Quote:Original post by brent_w
Realism and Immersion are not the same thing.

I think you're technically correct, but in most arguments for immersion a realistic depiction of things is what I hear being argued for, be it involving NPCs with complex lives or the the accurate scattering of light to simulate an atmosphere.

Quote:
A player is perfectly capable of being immersed in an extremely unrealistic game.

Can you give me an example? What game would you call very immersive without being realistic?

Quote:The wait function was not designed for the purpose of getting more cash on vendors. It worked for that purpose, but that's not what it was for.

What would you say that it was for in Morrowind? If I remember correctly, you couldn't heal with it in town, shops didn't have day/night cycles, and you didn't regenerate anything. You could use it to wait out bad weather, but not for someone to return or anything.

Quote:Original post by Kest
As badly as they design some gameplay, Bethesda are masters of immersion, so I think Oblivion would be the best example for this. Magic is a good example of an unrealistic element that doesn't break immersion.

I should have made my point better because what I'm trying to get at is a slippery distinction. Your next point is perfect for it.

Your original point was talking about realistic behavior and I sort of slipped into realistic graphics (sorry for that). What I was trying to argue was that even if we're talking a fireball, you'll get people arguing for how the fireball should cast light, which in turn should affect shadows and reflections, rather than whether or not the darn fireball should support complex gameplay. But OTOH, if someone calls something like World of Goo immersive, I'd be impressed (and want to know why).

More to your original point, I think arguing for complex NPCs with webs of social relationships and day and night cycles because it's realistic misses the opportunity of adding such things because they create complex, layered gameplay. Even if someone asks for it because it makes them feel like the world is real, I'm still going to be the cranky old man if you as a player can't leverage it in terms of gameplay. A shop owner going to sleep and having a family should present for me a series of strategic opportunities or it's a waste of time to put it in.

Quote:
While that line is pretty easy to see, a more complex issue is gameplay abstraction. Is toting around fifty weapons and twenty pairs of iron armor pants, while still being able to hop around and fight with a sword realistic? No one seems to complain about that.

Excellent point. I think when we think about something like this or your original point we have to go back to what the purpose of the game is in total. If an RPG is supposed to make the world come alive and put you in a particular role, there are far better ways to do it than simulated lives, especially if their only purpose is either realism or immersion.

Quote:Original post by JasRonq
A fireball with perfect light casting and such is great, but its purpose is to make the fireball not scream at you how unrealistic it looks when you see it so you can get on with enjoying how it makes people scream when they burn to death.

Agreed, this is a good point in terms of not being distracted. But I'd take a little distraction in favor of more options. But the trend in games has been for high fidelity and low options.

I was actually thinking about this recently as to how complicated the AI should get (wheteher realistic behavior is worth the effort vs the usual precanned choreographed stuff game companies have been doing forever). Does it add that much to the game (for the usual people who buy them) -- can it be as exciting when its really mundane behaviors done correctly.

Planners and such get extremely expensive (cpu-wise) for only modest increases in complexity (assuming you do it consistantly). The more complicated the simulation gets the more $$$ will cost the game companies to develop (tools as well as the game specific data) because of the emergent behavior (all the endcases that the players behavior might trigger). That is likely to be the greatest factor keeping AI improvements from happening.


Yet you can have simple improvements like a daily schedule that is semi-randomly varied at least can make the NPCs look less like lifeless mannekins running on tracks (if they move at all). The player shouldnt be inconvienced too much (as when the blacksmith is out sick for the week). But a varying time cycles (or situational variances) where different things happen (and different opportunities occur for the player) can make the game more interesting (and still be mostly choreographed/vetted). Several cycle intermeshing can make the environment look different (and yet can be pretested reasonably well).

Some challenges might be presented in the mundane environment (with the player having to spot the bits that dont belong).

The players have to be given multiple ways of solving their current problems and the solutions should be mostly logical and obvious (not too arbitrary). Different skills might be used for different approaches to solutions with the situation consistant enough that the player can plan it out (few sudden shifts of situation). Major shifts when they happen then should offer oportunities to the player (Town being attacked is a great time for a little looting...).


Have the NPCs involved adjust their behaviors to macro situation factors as well as be reactive to the player (actions). An NPC who the player attacks should fight back and call help from friends (if any) or run away and hide or various other possible reactions (depending on NPCs 'detail'). Dont ever get into a fight with the Blacksmith in his own shop.... More complex interactions with multiple NPCs can lend more interesting result for overt acts.

Larger macro game adjustments can be controlled by AI -- like spawn patterns (where , what, how frequent) in reaction to a macro plot (possibly/hopefully shifted by the players actions in the game). The game could use the standard 'dumb' mechanisms locally but the placements of the 'pieces' could vary in a logical way (eliminate repetative grindings on 'best' solutions).