Return to “General”

How superstitious should NPCs be?

Not at all -- NPCs doing dumb things would weaken the challenge of the game.
Total votes: 6 (8%)
A little bit -- it would help make the social world feel more plausible.
Total votes: 63 (81%)
Very superstitious -- it would be fun to see NPCs doing crazy things.
Total votes: 9 (12%)
Total votes: 78
Post

Re: Very Superstitious

#16
I actually see what I'm calling superstitious behavior in NPCs not as something to be programmed explicitly, but as an emergent behavior... and possibly a very cool one.

if NPCs are given the ability to form beliefs about the connections between their actions and world-events they can perceive, that creates the possibility that some of those beliefs may be wrong. Superstitious behaviors may emerge naturally from a sufficiently robust AI.

If so, the real trick will be not be deliberately adding artificial stupidity, but tweaking the degree to which NPCs can correct their own superstitious beliefs. That's pretty much what I was trying to get at with the poll in this thread.

And the value of this mild degree of brokenness is in creating a world that feels more plausible because it's not filled with inhumanly perfect robots. That emergent ability for NPCs to get things wrong sometimes, I believe, isn't just good for gameplay fairness -- it's more satisfying to be in a world filled with people who feel more human because they aren't perfectly rational all the time.

Assuming we're not all robots as Thymine suggests. In which case... never mind.
Post

Re: Very Superstitious

#17
Flatfingers wrote:Assuming we're not all robots as Thymine suggests. In which case... never mind.
McDuff suggested in another thread that if the player and other agents in the game were AI, that they'd be mind-uploaded personalities. So it makes sense that these agents would exhibit the same irrationality and stupidity that we do; they're just our consciousnesses hosted on a different substrate.

Even before McDuff suggested that, I was thinking along the same lines anyway; it wouldn't make sense for an irrational, human player to be controlling a supposedly perfectly rational AI. And in any case, the whole idea of a perfectly rational agent is almost paradoxical, since perfect rationality would lead inevitably to suicide.
Post

Re: Very Superstitious

#19
The main advantage I could see for the notion that all characters are AIs isn't gameplay; it's that it's useful for explaining away incongruous NPC behaviors.

I don't hate the idea. I'm just not sure it's needed if we can explain unexpected NPC actions as superstitions or other understandably human-like activity.
Post

Re: Very Superstitious

#20
Flatfingers wrote: I respectfully disagree with that. I think there are innumerable examples in the real world where -- unless one is a practicing "everything is connected" Buddhist -- some events are clearly not causally connected. Believing that they are is, I think, properly understood as superstitious behavior. (This isn't an assertion intended to go after any particular beliefs of anyone here, BTW. I'm speaking generally.)

This may be even more true in a constructed reality like that of a computer game, where the set of possible actions is constrained to a finite set by what the programmer allows through code. Emergent effects may be hard to predict, but atomic actions are still limited, as are the specific world-effects with which a character action can be correlated. Because in many cases it can be possible to know, through the logic of the code, what caused some world-effect to be produced, it's possible to know whether the character's action was the primary cause for generating that effect.

If it wasn't, and if the character believes it was, then that is a superstitious belief by definition. As the Pythons might say, " human perspective don't enter into it." ;)
(Hope I'm not necroing, not that that terms applies here). In a game with emergent gameplay, the line between action and effect can get unusually long. As the dev video has shown recently, even tiny perturbations of the system (tiny being relative, of course) can have dramatic effects on, say, market prices. Depending on how many CPU cycles you choose to invest, it's indeed possible to show that "everything is connected to everything", black boxes/closed systems excepting. Now the point of rational entities is to filter out all that noise and deduce proximate effects (what actions of my own are likely to cause the positive or negative effect that I observe)? Entities do this with varying levels of efficiency, with less efficient entities calling the other stupid/"sheep", and more efficient entities calling the other superstitious/paranoid.

(That is, in fact, a pretty good description of how conspiracies in real life come to being).

Then again, there are odd things like this: http://www.catb.org/jargon/html/magic-story.html ... not exactly pertinent, but a fun read.
Post

Re: Very Superstitious

#21
One of the many things I like about this forum is that it dispenses with the peculiar notion that thread "necromancy" is somehow undesirable. So no worries on my account!

Back to the notion of connectedness -- the most salient reference here might be that of the "clockwork universe." Everything operates by purely mechanical, physical rules, everything is physically connected, and if you could comprehend all the rules, as well as the state of everything at some moment in time, you could predict with perfect accuracy the state of everything at any future time.

As a theory, that's interesting. As a practical guide to prescience, it's useless. The number of states of things we can know, and understand and retain, is extremely finite, and the number of physical rules we know and whose effects we fully grasp is likewise extremely limited.

So even if everything truly is connected physically, that fact has no practical value because we can't apply it. Beyond the most narrow fields of action and windows of time, we cannot comprehend enough of the universe to perceive enough of the connections to never, ever make any mistake of causal belief. In fact, relative to the universe, our human ability to form accurate causal beliefs about the things we can see in our world peters out extremely quickly with increasing distance in space and time.

Note: I believe things can be worthwhile even if they have no practical value. I'm not a complete utilitarian! But we are talking utility when we talk about how to apply ideas of connectedness and understanding in a computer game, so the question of functional applicability is valid.

With respect to LT, I'm not saying NPCs "should be" programmed to form superstitious beliefs based on their inability (like ours) to grasp all causal connections. I'm wondering out loud if they might -- if they have a limited but real ability to form goals, perceive aspects of their world, and connect those two through their actions -- be capable of forming beliefs we'd reasonably call "superstitious" because the cause and effect don't have any actual direct consequential connection to each other -- they're just two things that happened close in time to each other.

Again, to me, that's not a failure mode. Within reason, that's emergent awesomeness. :)

(P.S. I bought the book version of ESR's Jargon File many years ago. A lot of its clever, funny, and amazing stories are still relevant, though many of them, such as "gorilla arm," have become legends that not enough designers remember.)
Post

Re: Very Superstitious

#22
@jimhsu

Can't link as I don't have much time and on my ipad but you have no idea much that's true (the run-on cause and effect thing).

The paper that I've posted elsewhere on which Josh is basing his own economic design reported that even changing the optimisation parameters on the compiler they used had a significant effect on the way their economy operated, due to microscopic differences in the way floating point numbers behaved.
Post

Re: Very Superstitious

#23
Superstitious npc's are fine, as long as it's obvious that what they're doing is being done for superstitious reasons. If you made an npc that could react superstitiously, 9 times out of 10 the player is going to think the AI is fucking up and broken.
woops, my bad, everything & anything actually means specific and conformed
Post

Re: Very Superstitious

#25
Question: Shouldn’t the fact that an NPC is superstitious depend on its personality? I mean yes a lot of people have stupid superstitions. But there are some like myself who have none, unless you count in my cynicism.

Also should not the incline towards superstition also be distributed unevenly, geographically speaking? On earth you see that the poorer and less educated a country becomes the more the population is inclined to believe all sort of nonsense.
Post

Re: Very Superstitious

#26
Superstition as a personal emotional assocation sounds wonderful, if the associaton is at least relatable. I feel that this would be accomplished by introducing a memory system into the world which NPC's tap into, which at high LOD gives each a personal memory and associated significance to objects and events. I posted an explanation in this post, which ties into the present thread at the last paragraph.
Post

Re: Very Superstitious

#28
Hadrianus wrote:Question: Shouldn’t the fact that an NPC is superstitious depend on its personality? I mean yes a lot of people have stupid superstitions. But there are some like myself who have none, unless you count in my cynicism.
You are simply not aware of them.
woops, my bad, everything & anything actually means specific and conformed
Post

Re: Very Superstitious

#29
Hadrianus wrote: I mean yes a lot of people have stupid superstitions. But there are some like myself who have none, unless you count in my cynicism.
LOL.

"I have perfectly logical beliefs, you have some quirky ideas, he's a superstitious nut."

I invite you to consider the possibility that nobody actually considers themselves to be stupid and irrational, and weigh up the chances that you, alone of all born men, have attained the super-humanity required for you to be the first person to correctly analyse his own mind perfectly.
Also should not the incline towards superstition also be distributed unevenly, geographically speaking? On earth you see that the poorer and less educated a country becomes the more the population is inclined to believe all sort of nonsense.
I dunno, as a member of a first world nation myself and a fairly well travelled individual, belief in nonsense like "my country is the best country" or "people from other countries are to blame for my problems" or "there is such a thing as a free market" or "democracy leads inevitably to maximised freedom" seems to happen much more frequently at the top where people are rich and coddled than in poor countries where they don't have as much insulating wealth between themselves and the brutality of their ruling classes.

I suspect this is simply a case of you classifying the sort of errant nonsense we all take as read in the west as "sensible", which of course would make your statement axiomatically true. Proof you can do anything you like with the right axioms.

Online Now

Users browsing this forum: No registered users and 19 guests

cron