Friday, May 31, 2013

Monthly Update

Well, I was going to wait until next post for this, but I've been busy with my volunteer project and don't have much to show in the way of the horror text adventure, so now is as good a time as any for the monthly update.

I think what I'm going to try for next month is a continuation of the horror text adventure. I'll try to solidify more of it into a legit design doc, settle on the way things work (at least for a first pass), and start prototyping in my usual clumsy way.  Maybe *gasp* I'll try some pseudocode first to make sure I understand what I'm doing.

Tuesday, May 28, 2013

Development Log: Horror Game Part 6

Let's skip the design notes today; they're slowing down anyway and it's time to test some of these ideas rather than philosophize about them.

So today I'm sketching out making safe rooms static and elaborate so players can meet in more than tiny one-room campfires, which was my first thought when the idea of this game hit me.

The player never sees this.  It's all in their MINDS!
We'll start off with the setup of a first year safe zone, being an elementary school.  I made it small, compared to some, perhaps, but I think it's fairly accurate to how my elementary school was set up, as far as I can remember.  Growing up, however, I lived in the middle of nowhere, and my one small elementary school took kids from seven towns, and we were proud when we reached a population of 200 students, faculty, and staff.  So it seems small at first, but I think it'll be large enough to house new players and get them used to interacting with others and learning about safe zones.

Supposing there were a 365 room cap per year, and there were multiple safe zones in each year, the safe zone rooms wouldn't count as part of that room cap.  Otherwise, 43 of the rooms would be used right here.

On top of this map, I've also begun throwing together a static room description list for this area, so it can easily be c&p'd when I program it.  Right now the descriptions get redundant, especially for large rooms and corridors, but I'll work on them to make them more descriptive and subtle, while trying to eliminate that "Text Adventure Vocab" that creeps in as I write room descriptions.

Anyway, this school would be a basic setup, probably for year one, and possibly with slight alterations for year two or three safe zones, before letting the school become dangerous in later years (and it would lose cohesion completely so it would become randomized).

Players would be able to congregate in various rooms to chat, and I'd probably offer other things to do as well, if I can come up with them (library books, for instance).  You can be sure that some players would enter a classroom and then type in the chat "plays Tic-Tac-Toe on the chalkboard" and stuff to pass the time.  That would be the kind of emergent social gameplay I would encourage but not enforce.  To do this, it would need to be in the room descriptions, like "The chalkboard is filled with games of Hangman." Etc.  Ooh, that's a good idea.  Time to work.

Friday, May 24, 2013

Development Log: Horror Game Part 5

More notes, this time starting with a list of possible physical areas, then exploring themes for scares once more, using children's horror books as examples, and sliding around the mechanics to see what takes precedence.  I think I like the idea in the last paragraph, of making monster be more of the continual ticking clock threat (a ticking clock where you don't know the time!), and puzzles be the real thought-provoking challenges.

Scraps and Notes and Ideas for Horror Adventure

5/23-5/24

Remember that the common places were the scariest in Silent Hill.  Elementary school = scary. Sewer != scary.

So, with that in mind, think back to being 8.  2nd grade, approximately. What scared ME?  What did I do, what were my hobbies?  What do OTHERS do? Since I lived in middle of nowhere, the experience is quite different.  Perhaps just start with my life, and work my way up to other things as I think of them.

Common Places:
School
- Bus ride to/from
- Class itself
- Lunchroom
- Recess
-- Swing Set
-- Slide
-- Pavilion
-- Monkey Bars
-- Kickball Field
- Nurse's office
- Bathroom
Friends
- Friends' houses
-- Friend's birthday parties
-- Sleepovers
- Friends' back yards
- Friends' woods
-- Forts/treeforts/snowforts
Home
- Back yard
- Woods
- Flower Garden/Vegetable Garden
Town Center
- Church
- Town Hall
- Library
- General Store
Babysitters' Houses
- Back yards
- Woods
- Field
- Barn
Cub Scouts (don't remember much?)
- Meeting Hall

Common Hobbies:
Movies/TV
- Ninja Turtles, Power Rangers
Books
- Goosebumps (discovered it in 2nd grade!)
Videogames/Computer Games
- NES, PC games

Special Yearly Activities:
Halloween/Fall
- Hay Ride
- Apple picking
Birthday
Christmas
- Visiting Relatives
- Stockings
- Waiting for Santa (did I believe at 8?)
Easter
- Easter Egg Hunt
- Visiting Relatives
- Almost Catching Easter Bunny (did I believe at 8?)
New Year's
Fourth of July
Dentist visit (twice a year)
Carnival

Was I 8 When...?
- Hiking

FEARS
- Zombie in the closet
- Monster under the bed
- Neighbor's evil dogs (Dobermans or something)

Now stop for a second. A.) Mostly woods, since that's where I grew up.  Life is completely different for a city kid.

Grr.

Well, anyway, new idea: what if what is a safe zone changes depending on year?  Since Elementary School was awesome, but Middle School sucked, for first few years the safe zone looks like a school, but later the school becomes dangerous.  It would lull the player into a sense of safety and think that the school motif is okay, and then later strangely there are no other players in the school, and monsters start appearing, making it terrifying and quite a shock.

So let's start with this idea of making common ground safe zones, like a full school structure, and work on randomized danger zones later?

Stop one second.  I'm reading Stephen King right now, and it's making my heart skip.  Let's look back at good ol' R.L. Stine for some lessons.  Was Goosebumps actually scary?  Some were.  Welcome to Dead House, Night of the Living Dummy, The Girl Who Cried Monster, The Ghost Next Door, The Haunted Mask...

What are the themes of these?  Because that's what really scares children. In order, we've got Ghosts, Toys coming alive, Parental distrust, Ghosts, and Losing control.  In fact, Night of the Living Dummy and The Girl Who Cried Monster are BOTH about losing control, as well, so we've got three stories where the kid loses control of a situation, and two more are simple ghost stories.

And what is a ghost, really?  A spirit of the dead, yes, but what do they DO? Sometimes you've got a poltergeist, but usually they're just something that you don't understand, and they're scary because you don't understand them.  Fear of the unknown and loss of control.

Move to the Scary Stories series by Alvin Schwartz. Besides the nightmare-inducing pictures, what were the scariest stories about?

Disturbing premises, for one: things off-kilter in a big way to set the mood, like a boy finds a big toe in the ground, and takes it to his mother, and SHE COOKS IT and THEY EAT IT. WTF?!

Ghosts, ghosts, ghosts, ghosts... Not being believed...

Also, expectations of fear! Some are about haunted houses, so people expect scary things to happen, and The Girl Who Stood on a Grave has the idea of hands coming out of the ground even when none does, she still dies of fright.

Premonitions... hmm, how would I work that one in?  Not just premonitions, but foreshadowing of any kind that the character understands, like in The Hook, the radio reports a prison break.

Modern scares, like High Beams and The Babysitter.  Work in telephones, somehow?  After all, that bit in Silent Hill was disturbing.  Phones can definitely find their way in, though I wouldn't know how to add cars.

So, all told: primary themes of these stories: Fear of the unknown, off-kilter premises (the world takes care of that?), loss of control, and expectations of fear/premonitions, which can also be interpreted as tone, to some extent.

So fear of the unknown is taken care of by exploration (kind of the opposite, but the tone makes players explore almost in morbid curiosity, scared to do so but knowing they have to), and what is left is loss of control.


How do you show loss of control through a game where you get the thrill of solving puzzles and such?  Perhaps if monsters aren't bosses in themselves, puzzles are, and the monsters are unkillable creatures that chase, and all you can do is hope to make it to a safe zone in time.  I think that will help tie things together.

Friday, May 17, 2013

Goin Hikin BRB

Off to the Appalachian Trail for a week.  When I get back, I shall continue my design work on my horror text adventure to finish up the month.

Appalachian Trail

Tuesday, May 14, 2013

Development Log: Horror Game Part 4


Well, as usual, more notes for today.  The theme today is mostly on the order of expanding the scope to accommodate a wider slice of life, and how to incorporate such things properly.

Scraps and Notes and Ideas for Horror Adventure

5/14

What about the seedier aspects of childhood; the kinds of things that aren't "normal" and are much more difficult to overcome?  If the "normal" (white middle-class suburban) stuff is terrifying, what to do about gang violence, sexual abuse, drugs, and things like that that some kids end up in?

Is there, perhaps, a way to make areas of the world that are even WORSE than normal, perhaps where there are fewer safe zones, everything seems more terrifying, perhaps more monsters, and you aren't likely to survive at all?  Perhaps the tonal shift to these areas is steep, so a player knows long before entering that if they go that way, voluntarily, they aren't likely to make it out alive, and there are warning everywhere to try to steer you away, but should you enter anyway, you unleash hell.  Once the door is open, it can't be closed, and the game gets ten times harder.

If a player has a proper amount of fear, they will steer clear of these areas and make it through each year without messing themselves up. A player who is wild and enticed by the prospect of it all, finding, somehow, it to be intriguing, makes things worse.

Or, perhaps, could their be a way to either randomly generate the chances of a player coming across these "opportunities", so one player might experience a less dangerous life than another (hey, just like real life), or perhaps offer difficulty settings at the beginning? I never really liked difficulty settings and have always been a fan of dynamic difficulty adjustment, but I wonder if there is a place for settings.

There must be a better way, surely.

Of course, being that this game is made by me, I can only create what I know, and my experience is on the lighter side of these things. If the complete experience were nothing but a white middle-class suburban experience, how would that effect the player? Considering this entire game is purely metaphorical, would the player ever notice?

Perhaps the best bet is indeed to offer an in-game "hard mode", above, where the player gets warning after warning not to open the gates of hell, so to speak, and they make the decision.

The tough part always is about how to make death and fear mean something.  The hellish option should be one that most players would choose to avoid, yet the desire for exploration is so wrapped around the whole concept of videogames (and even this one, it's a major mechanic), that it becomes tough to warn away players.  If the warnings are too comical and large, they'll think they're ironic/sarcastic/untrue warnings, and that they should be following them as part of the game, and if there isn't enough warning, players may unwitting stumble into those areas and get themselves killed.

So a balance must be struck between subtle but fair warnings.  An average player is likely to see them, and the scary nature of the rest of the world itself makes players not want to open those doors.  Players on the extreme ends of the spectrum will either miss the warnings and mistakenly open the doors, or see the warnings and laugh at them.  There's nothing that can be done about those, because when you fix one end you make the other worse.

So things will just have to be balanced.  It will take a lot of playtesting to see how such a thing turns out, and, of course, players interacting with each other is a wild card.  How to mold players to want to help others? Perhaps as long as most players are that way, the jerks don't have too much influence.

Or, perhaps, like how each "year" gets locked, perhaps there is a way to lock the doors of hell behind a player, so they can't get back?  Maybe, or maybe it depends on what they represent.  I kind of think that's a bad idea, and locking things out is already a little annoying.

Suppose there is no locking of years, while we're at it.  A player who is 16 can still interact with a player who's 10.  But, given the exploratory nature of the game, the player who's 16 is physically far away from a player who's 10, and unlikely to meet again. The player who is 16 is most likely using later safe zones than the 10 year old, and so unless the 16 year old has an oddly looping world, he's not likely to swing by.

And perhaps the rarity of players who DO do that allows me to have my cake and eat it too: SOME older players can meet younger ones, but it's not likely, individually, but guaranteed overall.

But to this hellish extra stuff:

If we take the "normal but upside down" approach that I mentioned last time, where the world is NOT all doom and gloom, just subtly creepy, would these other areas BE more doom and gloom?  Or find a way to make the tonal shift still not so drastic, with perhaps it just giving off a much scarier vibe.  Or offer warnings along the lines of "You can hear screams and shrieks and cries of pain and howling wolves through the door", etc.

Now, would the average player say "Er, yah, let's try door number two instead" or would the average player say "Ooh, sounds dangerous, let's see what's screaming"?  It should be the former, and no matter what you may do to steer away the latter, they are a lost cause.

You can't please everyone.

But heck, if I'm spending time to create that section of the world, I might as well make SOME use out of it, even if I shouldn't.  I won't encourage its use, but if someone falls through the cracks, well, that's life, right?

Friday, May 10, 2013

Development Log: Horror Game Part 3


Today is another post of design notes like Part 1, random thoughts and so on.  Sooner or later they'll be solidified and put in order.

This one mostly tackles the idea of monsters and player death.  It also gets into some questions about how to make things scary, what horror means, and considers videogame and movie examples for justification.

Scraps and Notes and Ideas for Horror Adventure

5/9

Consider how monsters truly work.

Should there be many, or very few?  Should they work basically as bosses only, and most of the world is puzzles?  Perhaps a monster roams the "year" freely, and gives chase when it spots you.  That way you are always running?

Perhaps that is that case, where there is one monster per year, and can't be defeated just by fighting them. Throughout each year, the player slowly learns some skill, or perhaps discovers how to defeat the monster near the end of the year, and once the monster is defeated, the player can move on to the next year.

Perhaps some years there is more than one monster. Maybe +1 monster each year? Although no two monsters can be defeated the same way.

The point of the game, of course, is survival, or part of it anyway, and you feel far too in control if you can easily defeat lots of monsters.  And even if you can't defeat them, and all you do is run, the more there are, the more actiony the game feels, and less scary.  Alien was scarier than Aliens. The fewer the number of monsters, the more weight each one gets.  Hence, Zombie movies aren't scary.

But there also needs to be an ebb and flow to danger, and I'm not sure if that's taken care of by safe zones.  Maybe you always want to be a little on edge outside, but you shouldn't necessarily be fearing for your life at all times.  Having too much of that is draining and overloads the player.

Gotta be very careful about overloading, because that's the biggest problem with horror games and movies that aren't scary.  After all, every great slasher flick has ONE killer, no more.

So, heck, would one monster a year be overdoing it? Considering how by the time you get to monster #10, you aren't scared at all?

Perhaps, each player gets ONE monster that chases them throughout the entire game, one monster that represents a culmination of all fears, and it's the puzzles the player has to solve combined with that one monster giving chase that create that sense of panic and fear.  "I gotta get this done and escape the level before the monster gets me!" etc.

Perhaps in year one and even two, you never see the monster, but it's alluded to (you hear it in another room, you see evidence of it destroying stuff in its path) which leads up to a creepy, chilling atmosphere.  So that the player DREADS meeting the monster long before he sees it.

Hmm... if each monster were unique, or unique enough per person, how do you kind of prevent too much metagame chatter? "My monster is blah blah", because that might ruin the illusion.  Unless each player has a "personal demon" which is more part of the world lore.  That may work.

It especially works with the yearly separation, because players in the same early year shouldn't KNOW about their demons for a little while, so they have no way to compare notes.

By the time the player gets to later years, there are fewer players anyway, because many of them have died.

MAYBE.

Seriously consider what death means in a videogame.  I play permadeath, but many players would HATE to get far and die, and throw a tantrum.  Not to say that's the core audience, but I would hate for any player to get frustrated by dying.

How can we make games that are horror and even SURVIVAL horror without ever truly being in danger?  Perhaps there is always an escape route, and the game is always fair to you, and you can even be injured without death? There has to be a solid way of freaking out the player, bringing them to the edge, even dangling them off, without dropping them.

Perhaps, as danger approaches, drop the player hints about how to escape.  Lead the player on to get them to survive, without making it completely obvious.  If they can't take the hints...?

Because this game is so high tension, it's no Dwarf Fortress, and losing is NOT fun.  I can't see how I could die, have to create a new character and start all over, and think it's all in good fun.

And there is NO WAY this is not permadeath.  If the player dies and can resurrect easily, there is no scare.  How often has the mood been completely killed in Silent Hill by dying?  Lizard boss in SH1, Apartment meeting with Pyramid Head in SH2, dying in the mall in SH3. All of these suck because you die from the slightest of errors, and then you suddenly come back.  How is the fear of death scary if I just saved?

Now also multiply that feeling of annoyance from because it happens later in the game.  Just like any awful MMO, death is a waste of time.

So instead of thinking about what the character goes through, let's take a different tack:

What does a reader go through when reading a horror novel?  Sure, sometimes fear for the lives of the characters happens, but the bigger fear is what? Tone and setting things up creepily?  Wasn't one of the scariest Stephen King moments in Bag of Bones when the protagonist is talking to his dead wife through knocking?  It made me shiver!

So it's best to make everything as subtle as possible.  Make the WORLD itself less fantastical.  Perhaps the more "normal" the better?  Which was scarier though, in Silent Hill, the "normal" or "nightmare" world?

Quite frankly I don't know anymore.

I think perhaps the scariest moments were the places where it wasn't quite nightmare world, but was rather the real world tipped a bit.  Like the "nowhere" level of SH1, which was real-world, but the map was messed up.

So is it perhaps more that I should be making the world less creepy than I think, or much more subtly defined, rather than full of blood and macabre atmosphere?  Because the randomized nature of the world takes care of a lot of the "messed up" nature of scary worlds.

Remember that Terminator was scary because it took place in "present day", and that only those brief glimpses into the future were necessary, because any more and we'd be flooded with post-apocalyptia (OMG "Apocalyptia" is an AWESOME word!), and that was part of what made Terminator 4 suck.

Similarly, Road Warrior was a great action flick, but Mad Max was closer to "normal" and it was more dramatic.

Remember there is horror is silence!  A lot of the best drama is visual.  Consider giving the world character through subtle detail but imagery that is realistic yet creepy.  And find a way to make safe zones feel safe but only in a subtle way so as to keep the overall tone with only slight shifts.

Crap, the biggest test for this game sounds like my literary skills, not my programming skills!  Oof.

Tuesday, May 7, 2013

Development Log: Horror Game Part 2

Well, not too much today.  I've just been considering how best to split up the interface.  Considering it's a text adventure, essentially, but with extra elements to it, it might be helpful for the player if each element was separate so the player doesn't have to dig through text to find the relevant info they need.

What I'm thinking might work.
Although my quick Flash test was four boxes, I think six boxes might be necessary, or at least the most useful.  This seems to be fine, considering that the sample text I'd used in the Flash test was very small and tons more could have fit in those text boxes, which means I can cut their sizes down without issue.

The one major thing I've been considering it whether to develop a text parser, or to rely on a menu-driven system.  Of course a robust text parser would ultimately be the best, player-side, a menu would be easier for me to make.  Ah, that's the way it always is, of course.

Friday, May 3, 2013

Development Log: Horror Game Part 1

Well, I spent the week making design notes more than anything, in an attempt to flesh out and experiment with how the game world will work.

Not all of what I wrote matters; probably most of it will be trashed, but that's part of what the experimentation is all about.  I'm trying to find a way to make a solid horror experience where the player gets sucked in to the text and forgets they're playing the game.

But the notes don't cover everything, and they assume you know what's going on in my head, because, of course, I wrote these notes for myself.

So, in essence, what I had in mind before writing my notes was this:

A horror text adventure that would be online, where the player may meet other players in specific areas.  Where players meet would be "safe zones" so they can chat, discuss what they've seen, etc.  I might offer a few "Channels", such as one for Newbies and Mentors, and one for Roleplayers who like to keep in character.

But beyond safe zones, the world would be randomly generated for each player, so players could not meet outside of safe zones.  In the world, it is creepy, dangerous, and you are very, very alone.

The goal of the game is escaping the nightmare world.  The player is trying to find a somewhat-legendary exit, and must explore endless rooms to do so. ("Rooms", of course, being text adventure vocab, and not necessarily literal.)

What follows are my actual notes, which meander and babble, and just have random idea after random idea, expounding on each for a bit before getting off-track.

Scraps and Notes and Ideas for Horror Adventure

5/1-5/3

Player ideas:

Player is a child ages 8-18 -- Scares metaphorical of puberty, growing up, sex, schooling, abusive parents/authority, responsibility.

RESPONSIBILITY -- As you get older, you are given more mechanics, have to do things more on your own/take care of younger players.  You start off with little to be able to do, but as you go you have fewer people above you and more below, and you must guide and defend others/show them the ropes.  But as a result the world gets more dangerous for you, making the game much tougher as you gain more responsibility.

You win if you survive long enough--Once you hit a certain age, perhaps, the possibility of escape is made.  Age, however, is constant, so the game is essentially timed.  You make it to age 18 and the door opens, hit age 19 and the door closes.

So you have 1 year to get to the end, once the end has become available.  However you start at age 8, so you have 10 years to survive before the door even opens.

YOU CANNOT TAKE ANYONE WITH YOU.

Once you have survived all the way, your character gets "hall of famed" whether personally or globally.

Is age in realtime?  #Turns? Fraction of realtime?  Don't want to use anything like exp or leveling because some players may figure out tricks and beat the game fast. # Rooms discovered?

3650 Rooms for 1 room a "day"!  As long as you don't discover anything new, you stay same age? :(  But I guess it does have advantages. It allows for "sprints" where the player is trying to go from safe room to safe room.

So safe rooms are not quite random, but you only discover safe rooms at intervals, and everyone discovers the same safe rooms?  So like Players who find the first safe room, that is the same safe room for everyone?  Or have multiple saferooms in each "year" that might randomize, for instance. -> * problem here is I like the idea of older players helping younger ones, and cutting off between years doesn't work.

Lock off each year with a "boss" be it monster or puzzle (or monstrous puzzle?), which in turn also changes the theme of the year.  (One might be a puberty theme, next might be authority theme?)  Or interweave minor themes together.  Make corresponding locations as themes, e.g. school/church = authority, woods = puberty, etc. (But think them all through very well).  You get a mix of each every year, but some years get more of one than another?  Perhaps research that kind of stuff, as well as diffs. between male and female!  Perhaps use statistical chances?

In the last year, you are slowly stripped of your friends (i.e. your path becomes more solitary).

Each time you log on, the description of yourself may change slightly as you age.

Player has opportunity to gain skills through:

Practice - Player could take up an instrument and practice, getting better at it
Exercise - Player could run around a safe zone to increase speed

Hmm... I wonder though, should "stats" ever come into play?  Should things be quantified in that way?  Of course, PLAYER should never see them anyway. But should there be stats hidden to the player that effect the game?  Little things that the player wouldn't necessarily know to do?  Like running from a monster for long periods as it chases you would increase some kind of endurance stat-- but what if we did some Cthulhu type stuff where prolonged exposure to monsters drives you crazy?

And the crazier you go, the more monsters you see? That might translate poor, especially if the player doesn't know.

What would sanity do? Insanity?  If a player's sanity went down "all the way" what would happen?  We wouldn't want to have them see other players as monsters because they would try to kill them.  That would be a bad experience to innocent players in safe zones.  Perhaps if insanity went down all the way, the layout the player has been discovering the whole time is erased, and replaced by more fixed rooms that always lead to death?  Or perhaps there is one safe room, and if the player goes there and stays there long enough, they gain their sanity back?

But what would such a room be like?  The ultimate comfort of a real-world bedroom, perhaps?

And what would a breakdown like that REPRESENT?  If this is a metaphorical game of growing up, what does it mean?  Does it mean an emotional breakdown, or something else?

For that matter, what do the monsters mean?  If monsters that represent the confusions of puberty go away after a time, they get replaced with monsters that represent later teenage fears -- graduating, going to college/getting a job, the pressure of parents.

I guess confusion is taken care of in a sense because the player, as s/he goes, memorizes the layout of the gameworld, perhaps finding a hub that they are comfortable with.

But how do you represent picking a college with a monster?  Perhaps that is something for more intricate details.

In fact, maybe special life choices like that are "bosses" while general stresses, like high school classes and grades, are more normal monsters.  However, they must ALWAYS be carefully described, so they represent these but are NOT obvious.  They CANNOT be obvious.  They should be scary and original, and a player would have to seriously read between the lines to get it.

Back to insanity, because that's a mechanic--

Let's see, being with players, as in peers, raises sanity and calms you--or only if you're an extrovert?  Let's suppose you can pick whether to be an introvert or extrovert--to be an introvert means non-multiplayer safe zones restore your sanity.

Not sure there :/  Wouldn't want a player to go insane because they're in a multiplayer safe zone and an introvert.

Unless safe zones are complexes, with beds and lots more; they are more like fortresses than single rooms, so you can go to a bedroom and sleep or you can chat with friends by the fire.

So even if you're introverted or extroverted, being in a safe zone is universally a good thing.

Maybe if you're one or the other, should you get a warning saying "you need rest, find a safe empty room" or "you feel the need to talk to someone"?

Maybe not that, in fact, you no longer need an introvert/extrovert character, because if safe zones cater to all possibilities, the player will do what is natural to them.

So safe zones are definitely complexes.  Perhaps even tell the player they are in a safe zone?  Not in those words, but would something like "You feel safe here" still be too gamey?

Should I let the player know they are safe--only if so then through NPCs, or descriptions that speak to safety (the building is clean and warm and friendly-looking, brightly lit and unshadowed, so nothing can hide).

Back to insanity--losing and rearranging the world isn't something with too much real-world reference, except basic disorientation at having a breakdown and not knowing who to trust.  So perhaps instead of a world-rearrangement, the world creates a bee-line for the closest known safe zone, and everything else is cut off.  You make a headlong rush to get there, evading monsters that have previously been the ones who have knocked the most sanity out of you, since those are the monsters you fear most (maybe, but what if you are unafraid of a monster and know you can kill it, so you do not run, and you laugh in its face, and let it try playing cat-and-mouse with you.  The player does not really fear it, so sanity should not go down.  So how can that possibly be represented in game?)  Maybe don't use time facing it, use harm caused or an equation of harm caused/time, or something.

Maybe no matter what you do, you slowly lose sanity if not in a safe zone, you just lose MORE when monsters are around, you discover a new place, a new theme, etc.

Perhaps you gain sanity back when you defeat a monster, solve a puzzle, etc?

Perhaps if you go insane, yet survive and restore your sanity, your sanity meter is changed, either you are quicker to go insane later, or perhaps you build up endurance so it takes more for you to go insane.  Perhaps the latter if you don't go insane, but get close to it and get back just in time?  Hidden vars, of course.

New idea -- how to make puzzle solving something randomly generated like rooms?  Perhaps come up with a few types of puzzles that can vary in their details but basically play the same.  Like if the puzzle were a rube goldberg machine (for instance) then there would always be ten pieces, and piece number 3 could be randomly generated as one of four different pieces, but they all ultimate do the same thing (connect piece 2 to piece 4 properly).  I would just have to be careful to make that kind of stuff work with all variants.

Alternately, puzzles could be less "in the world physical" and more like brainteasers (Silent Hill 2's riddles, or Sudoku-type puzzles).  Could also have puzzles be of the adventure game "find the inventory item and put it where it needs to go" style.  If there is variation but logic to them, there could be something there.  Like think of Raiders golden statue scene where he uses a bag of sand as a weight.  You could have six different objects to put there of the same weight.  In your game, one of them is randomly picked as the item that appears in your world.

There is always debate on those kinds of puzzles about whether you should have the item first or find the slot first, but a randomly generated layout leaves it up for grabs.

Think about a way to do what tabletop RPGs do: make the player feel as though this is THEIR story, and not just a blank slate.  Try to figure out how to make what happens feel like it happens only to them.  But think about how that works if players in game safe zones were to talk about their adventures.  They can't sound all the same.  One player talking to another might say "You'll never believe what I ran into!" and another player will NOT have encountered the monster, or is very unlikely to, because the details of the monsters are changed enough so that even if they play the same, you think it's YOUR demon.

Find ways to ask questions at character creation?  Or at some other point, maybe all during first year, you pick your traits and hobbies and likes and personality which gets displayed later.  I'm kind of annoyed by pre-creation, so what things MUST player choose?

Gender might be required, and some kind of basic description of character (for the growth part), although maybe the player can fill that in whenever.

Find a way to integrate things so perhaps first year is semi-random, but as the player goes on s/he develops a play style, needs, etc., and these are saved to help make further year stuff.

Completely randomized is easiest, but player-based MAY feel more customized.  But would player notice?  Would completely randomized be better because there is LESS chance of two players having similar experience?

Consider having some X number of monsters/puzzles/etc. per year.  Not predictable.  Perhaps player does not necessarily run into every monster/puzzle per year?  Maybe the number of each switches depending on first year experiences? Or ongoing experiences?

Is there a way to encourage friendships and even relationships between players?  Or player-character's complicated relationships are represented by puzzles and monsters?

IS THIS GAME ABOUT: Discovery? Survival?

You CANNOT solve the mystery, or at least the point of the game is NOT to figure out where you are.  For all you know, you live in this world, and there is escape, at the end, but you do not necessarily try to figure out WHY you are there.  Perhaps fill the world with lore of escape to make the player keep exploring to escape, so they don't sit in the same place forever?

As for the older player-characters telling younger ones and guiding them... does that really MAKE SENSE???  How often would that be?  Think of your peers, perhaps unless you have a sibling, you'll stick to kids your own age, so unless one player "adopts" a younger one as a younger sibling, that's not likely to happen.  And I don't know how that's possible technically, so Perhaps the yearly bottlenecks will work best.

Of course, players who are faster than others would get older faster, and would leave their friends behind.  But in an abstract way, that happens in life, not that people get older faster, but some certainly do mature faster, or move on and find new groups of friends.  Your friends at age 8 are not usually your friends at age 18.