To RNG or not to RNG

[gadjung]

The core issue, to me, is the colony death spiral.

Yup. Game is designed from beginning (Tynan please correct me if i'm wrong) in a way, that colony is set to be non-sustainable long-run, the end and main goal of game is to build ship to prevent this annihilation.

As for RNG i would not say that it should be influenced by storytellers because that will make it even harder to pick for most players which difficulty they want. More 'configurable' RNG might be a compromise here.

[9ofSpades]

1. Should the game have a such a thing as bad luck outcomes that's not induced by some obvious, non-pressured, voluntary player decision? Or should I make a universal design standard that nothing bad ever happens unless the player actively induces it or makes some clearly-traceable mistake to cause it?

When I hear "bad luck outcomes" I interpret that as you saying a scripted event that will bring about a challenge or negative result to the colony. If I am interpreting that correctly then this reminds me of the Crisis events in Stellaris. A scripted yet randomly (confined to late game) determined event that radically changes the games course. I personally would enjoy such an event in rimworld BUT think it should be a selected parameter during the map setup. On the other side I feel like there is no need to make universal standard where nothing bad happens. It is my opinion that a sandbox mode or Chillax setting should be adequate. Perhaps adding the option to remove certain types of threats (e.g. disease, raids, blight) could be a method to find middle ground.

2. Should I just ignore some classes of player feedback as simply not linking up with what RW is? Are some players worth leaving alone to try to make a game that's different from the usual assumptions? Even if it leaves them pissed off because they interpreted a story generator as if it were a skill test?

[Not a game designer or businessman]!! but a product that is perfect for everyone is impossible. So any product will inevitably have to be tuned to a particular population. So it is my opinion that as the game dev you should seek to make RW the best that it can be, following first your own desire for the game while continually considering the many fans that have come to love RW. Some players may never be satisfied (like those who desperately want z-level construction of which I was one for a long time) so there are players voices who you can consider outliers of your normal fan base and dismissible. Now how you balance the artistic pursuit of making RW how you envision it or maximizing its profitability is not my place to advise you.

3. Should players be able to consistently avoid losing people/resources even at high difficulty? At any difficulty?

This is a difficult question to answer because it is hard for me to remove my bias against what I like about RW when trying to consider what other fans may enjoy. For example, the many excellent and beautifully designed colonies that you share on twitter or are posted on Reddit are well beyond my skills in RW and not what I enjoy creating.
At high difficulty I want to sweat when a raid comes and be at the edge of my seat trying to hold back the overwhelming forces but I never enjoy losing pawns to events like disease or mental breaks. Those are a necessary part of the game. I would enjoy being able to (within most circumstances) avoid loss for events like disease or mental breaks. However, if a raid puts me in a position where I cannot avoid loss during a disease outbreak or mental strain then I would be content with the loss. A fine tuned position where a loss is (nearly) guaranteed at some points but reasonable to handle on its own is where I would like RW to be. When it comes to play difficulty I would suggest adjusting that gap, according to difficulty chosen, (i.e chillax has very small chance of loss to insane where you expect a loss)  between almost guaranteed loss and just a stressful but manageable threat would be ideal for me.

4. Is there a way to set expectations (relative to the whole game, or relative to a given difficulty level) to encourage players to accept some degree of randomness to game outcomes? Or will they always reject this randomness and demand to be rewarded in accurate proportion to their skill/effort?

I feel my response to #2 & #3 should lead you to expect that for me the description given already should lead people to anticipate randomness and look forward to dealing with the challenge or blessing. Since I don't find my expectations unmet by the descriptions I don't think my feedback is as useful.
I feel that it is part of the RW lore as the story of your colony unfolds sometimes things just go wrong and I enjoy that but for those players who are wanting preserve their year 7 colony I think they also should be able to negate RNG as much as they deem necessary. Hopefully that is possible and not just a bunch of theory crafting.

[I Am Testing This Game]

1. Should the game have a such a thing as bad luck outcomes that's not induced by some obvious, non-pressured, voluntary player decision? Or should I make a universal design standard that nothing bad ever happens unless the player actively induces it or makes some clearly-traceable mistake to cause it?

Temporary bad things should happen randomly.

Permanent bad things should be a result of player mistakes. (This mistake includes "engaging in direct melee / ranged combat.)

2. Should I just ignore some classes of player feedback as simply not linking up with what RW is? Are some players worth leaving alone to try to make a game that's different from the usual assumptions? Even if it leaves them pissed off because they intepreted a story generator as if it were a skill test?

You should take into account what Rimworld actually IS and has been throughout the betas / alphas.

How did people play it all that time? Did they play it as a permadeath story generator? Or as a base builder? An RTS?

Are you finishing Rimworld, the game that existed in the betas / alphas? Or are you making a new game?

3. Should players be able to consistently avoid losing people/resources even at high difficulty? At any difficulty?

Yes

4. Is there a way to set expectations (relative to the whole game, or relative to a given difficulty level) to encourage players to accept some degree of randomness to game outcomes? Or will they always reject this randomness and demand to be rewarded in accurate proportion to their skill/effort?

Yes, you can get players to accept randomness and permadeath.

However the game has to be designed around it.

For example, if you had 40 dwarves in your colony, with some being grunts, each random bad result hurts you less, and is more acceptable.

If you had 6 colonists in your colony, all with an important specialty, each random bad result hurts you more, and is less acceptable.

In Rimworld, I start to accept permadeath of newly recruited grunt characters in the later game, as colonists have become more disposable. But at this point the game is starting to break down, mundane non-combat tasks have  become extremely tedious, and there is no more story or ability to focus on individual pawns. In the early game, with low pawn numbers, when you can focus on individual pawns and their stories, they are also often irreplaceable and vital specialists.

Or, if the results of losing a finger are relatively minor, like a work speed decrease, people will be more tolerant of losing a finger. But, if the results of losing a finger are more severe, like a permanent failure chance, regardless of skill, which makes a skilled, injured character worse than a novice, people will be less tolerant of losing a finger.

A permadeath bad things happen story generator would need to have more granular outcomes, where you could see a lot of bad stuff happen and still have the colony continue.

Permanent 5% loss of colony efficiency, with the prospect of making it up? Ok. Permanent 25% loss of colony efficiency, with no prospect of making it up? Well, you'd need a darn good set of recovery mechanics to get people to accept that.

Or, if you wanted people to accept that colonies permanently fail due to bad luck, it would need to be less time consuming and tedious to restart.