Re: realism and historical accuracy in video game design.
This discussion resurfaces every time a new medieval-inspired fantasy game is announced. The two sides almost always seem to talk past each other.
There is a sense in which a hyper-realistic realistic game would be boring, because it would involve so much tedium. For example, if you want to convey the character renting a room at an inn, sitting down to eat and drink, and then changing clothes before going to bed it would not necessarily make for riveting gameplay to have the player spend several minutes to manually go through the motions. In cases such as these, unless we're talking about sim-level games, I think people would largely agree some degree of abstraction is permissible.
The other concern in favour of dismissing realism and historical accuracy seems to be that researching and creating believable and accurate representations of real objects consumes much of the developer resources that could be spent elsewhere. Again, I think it is intuitively agreeable that major mechanics need to be fluid and working so that the game has some semblance of completion before resources are spent on other things. This is of course not relevant in the case where the studio employs people specifically to deal with aesthetics or research.
However, in order for a game to be immersive, which constitutes a significant part of the experience when it comes to games as a medium, the game ought to give you the opportunity to suspend your disbelief. For example, I can accept that there are dragons in Skyrim because they're basically timelords from another dimension. However, I can't really understand how my horse is capable of traversing mountains at above 90 degree angles. This is not believable, because the game doesn't offer you a reasonable explanation. There are basic physics in the world of Skyrim, and they seem to apply to all races, and nowhere is it suggested otherwise.
Furthermore, there are certain "hinge propositions", on which our understanding of the world relies on. Things we just take for granted, like Newtonian or Einsteinian physics and everything else assumes this. Video game universes usually accept these propositions, although we might ask if we should presuppose that they do, as sometimes these presumptions can be false. For instance, Hesiod tells us that it takes
ten days for an anvil to fall from heaven to Earth, and the same time for it to fall through the Earth to the underworld. We might assume that this means Earth is equidistant from heaven and the underworld, but the ancient Greeks had a very different understanding of physics and some phenomena were thought to simply be dependent on where they happen in the world. Maybe the anvil falls faster in Tennessee than it does in Greece just because Tennessee is not Greece, so we can't draw the aforementioned conclusion.
However, it doesn't seem reasonable to forego fundamental assumptions about the world when we play video games, and you don't really have to play long before understanding that barring technical/mechanical limitations, pretty much ever single game is based on a contemporary understanding of physics.
This argument of believability extends into, in this case, the concept of castles. The information we have is something along the lines of: small centers of population, but infinite amount of people. So you might ask how do they feed this many people? There's only one season in the game, but you can clearly see fields and similar being represented in the game. Is this a reasonable abstraction? Perhaps, but everything still points towards individual objects being accurate representations of the real world. What I mean is that humans look human, they eat the same food we do and exhibit similar behaviour to real humans. Why, then, is there a disconnect when it comes to the representation of castles?
I think nobody would have difficulty to accept inaccurate descriptions of castles if the game didn't take itself seriously and didn't aspire to create accurate representations of reality.
This leads us to the following point which is that in the games industry "realism" is often used as a catch-all term, similar to RPG. It's also pointed to as a scapegoat for many poor design decisions. The developers might have a specific vision for their product which is to some degree objectively disagreeable, and they might try to legitimize this vision to their customer basis through vaguely formulated expressions of realism or historical accuracy. This creates a lot of confusion in regards to what realism in games really means and is probably a source for many disagreements over whether or not games can gain something from being realistic.