented by stewards for that ruler, rather like Castle Skingrad was designed. If you wanted to see Janus Hassildor, you didn't just walk in to see him. You got in touch with his steward, who then went and contacted him for you. However, whereas Hassildor had a rather unique reason for doing this, rulers in TESV cities should practice it as a matter of course. Why? Well, because they're nobility and you're not. Perhaps if your reputation with the nobility faction were high enough, they would welcome you and hear you, but that reputation needs to be difficult to attain. In order to gain this reputation, perhaps the stewards would need to be coaxed via favors (quests). In fact, I would find it much more realistic if you had a major problem that in reality was worthy of a ruler's immediate attention, but their steward judged it to be less worthy and either turned you away or asked you to "prove your loyalty and worth" via tasks.
7c. Ruling families need to be present, and they need to provide yet another layer of complexity. Say you've managed to gain audience with the King of Winterhold to talk about something vital to the main quest or a side quest, but he is opposed to you. End of story, right? Ah, but say you know how to play the gossip game rather well, and you learn that the King's son, the Prince, is not fond of some of his father's decisions. Upon following that up, you learn the Prince is actually slowly plotting to usurp the father, and in return for some favors or assistance, will gladly provide you with what you need for that quest. Have royal families, have nobility, have those families sometimes be united and strong, but also have them sometimes be fractured and divisive. And have those politics leak and filter down into the city as well. In a city with a divided nobility, have agents within the city, perhaps attempting to coerce public thought or accomplish certain strategic tasks secretively. Have a populace who only hears rumors of power-struggles among their leaders every now and again, and can only theorize about what's going on. Have people who have a preferred side or stake in the issue.
7d. Further, ruling families should have extended families that put the region(s) surrounding their bloodline's city to industrious use. Rather like how cities in Cyrodiil had counties as territory ascribed to a city, so too should city-states in Skyrim. And what better use of such a land-bounty than to give it to sycophant relations to benefit your political, economic, and military power? Large homesteads, strongholds, and manors should dot the immediate and not-so-immediate countryside, and each should be with its own purpose, either implied or explicitly noticeable. Perhaps one such land-grant would be for defensive purposes, giving a minor lord with some tough followers a small barony between their home city and a rival city. Perhaps another such land-grant would place a relative in charge of another small barony involved with harvesting natural resources, such as timber or ore.
7e. Nobility exists as the friends, strong allies, and loyalists towards the ruling power in the area, and noble families are a must for city-states. Further, nobles are often in deep competition with one another for power, influence, and favor with the object of their loyalty, meaning potential for an even thicker political atmosphere and politics-based questlines. Also, just as rulers have spies and agents embedded in the surrounding city-state, nobles are bound to do the same to collect useful information and subtly sway public opinion. And, depending on the social structure Bethesda wishes to create for TES:V, perhaps nobility and rulers would have a whole separate law-code to abide by, or at the very least have access to legal perks, favors, and loopholes that the common citizen certainly doesn't.
Almost all of these ideas are very interesting and seem like they should have already been thought of!
To add on though, in the nobility section:
Each city has acres and acres of land outside of it. instead of just having "outdoors" I guess, there should be 4-6 noble families in a province as you suggested, and each noble family should recieve their own castle, which in turn has a fuedal bond with 3-4 surrounding villages, maybe a priory and farms or mines etc. that are in the area. Usually it would be a peaceful realm of hard working underclassmen and vigiliant knights protecting them. The Noble families should have a head, who has their own priorities/lifestyle. He/she could be a duke/duchess grateful to the divines for what he has, chivalrous, caring and generous to those under him - and respectful to those above. OR he/she could be careless of those under them, abusively working them and overtaxing them coldy. All of the rulers -as all are, they are unique- and their personalities range inbetween. The chivalrous leader would provoke love and admiration from people, and his realm would see mostly peace - with the occasional farmer being terrorized by wolves(quest for the lower levels to deal with) or invasive, migrating goblin tribe(quest for the higher levels - note: they don't have to be goblins, maybe ogres or bandits or any
group of enemies really - and the reward would be money or a title in some chivalric order [ie the knights of the white stallion]). This being said, the greedy duke/duchess would see the same problems, but hire out mercinaries or take care of it themselves. eventually from their poor rule, the lower classes begin to grow restless and unloyal towards their lord. To the point where one day the adventurer will wonder into the inn or local hangout of one of the underlaying towns and over hear a group of people murmering words of dispite towards their lord just before a sip of ale. it would then become a topic, and continue from their, the quest could continue to, as Diviner said, the player can try to find and replace the current leader with a more truly noble family member - or (off topic really quick, there should be 1 of each unstable soverignty such as this in each province, and the adventurer may choose one to place himself upon the seat of power in) he could start a rebelion agains the duke/duchess to o'er take the castle himself.
The options are both difficult in their own manner. The option of going secretively to replace the current leader with a family member would require an adventurer of subterfuge, of smooth speech and intelligence. If the replacement is succesfull, the populace in the surrounding area will love the adventurer, and the new leader would feel respect and a debt of gratitude towards the adventurer. The new duke/duchess would go around to their nobility friends and tell of the adventurer's noble deeds, increasing his
fame in chivalry -go down to see a1. The option of rebelion would be difficult, because the adventurer would have to try completing it with as little loss of life as possible. The lowerclass - pretty much any citizen not living directly in the castle itself (maybe 10 peasant warriors would spawn approch you when nearing the caslte instead of using actual NPC's) would storm the castle. However be weary - for the over-bearing duke/duchess will have a blood thirsty group of knights highly loyal to his fat pockets, and will slaughter then whole of the popluation revolting and revel in it. Should the rebelion be successful, the count will be banished, the adventurer would immediately be loved by those underhim(that may or may not change based on his coarse of action), who would tell all whom they meet in their travels about your glorious victory, increasing your
fame in honour -go down to see a1. The overthrown noble family would take on the adventurer's last name, and the adventurer would have to be weary of them for they will be jealous, and attempt to take back their power, honor - and their family name.
In the event of failure when proceding with the more laying-low kind of way, the prince that was to replace the noble will be excecuted for treason, and you will immediately get 2 infamy and a bounty of 500. until the event dies down, (give it two weeks or so) the lower class men in that province will protect you - in otherwords so long as you remain in the area controlled by the count you attempted to overthrow, but not in the castle, the guards will not attempt to arrest you. once it dies down your bounty will disapear, but your infamy will not.
In the event of failure when proceding with the more aggressive approch, that would mean at least 8 civilians died before you took complete control of the thrown room(as Diviner said again, would be placed in a more strategically defendable position). Your infamy would increase by 10, and a bounty of 10,000 on your head. again, the people in the direct area would protect you, but be weary of the black knight patrols. The really is no "dying down" for a straight up rebelion, so the only way to get rid of your bounty (and infamy in this case) would be to attempt another rebelion. however the villagers would not be so ready to rebel for fear of their lives.
a1 - So this is my idea that I kind of threw in there as well: a differentiation between status. example: Fame would be Chivalry, Honor, and Celebutant added together. you attain chivalry through quests, honor through glorious battles and celebutant would very slowly increase as you talk to more and more people (that respond positivley) throughout cyrodill. Infamy would be cruelty, cunning, and crudeness. Cruelty would be attained through carelessness on quests(acceidentally or purposely killing the innocent), and blatantly murdering or stealing so that all may see and increase your bounty. Cunning would be only noticed by fellow thieves/assassins really, but the infamy it is totalled into is noticed by everyone still. what I mean by that is this: infamy is what the public sees you by, but having a high cunning would earn the respect of anyone in things that would be like the dark brotherhood or the thieves guild, and even rogue thieves/murderers. Cunning goes up when you assassinate some one big or pull a larger hiest. (I just thought of this, by maybe you could assassinate dukes/duchess too as a part of the more secretive path). Crudeness is the exact opposite of Celebutant, the more people that respond negitavely to you when you talk to them, the higher your crudeness(which also grows very slowly)
PHEW. I typed ALOT :/
-=edit=-
btw sorry, whenever you see me giving credit to "Diviner" I meant thatonguy, you probably already guessed that, but sorry haha I was looking at his rank??(I think) and not his name haha