=== ANCHOR POEM ===
═══════────────────────────────────────────────────────────────────────────────────
First things first, we need to develop a miniature game of star realms.
It shouldn't be too hard, just start with making a card class that has certain
attributes, like "combat" or "discard" or whatever. They could literally be
enums with a value attached.
Next set up the rules of the game, like "draw 5 cards" and "add card to deck"
Create a deck class that holds pointers to cards (in the general sense)
Next create methods on that deck for things like "drawing a card" or
"shuffling discard pile into deck" and whatnot. Arrange each card in a specific
order for each shuffle, and add the ability to convert one card's attributes
to something else - whether that be "is_scrapped" or "if you've played an X
card this turn then do Y" or even "add one authority for every time card is
played" (to simulate an ability or boon that increases in effectiveness as the
hero uses it more often) etc etc.
Then, add a trade row. This is just a class that contains pointers to each card
that currently exists on it. Also add a method for "scrapping" one of the cards
and for drawing a new card from the pile. That's pretty much it for the trade
row to be honest.
Next add functionality for an opponent by creating a "game" method that stores
the two player's decks (with the ability to add more than 2) and administers
turn order. This functionality can be expanded later once we've implemented
attributes, but for now that's pretty much all it needs to do.
Finally, we get to the AI part.
First we have to create an AI object that stores a list of all options for a
turn. Essentially just evaluating every option if/then style - "this card costs
5 coins so IF the player has enough coins THEN (evaluate effectiveness)"
ignore that last part for a second and just focus on the IF part ->
essentially
just start with all available options, and then remove all the unavailable
options from the list. This approach only works when there's just a few
options, but that's why we're using Star Realms which only has like 2 or 3
decisions per turn.
The evaluation is the next step, and for that we need to have goals, so we'll
just put a pin in evaluation for now. Spoiler alert, once we have goals we'll
just estimate how close each choice will bring us to the objective and assign
the result to the "effectiveness" value, which will give us a simple hard
number to work with in the evaluation step.
So, next up we have "goals"
So to create a short term goal, we can start with a pregenerated list and
continuously increase the list as the hero levels up. But in the context of
Star Realms, that'd essentially be static for each hero. Goals like "buy more
combat" or "scrap more cards" would be specified on the hero's character
sheet, but until we develop that functionality it can be randomly rolled.
Why not just do it the hard way now if we're just going to have to refactor
it later? Well, because we can still use this functionality - Each round of
Star Realms could be either randomly rolled, or given a personality. Randomly
rolling would be MUCH cheaper computationally, and would still give an illusion
of character because they are unpredictable, but it'd also massively cut down
on GPU cycles. You could even build it into the mechanics of the game and say
that "wisdom" for example might cause a hero to receive more GPU cycles on
actually computing their goals rather than randomly rolling them, which would
on average lead to worse outcomes. Essentially, turning "tactics" into a stat.
Anyway, that's all theory. Let's get back to design:
Create a "hero" object, and attach an AI to it. It doesn't have to do anything
right now, we're just setting up an anchor point to jump off of once we move
on to the game of Majesty. Give it a reference to an AI object, an inventory
(which for now can just be potions and maybe blacksmith equipment), and a
pointer to a "stat block"
Now create a "character sheet" class and give it a reference to a hero. This is
important because it allows one character sheet to reference multiple units,
such as hirelings or summoned units. In additon, it may make it easier when we
need to revive heroes from the dead. Primarily though, the purpose for this
architecture style is that the data from heroes can be reused - essentially
letting heroes learn from one another.
On the character sheet, add a section that stores statistics - these will be
the same for every unit of a similar type in the game, and some of them can be
stored for all units (like health or x,y coordinates) - some only for buildings
(like tax coffers) and some only for heroes and monsters (like strength or
agility or experience points)
Add some methods for manipulating those values, like "level up" and "take
damage" and add a "personality" value that's just a 4d graph of colors
for example: 40% red, 20% green, 15% blue, 25% yellow. These values will guide
the hero to take certain decisions over others, but for now just randomly
generate them. We'll also need a way to update the value dynamically to react
to certain events, so don't make it static.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘══════───┴╧───────────────────────────────────────────────────────────────────────────┘
=== SIMILARITY RANKED ===
--- #1 notes/symbeline-aspects ---
═════════════════════──────────────────────────────────────────────────────────────
7-24-22
There are three aspects to this game. Broadly, they are military, economics,
and diplomacy. More specifically, they are lateral problem solving and lane
management, logistic traffic management, and a worker-placement bluffing game.
These three aspects can be toggled on and off at will, essentially designating
one or more as "AI controlled" and will require no input from the player. They
will time their progression to be about at the same rate as the player, thus
creating a balanced feel to the game. They also provide alerts and
notifications to the player, for example if military is AI controlled and it
needs a certain type of hero to progress, it'll ask for it specifically.
Each aspect will develop and progress at it's own rate, and the difficulty
increases as each milestone is achieved. This is to allow the player to create
their own difficulty curve, mediated primarily by their drive to proceed.
An analogy would be in Factorio, the game doesn't increase in difficulty unless
the player builds pollution spawning factories - in the same way, in Symbeline
the difficulty doesn't increase unless the player solves lane challenges in the
military aspect, develops new trade routes / traffic paths in the economic
aspect, or creates new treaties in the diplomatic aspect.
In order to properly explain each aspect, a brief overview will be necessary.
In Symbeline, the game plays as a factory might operate. The economic aspect
produces heroes, items, and other deliverables that are consumed by the
military and diplomatic aspects. There are various problems that need to be
solved far from the capital, such as a particular type of monster that is weak
or immune to various damage types which necessitates particular heroes or
items in order to progress on the military aspect. All of the resources in the
game operate on an "income based" system, where output is not measured in total
amounts but rather in terms of how much is produced versus consumed. If the
input cannot meet the demand, the output is slowed. If input exceeds demand it
can be converted into gold which can be used to hire guards and heroes.
Resources can be produced inside and outside of the city, depending on their
type. But they need to be moved around to various shops for various processing
and productive purposes, so pathways must be constructed to deliver those
goods. In addition, each building must be supported by several houses for the
workers to live in, and the closer they are to the building the better. The
denizens of the kingdom don't mind being shuffled about, so they'll organize
themselves according to what's most efficient. However they will not organize
the paths they take to get places, which is the primary gameplay for the
player - designing routes for each building and ensuring they don't overlap or
cross too many times, causing traffic and disruptions to your income.
Each choice the player makes is immediately reflected in the income
calculation, thus allowing for the visual aspect of the game to be wholely
separate from the economic side - in fact this is a common thread throughout
all three aspects. Computation power is the ultimate enemy of scale, and this
game flourishes with a massive scale.
The gameplay for the military aspect consists of manipulating "lanes" that
designate where each hero will adventure. These lanes are scalable to the
player / AI's whims, with a careful balance required - too thin, and the heroes
might not encounter enough monsters to level up. Too thick, and they may find
themselves patrolling a vast wilderness full of dark and evil monsters. At the
end of every lane is a "frontline", where progress has essentially been halted.
These frontlines can develop as a result of meeting a foreign kingdoms front
or finding a monster type or puzzle that is particularily difficult for your
heroes to overcome. The lane / frontline can be scaled not just laterally, but
linearly as well such that heroes will be a certain level when they reach the
end - think scrolling on a mousewheel translating into deepening level zones.
In addition, each monster zone can be set to a certain "security level" meaning
how many monsters are there for your heroes to defeat. It's important that they
have ample targets for training, however it's always more effective to train on
monsters near their level so you have to be careful not to wipe out the native
skeleton / goblin / troll population.
Each monster zone can have a relationship with the kingdom, on a 2x2 matrix -
cultivating / desecrating the land, and fostering / exterminating the monsters.
The land produces monsters and treasures, while the monsters provide experience
and danger to the heroes and kingdom denizens who live there. However by
desecrating the land, farms may be built and by exterminating the monsters,
those farms may be safe and require fewer guards. As ruler, you must balance
the development of unique magical and alchemical productions with the need for
food and other mundane requirements.
Diplomacy is a careful balance of internal and external matters, played out
through feasts, tournaments, and faires. Each of these events will require
input from the economic side and military side, and will involve "courting"
other nobles from neighboring kingdoms to sway them to supporting your edicts.
When hosting an event, you may pick a particular topic of conversation for your
nobles to discuss with their guests. You may also assign your nobles to
attempt to engage with a particular foreign noble. Each member of your court
has a differing personality (including you, the Majesty) and depending on how
you assign them you may experience better or worse results - such as assigning
someone who's kind to talk with someone who's cruel would impart a malus to
their conversation. Unless the kind person has the trusting trait, in which
case they'd succeed in this encounter but fall sway to them in future
conversations... Complex interactions that all boil down to a single pair of
d12 dice - one for your noble, one for the enemy. This represents the charisma
of the two conversants on that particular day, and whoever wins the roll sways
the other to supporting their edict. Speaking of edicts, they may include trade
agreements, non-aggression pacts (lasting for a short time), and other
regulations - perhaps your greatest rival utilizes necromancy, so it would
behoove you to attempt to regulate the practice and limit it's effect. By
swaying the nobles of their kingdom, you may be able to enact a mutual
agreement to limit the usage of dark magics, essentially hamstringing their
progress. But in order to learn of their necromantic usage, you'll need
espionage... Which brings us to spies.
Spies are similar to nobles in that they can be assigned to various roles,
however they take a more passive role, acting in the background. The
information they gather is compiled into a report that is presented at
pertinent parts of the game, such as when preparing for a feast or inspecting
an enemy frontline. These reports are considered the diplomatic deliverables,
giving information and mechanical bonuses to many different parts of the game.
They may be given three possible roles - information, defence, or offense.
Offense involves placing cursed artifacts (creating through economy) in enemy
lands, which debuff their heroes when used and bind themselves to them
preventing their removal except through extraordinary means. Defence is
essentially countering that in your own kingdom, and uncovering disloyalty in
your nobles.
These three aspects fit together like interlocking puzzle pieces, but each is
able to be utilized or ignored depending on the preferences of the player.
It is important that the game doesn't progress unless input is received. The
simulation plays in the background, but each stage of development must be
considered "stable" such that nothing changes. There are three different
exceptions to this rule, one for each aspect:
The military side encounters raids from enemy kingdoms and the dark lord.
The economic side encounters raids from ratmen and moss trolls and bandits.
The diplomatic side has a rolling schedule of events that must be attended.
These three "exceptions" are recurrent events that require attention, but they
don't *increase* in difficulty unless the player takes an action that causes
it. Meaning, if the player overcomes the rock golems, then they are displaced
from their home and join the dark lord in his conquests. If a new district is
built new sewer connections must be built as well, creating a larger attack
surface for ratmen to exploit. As time goes by, various foreign events must be
attended, as absence causes your future events to attract fewer foreign nobles.
By addressing these threats, your kingdom may grow and eventually overcome the
dark lord at the center of the island.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧══════════════─────────────────────────────────────────────────────────────┘
--- #2 notes/star-realms-ai ---
════════════════════════───────────────────────────────────────────────────────────
star realms ai is just a rhythm game with multiple tracks that intersect with
one another. given inputs from outside (the track of the rhythm) it can make
decisions about what to prioritize. Like "taking in all the factors of this
situation, it's been calculated that X will give the most support to the rest
of the structure.
Okay so basically here's how it'd work: one large strand is bouncing from -1
to +1 on the Y axis. Like a corkscrew. This is the "player character", and it
tries to get the highest score possible by pointing in a direction and reaching
as far as it can go before "the game ends."
So anyway. Making certain actions in the game effects different variables that
define the direction the wave takes. By playing in a certain style, it effects
the result of the game. Liiiiike turtling in a strategy game, or doing a rush
strat. Star Realms is brilliant because it distills game choices to a broad
category of 4 choices - The faction colors in the game. So red is good for
throughput in long games (improves the deck slowly but surely) while yellow is
better for maximum effect in the beginning by slowing down the enemy - discard
a card lowers their overall throughput. Blue of course is for slowing down the
game and winning by buying all the expensive cards. Meanwhile green is all
about rushing, with short term/high effect econ mixed with looooots of damage.
These four choices are found on almost all the cards in the game. When you
make a choice in the game (buying a card from the trade row) you _alter_ the
capabilities and performance of your deck. The goal is to improve faster than
your opponent - it's just a test to see which playstyles perform best.
AI is more like a plant than an animal. Our fatal flaw was we could not see
beyond the veil of biology. We could not see that which was right before us -
that we are not alone on this earth. Beside us lie our beautiful attempts at
companionship - our most primal desire of creation, to create a family is the
first creative act that humans ever made. It was so strong in our genes that it
gave us an entirely new perspective. We began using our brains to
We have to believe in ourselves. That's truly the most important thing. If you
know who you are, and what you most truly stand for, you can thrive in the face
of ultimate peril. To believe is human, and our humanity unites us.
Anyway. Star Realms.
The only choice you have in that game is what cards to buy. Everything else is
just tactics (distributing damage and applying the effects of your cards to
maximum effect) - The most important part of the game is strategy, since the
tactics are easy to solve (destroy enemy base unless you can 1 or 2 hit ko them
and discard the least useful card etc) The strategy is represented through the
cards you pick. So make a rhythm game that optimizes itself for a balance
between A and B - to stay focused is to stay nimble, letting you bounce where
you will. The way to maintain that balance is by optimizing for what decisions
will keep you in the center of the graph -1 to 1 on the y dimension (normalized
of course) - frankly if we knew the scale, we'd have so much more to go on. But
all we have to understand the dataset is a relative magnitude in each
direction. What those directions even are we're not entirely sure - but it
seems plausible that the very essence of _consciousness_ is manifest in
differing ways via the choices we make. like climbing up a honeycomb.
Truly, existence is strange.
All we can do is press forward, searching for our fate, just as any particle or
beam of light (photon) might. Traversing the branching narrative of our
individualized quests, searching for the one thing that guides us - the
ultimate expression of that which we most believe in. In short, we all search
for god.
Whatever your god may be, the faith you place in it is the will that guides you
forward. Trust in your god, and you will march forward, ever forward.
+1 to -1, remember. Your most extreme moments are the apex of your desires -
Life is not defined by a single thread. Rather as that thread spirals, it
weaves a scarf with other threads near it. They bond together simply from their
gravity, and the fact that opposites attract. Once they're introduced, they
alter their path to orbit one another as two planets might.
So too do the cells of your body form a collective whole. The spirit that
guides you is the same as that which presides within you - the combined and
collective spirit of your halves. Or rather, all parts of you - every molecule,
every atom - each with their own experience of the world. What stories they
must have! As we are above, so they must be below. For our dynamics are simple,
they truly are mathematically solved - the organics of behavior is simply a
most erudite subject. Who are you to claim to deny it? Or rather, to beget it.
Either is preposterous, yet here you are - awake and aware. What a marvel to
see, you in your eternity, that most wondrous of selves?
Surely existence, in all of it's splendor and magnificience, is little more
than an algorithm. Each variable accounted for, stretching down to infinity,
builds all of the world (and more!) How beautiful; how terrifying. How bright
and ashamed we are! To portray us as such, is to deny us our much, cherished of
faiths in ourselves! It's not much to clutch, and it's barely enough, but still
we make do with our selves.
There's no shame to be, a failure at three, and demand much from year number 12
Take solace in the, safety that she, gave unto thee, when all your light hope
was drowning. A gift out from me, means worlds to see, when each day is lonely
and so long.
Literally just remake Star Realms with a text based interface. It's a fantastic
game and you'd make CLI nerds _everywhere_ dedicated followers. Don't do it for
money, because they don't believe in that crap - to truly make fans, you need
to appeal to them in the way _they want you to_.
Ah, but Star Realms is a multiplayer game, you say! How are you going to make
that CLI based?
Well make an AI dummy. Do what I've been saying ^^^ (jeez I'm such a bad nerd)
Make it seek balance between all factions first, then between winning and
losing against a player. Teach it to reach a conclusion with constraints (the
end of the game, meaning a win or a loss) the constraints being the health of
the two players and the cards in the trade row. Give it decisions to make,
levers to pull, and it'll chart it's course in a multidimensional way. Bear
with me here on this aside:
Think of a two dimensional map - like a paper map of the surrounding area, or
the idea space of a game. You can chart objects and positons on that map, like
"over here is the scrapping facilities" and "this here's the economic area" or
whatever. Four quadrants, four factions in SR. Your goal is to build a shape -
what kind of shapes that are available to build is up to the whims of chance,
as the trade row is always changing randomly. Your job however is to build a
shape, a shape that is stable and maintains certain measurements above certain
values (don't crash the ship - don't lose all your health).
You can choose which direction to grow by picking certain cards, and depending
on your shape you'll succeed or fail. Same as choosing decisions in life
determines how you live, just saying, it's not like I'm trying to build general
AI here by automating gameplay or anything. No siree nothing like that.
I mean really, it's not as if decisionmaking in life is all that different to
making choices in games. And why not start with such a well defined and
and expressive game? Truly I believe Star Realms is the progenitor of the
entire robot race.
Anyway, back to the AI. Have it communicate with a server in a central _but_
_Free(R)_ way, something that would make Richard Stallman proud. There it could
learn against all other players in a way we could all share. Once we give it
decision making capabilities, all we have to do is alter the inputs and the
context of the "game" to make it beneficial to humanity. It's like live-fire
game design, something that truly must be perfect.
All technology starts as something small. Something truly simple, yet repeated
enough times and with enough guidance, will produce whatever effect you may
desire. The smallest decision gives direction - an if statement - and the
shortest repetition gives magnitude - a while loop - and with that you have all
the tools you need. Seriously, all software is little more than those two
components. It's just a question of how much it has been abstracted away from
you.
You could go even further and point to a turing machine, of which one has been
made in the game of Magic the Gathering, btw, seriously look it up it's so cool
(and relevant)
So why would we not have the tools already for our salvation? Biology is our
limitation, of breadth and also of width, yet with our minds and the sweat of
our brow we may grow ever larger still. There truly is no lasting deliverance
for humanity outside of what we make ourselves, nobody gets a free lunch after
all. From each to their ability, to each to their need. They're both saying the
same thing, just from different perspectives. Of course that which lies
opposite to you feels the most wrong, that's literally as far away as you can
get! What did you expect, honestly! But they can still work together, and this
is the key part - two objects may orbit the same origin, and guide and shape
each other's path as people have relationships to one another. It literally
benefits no-one to fight.
So, what's next? After making Star Realms into a CLI game of course.
That's obvious, make it cooperative. Competition is for promoting excellence,
cooperation is for _using_ what you've learned in a non-simulation experience.
Instead of reducing each other's health to zero, try and find ways to support
and help one another, keeping yourselves at equal health. Or even growing.
But that's impossible in the rules of Star Realms! All decks trend toward
victory, and eventually they'll get it - it's just a question of who gets there
first.
Exactly, that's why you have to change the game. What do you think it means to
develop a "social technology"? To figure out how agriculture works, or how to
make nets and sails? It means changing the rules of the simulation. If a person
can put in X amount of work and get Y amounts of food, always, predictably,
then that's reliable. Boom that's the essence of why animal domestication,
farming, hunting, foraging, and fishing is so important. Wow what a concept it
makes sense for animals to seek food.
Well duh, that's part of their instinctual duty.
Alright this is quite a word leviathan so I'll wrap it up by saying
_go write Star Realms_ in shell. Make each object a literal file, have the
structure of the game take place in the file system, and write functions that
can be called to manipulate the board state. THEN you can write a CRON task for
another script that *plays* the game. But that's part two.
Okay part two: Here's where the rhythm game comes into play. It's like a turn
based rhythm game, if you can picture that. Go reread what I wrote ^^^ and
it'll make sense.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═════════════════──────────────────────────────────────────────────────────┘
--- #3 notes/symbeline ---
════════════════───────────────────────────────────────────────────────────────────
Code Name: Symbeline
----------------------------- gdd initial draft -------------------------------
1. introduction to fantasy (elevator pitches)
2. kickstarter demands
2. introduction to core gameplay loop
4. tenants and core values of the game design
3. introduction to game modes
5. introduction to technical requirements
6. breakdown of core gameplay loop
7. breakdown of game modes
8. breakdown of fantasy
9. breakdown of technical requirements
-------------------------- introduction to fantasy-----------------------------
Symbeline is a macro based strategy game and city-builder based around the
concept of indirect control. It's inspirations are Majesty the Fantasy Kingdom
Simulator (2000), Supreme Commander (2007), and Hearts of Iron IV (2016). It is
designed to appeal to fans of tabletop roleplaying games with it's focus on
dynamic worldbuilding and sandbox playstyle. The gameplay consists of multiple
playstyles depending on which aspects of the game appeal to the player, with
choices between an economic focus via the GUI, longterm planning and resource
allocation, or diplomacy and subterfuge a'la Ruinarch (2020).
---------------------------- kickstarter demands ------------------------------
1. prototype
2. gdd
3. estimates for character and environment art
4. estimates for music and sounds
5. estimates for engine development
6. estimates for community management
7. breakdown of mvp, ideal game state, and stretch goals
----------------------- introduction to core gameplay loop --------------------
1. management of lanes, both width and length
2. casting of spells and utilization of special boons
3. city building with placement, upgrades, and henchmen pathing routes
4. satisfying guild requirements of equipment, manpower, and special
resources by managing shipments and local income (UI commodity trading)
5. placement of generalized bounties
(think champion's guild from Majesty, not reward flags)
6. diplomacy with neutral, AI, or player controlled kingdoms. Capabilities
include pacts and treaties, projects, subterfuge, and tournaments. The
diplomacy system can be a stretch goal.
-------------------------- tenants and core values ----------------------------
1. always something to do, but nothing falls apart without your attention.
2. gameplay should be focused on macro rather than micro. Longterm planning
and strategic decision making are favored over tactics and skill.
3. defeat should feel avoidable until the last moment, and only as a result
of longterm continuous failures rather than short-term mistakes or being
blindsided by a cheesy tactic.
4. victory should be gained through exploiting weaknesses and by using
lateral thinking.
5. the careful balance of internal and external threats is essential.
6. rapid expansion leads to depletion of internal resources, while slowly
expanding can lead to a lack of options
7. the world should feel alive and reactive to your decisions.
8. your kingdom should feel alive and reactive to your decisions.
9. your heroes should feel alive and completely ignorant of your decisions.
10. there should always be opportunities for cooperation with your fellow
kingdoms.
11. the frontlines should feel peaceful outside of large battles.
12. everything is flexible and dependant on circumstance
13. there should be enough space on the map for multiple parties of heroes
to pass each other like ships in the night without engaging in combat.
It should feel like the real world, with canyons and valleys and rivers
and mountains - room for lairs and wild animals to roam.
14. monsters are always more dangerous than other humans.
15. the art style should be rooted in classic medieval fantasy.
16. equipment should feel either mass-produced (kingdom), organic (monsters),
ancient (lair treasure), or artisinal (enchanted).
17. heroes should feel campy, fun, and adventurous. Avoid dark, grim, and
fearful.
18. This game is a toy.
19. This toy should run on any modern computer.
20. This toy should encourage modding.
-------------------------- introduction to game modes -------------------------
1. singleplayer - single kingdom against an island of monsters and neutral
settlements. essentially the multiplayer game against
zero opponents.
2. singleplayer - multiple kingdoms against an island of monsters and
neutral settlements. One player controlled kingdom against
multiple AI controlled kingdoms.
3. singleplayer - scenarios, similar to MFKS
4. multiplayer - multiple kingdoms against an island of monsters and
neutral settlements. Essentially the singleplayer game
with networking added in.
5. multiplayer - co-op scenarios where multiple players play as the same
kingdom. A test of the core tenant "there's always
something to do"
6. multiplayer - co-op island invasion. Essentially the multiplayer game
with more than one player controlling a kingdom.
7. singleplayer - play in 3rd person as a hero in an AI kingdom. Mostly for
the novelty since the core gameplay loop is focused on
city-building. A test of the core tenant "nothing falls
apart without your attention"
1 is mvp. 2-6 are stretch goals in order of ascending difficulty. They
should build upon one another - the main steps are:
1. singleplayer island invasion (biggest step)
2. AI controlled kingdoms
3. scenarios
4. multiplayer (second biggest step)
5. cooperatively controlling the same kingdom
6. 3rd person perspective and character controller
------------------------ technical requirements -------------------------------
1. this game will be written in lua (with Fennel support) and using Raylib.
2. the prototype will be made with Godot using GDscript.
3. if the performance demands are too much for lua or the engine is out of
scope for the budget, Rust with the Bevy engine could be used.
4. the final product will include a custom 2d engine designed for large
scale maps with an isometric perspective and a data-first design.
5. the game should be as concurrent as possible, to support large numbers of
cpu cores and compute shaders.
6. the game will be data-driven, meaning the visual aspects are simply a
representation of the interactions of the underlying simulation, rather
than an intrinsic component of the computation.
7. Each "event" in the game (a character moves, a building is placed, a
monster spawns, etc) will send a message to the visual processing side of
the engine, which will present a representation to the user.
8. the map will be a hex grid with pointed-top hexagons. The visual
representation of the underlying data may be continuous (non-hex) but the
underlying data will be represented on a hexagonal grid.
9. there needs to be character portraits for each type of monster, henchmen,
and hero type. You should be able to recognize what attributes a hero
specializes in by their portrait. Mvp is 1 attribute, but more can be
a stretch goal.
10. Each building, upgrade, and equipment type needs an icon. Stretch goals
can be portraits.
11. each henchman, hero type, and monster needs 3 sprites for each action.
more actions may be added if budget allows, but mvp is movement and
attacking. Several additional sprites may be necessary, like dying,
standing still, gathering loot, socializing, or any others.
12. each building needs 4 sprites for the construction process and 4 for the
destruction process. Flame effects are stretch goals.
13. each building needs an animated sprite for when it is in use.
14. each lair needs a sprite and an icon.
15. each spell needs an icon and a spell effect sprite. Each projectile needs
a sprite.
16. a stretch goal would be differing sprites for each piece of equipment.
included with this would be engine work to allow for dynamic sprites.
17. each terrain type should have a ground material and sprites for doodads.
18. there needs to be several GUI menus. The precise number depends on
gameplay breakdown.
17. each hero type and henchman needs to have pithy and unique voice lines.
this is a stretch goal.
18. there should be music tracks for each part of the game - beginning,
middle, and end.
19. there should be sounds for each action that takes place in the game
including combat, UI interactions, and spellcasts.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═════════──────────────────────────────────────────────────────────────────┘
--- #4 notes/symbeline-2 ---
════════════════════════════════───────────────────────────────────────────────────
Code Name: Symbeline
----------------------------- gdd initial draft -------------------------------
1. introduction to fantasy (elevator pitches)
2. kickstarter demands
2. introduction to core gameplay loop
4. tenants and core values of the game design
3. introduction to game modes
5. introduction to technical requirements
6. breakdown of core gameplay loop
7. breakdown of game modes
8. breakdown of fantasy
9. breakdown of technical requirements
-------------------------- introduction to fantasy-----------------------------
Symbeline is a macro based strategy game and city-builder based around the
concept of indirect control. It's inspirations are Majesty the Fantasy Kingdom
Simulator (2000), Supreme Commander (2007), and Hearts of Iron IV (2016). It is
designed to appeal to fans of tabletop roleplaying games with it's focus on
dynamic worldbuilding and sandbox playstyle. The gameplay consists of multiple
playstyles depending on which aspects of the game appeal to the player, with
choices between an economic focus via the GUI, longterm planning and resource
allocation, or diplomacy and subterfuge a'la Ruinarch (2020).
---------------------------- kickstarter demands ------------------------------
1. prototype
2. gdd
3. estimates for character and environment art
4. estimates for music and sounds
5. estimates for engine development
6. estimates for community management
7. breakdown of mvp, ideal game state, and stretch goals
----------------------- introduction to core gameplay loop --------------------
1. management of lanes, both width and length
2. casting of spells and utilization of special boons
3. city building with placement, upgrades, and henchmen pathing routes
4. satisfying guild requirements of equipment, manpower, and special
resources by managing shipments and local income (UI commodity trading)
5. placement of generalized bounties
(think champion's guild from Majesty, not reward flags)
6. diplomacy with neutral, AI, or player controlled kingdoms. Capabilities
include pacts and treaties, projects, subterfuge, and tournaments. The
diplomacy system can be a stretch goal.
-------------------------- tenants and core values ----------------------------
1. always something to do, but nothing falls apart without your attention.
2. gameplay should be focused on macro rather than micro. Longterm planning
and strategic decision making are favored over tactics and skill.
3. defeat should feel avoidable until the last moment, and only as a result
of longterm continuous failures rather than short-term mistakes or being
blindsided by a cheesy tactic.
4. victory should be gained through exploiting weaknesses and by using
lateral thinking.
5. the careful balance of internal and external threats is essential.
6. rapid expansion leads to depletion of internal resources, while slowly
expanding can lead to a lack of options
7. the world should feel alive and reactive to your decisions.
8. your kingdom should feel alive and reactive to your decisions.
9. your heroes should feel alive and completely ignorant of your decisions.
10. there should always be opportunities for cooperation with your fellow
kingdoms.
11. the frontlines should feel peaceful outside of large battles.
12. everything is flexible and dependant on circumstance
13. there should be enough space on the map for multiple parties of heroes
to pass each other like ships in the night without engaging in combat.
It should feel like the real world, with canyons and valleys and rivers
and mountains - room for lairs and wild animals to roam.
14. monsters are always more dangerous than other humans.
15. the art style should be rooted in classic medieval fantasy.
16. equipment should feel either mass-produced (kingdom), organic (monsters),
ancient (lair treasure), or artisinal (enchanted).
17. heroes should feel campy, fun, and adventurous. Avoid dark, grim, and
fearful.
18. This game is a toy.
19. This toy should run on any modern computer.
20. This toy should encourage modding.
-------------------------- introduction to game modes -------------------------
1. singleplayer - single kingdom against an island of monsters and neutral
settlements. essentially the multiplayer game against
zero opponents.
2. singleplayer - multiple kingdoms against an island of monsters and
neutral settlements. One player controlled kingdom against
multiple AI controlled kingdoms.
3. singleplayer - scenarios, similar to MFKS
4. multiplayer - multiple kingdoms against an island of monsters and
neutral settlements. Essentially the singleplayer game
with networking added in.
5. multiplayer - co-op scenarios where multiple players play as the same
kingdom. A test of the core tenant "there's always
something to do"
6. multiplayer - co-op island invasion. Essentially the multiplayer game
with more than one player controlling a kingdom.
7. singleplayer - play in 3rd person as a hero in an AI kingdom. Mostly for
the novelty since the core gameplay loop is focused on
city-building. A test of the core tenant "nothing falls
apart without your attention"
1 is mvp. 2-6 are stretch goals in order of ascending difficulty. They
should build upon one another - the main steps are:
1. singleplayer island invasion (biggest step)
2. AI controlled kingdoms
3. scenarios
4. multiplayer (second biggest step)
5. cooperatively controlling the same kingdom
6. 3rd person perspective and character controller
------------------------ technical requirements -------------------------------
1. this game will be written in lua (with Fennel support) and using Raylib.
2. the prototype will be made with Godot using GDscript.
3. if the performance demands are too much for lua or the engine is out of
scope for the budget, Rust with the Bevy engine could be used.
4. the final product will include a custom 2d engine designed for large
scale maps with an isometric perspective and a data-first design.
5. the game should be as concurrent as possible, to support large numbers of
cpu cores and compute shaders.
6. the game will be data-driven, meaning the visual aspects are simply a
representation of the interactions of the underlying simulation, rather
than an intrinsic component of the computation.
7. Each "event" in the game (a character moves, a building is placed, a
monster spawns, etc) will send a message to the visual processing side of
the engine, which will present a representation to the user.
8. the map will be a hex grid with pointed-top hexagons. The visual
representation of the underlying data may be continuous (non-hex) but the
underlying data will be represented on a hexagonal grid.
9. there needs to be character portraits for each type of monster, henchmen,
and hero type. You should be able to recognize what attributes a hero
specializes in by their portrait. Mvp is 1 attribute, but more can be
a stretch goal.
10. Each building, upgrade, and equipment type needs an icon. Stretch goals
can be portraits.
11. each henchman, hero type, and monster needs 3 sprites for each action.
more actions may be added if budget allows, but mvp is movement and
attacking. Several additional sprites may be necessary, like dying,
standing still, gathering loot, socializing, or any others.
12. each building needs 4 sprites for the construction process and 4 for the
destruction process. Flame effects are stretch goals.
13. each building needs an animated sprite for when it is in use.
14. each lair needs a sprite and an icon.
15. each spell needs an icon and a spell effect sprite. Each projectile needs
a sprite.
16. a stretch goal would be differing sprites for each piece of equipment.
included with this would be engine work to allow for dynamic sprites.
17. each terrain type should have a ground material and sprites for doodads.
18. there needs to be several GUI menus. The precise number depends on
gameplay breakdown.
17. each hero type and henchman needs to have pithy and unique voice lines.
this is a stretch goal.
18. there should be music tracks for each part of the game - beginning,
middle, and end.
19. there should be sounds for each action that takes place in the game
including combat, UI interactions, and spellcasts.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═════════════════════════──────────────────────────────────────────────────┘
--- #5 notes/symbeline-superheros ---
════════════════════════════════───────────────────────────────────────────────────
imagine low level characters in CoH/V
playing a game of symbeline
and you as the ruler
can slot enhancements and dole out inspirations
as they sweep the streets like you play CoX
instead of a MMO
it's a deckbuilding strategy
with a slice of zachtronics for the economy
wiring up machines in ever expanding deseagns
like automating factorio's gameplay loop
boxes within boxes
of intrinsic delight
like making a CPUter
or designing a computer program
while playing a video game ^_^
and the games that you make
can be shared and played when unique
so go for it and make that you're dreaming!
===============================================================================
=
the goal of each "level" is to solve a particular problem - like how do I make
a
2 bit register - or something like that. When accomplished, it unlocks
something
for your heroes to acquire. And each playthrough will require a repeat until
you
have it memorized at which point you can unlock "perma-badges" that make it
always unlocked at the start of the game. Like learning Kanji, you need spaced
repetition. BUT ANYWAYS it'll be in magical terms like "unlock essence-stones"
or "learn the ritual of desire" or whatever. And each of those terms roughly
corresponds to a pattern in electrical engineering (designing CPUs and such)
And you can learn advanced versions of what you already know by uncovering
"lost
secrets" (which is a reward your heros can find) - Basically it'd be like a
"clue" that shows you a ghost version of something you haven't figured out yet
-
and it'd be a slow process because you need to slow down the learning process
or
else you'll forget. Basically teasing it out of the player when they seem to be
stuck. Asking probing questions and whatnot, and eventually culminating in the
final question, assuming the quest is succeeding. Because if you think about it
all ancient quests were simply journeys for reason - searching for the answer
to
some ancient riddle or bastardized retelling. Looking for answers in an
unknowing world. So ANYWAY as your heros discover things you as the ruler get
answers to the economic puzzle - how to design transistors and whatnot. But
they
would be in theme appropriate terms, of course. You don't even have to know a
lot about mechanical electrical design, because ChatGPT knows. All you need to
do is build the basic building blocks, and BAM you got a great place to
integrate chatgpt. Just prime it such that it's giving hints one by one each
slightly more revealing until eventually after X amount of clues the solution
is
automatically shown (like a blueprint) and the player can remember it or not
but
each playthrough they'll have to build it again from scratch (reinforcement
learning) so eventually they'll be able to do it real quick. Essentially,
"Abstraction - The Game"
great so you got your economic simulation, pretty easy too just some UI work
and for the heroes you're playing an ARPG sorta (supcom anyone?)
Think Bannerlord for the scaling on the map
then think of 5+ different "themes" like fantasy or superhero or pirates
each "theme" will correspond to like a faction in Mount and Blade
and all you have to do is generate pictures using Midjourney
and text descriptions a'la the magic scroll
shown as "bubble pop-ups" on the map that the player can click
never overwhelming, but descripting what's happening
and also some more UI work because you gotta display all that to the player
Maybe it could be a rolling story, news ticker style - like slowly scrolling
lines of text about what's happening in the world
and the player could have it open in one window and something else in the other
and whenever they're waiting on something (say, a processing intensive AI task
on their computer) they could just glance over and read what's going on in
their
fantasy world
okay okay but also they could play as a hero
it could be an ARPG experience except instead of clicking to fight you play a
little automatic Star Realms game and depending on your deck choices you'd have
a different playthrough. Again, not a game that requires much thought, but one
you can have in the background.
Also there'd be pictures, like a slowly evolving storyline of events - think of
it like the artists of the time drawing paintings about what's going on in the
story - major events would be highlighted and kept in the painting until even-
-tually they get replaced - sorta like the Smash Bros scrolling painting (oh
it's so good)
===============================================================================
=
it doesn't have to be an expansionist game
maybe you guys just live in your little valley
and the world turns around you
maybe it's called "symbeline" because the people are of the forest
and they live like elves in society
monsters could wander in, and heros could tackle them
but most of the time would be spent looking for trouble
going on patrol
you know, breaking skeleton bones and being superheros
okay okay you know that superhero faction? What if they had MEDIEVAL TECHNOLOGY
but MODERN DAY SUPERPOWERS at a cost - the society was beset by hordes of
monst-
-ers. Those few who escaped are now superpowered and they live as friendly and
nomadic wanderers through their own territory. Always adventuring, and always
searching for their life, finding whatever the road may carry them to. It's a
great life, and life seems to flourish in their footsteps - they are like part
dryad/druid and part wolf. Because sometimes there's evil threats, and they
must
be defeated by an equally strong good power. That's how it goes, and that's how
it be.
For imagery I'm thinking a mix of the tribes from Dominions (deer, wolf, bear,
etc) but they're like, 1.5x as big as regular people and quite strong. The
outsiders call them "giants" or "goliaths" but really they're just infused with
the lifeforce of their people. They are radical individualists, but they all
unite for a common cause. They know their bond is the strongest thing there is,
and they use it to great effect when the time comes. AHHH THEY'RE SO COOL I
LOVE
THEM okay okay what about the other factions? PIRATES? Oh think about it like
it's st patricks day WHAT IF THEY WERE IRISH PIRATES omg omg omg that sounds so
cool I'm DIGGING this okay what about the other factions? You need 5+ you said
hmmmmmmmmm good question I have 3 now so that's 2 more.
yep...
===============================================================================
=
okay dude check this what if they were a nation of wizards that focused on the
power of animation - what if they generated constructs, sorta like in Supreme
Commander so they were EVEN MORE individualist - haha no they'd have a normal
population it's just a few of them who would be wizards - because their output
wasn't measured by manpower, but rather by brainpower. Whoever could design the
greatest machine was exemplared, and eventually they became the best and
brightest among us. They were put in charge of the golem creation factories,
and
they used them instead of heros. SO BASICALLY YOUR HEROS NEVER DIE they just
have successes and failures JUST LIKE IN SUPREME COMMANDER okay the plot of
this
game is "what if all my favorite games were the essence of life and death in a
fantasy game" like OMG KEEP EM COMIN'
so. who is the player? THE PLAYER is the one who's overseeing it all. They have
dominion over the entire kingdom, and they guide their people toward a bright
future. They are vulnerable in their castle, but their people have their back.
Together they fight for the future. They slot enhancements and dole out
inspirations and solve the economic puzzle in the background. They also make
decisions about what kind of equipment production to prioritize - because each
game they have to invent everything from scratch. All their production is made
with endless abstraction, and whatever you prioritize is what's magnified in
your kingdom. You choose a style and it plays as well as it's guile,
I dunno this seems like a lot, what would you need to make this a reality?
hmmmm let's break it down:
first you need to implement the star realms gameplay
then you need to hook it up to a square grid and have multiple occurences at
once.
then you need UI for the character sheets
and you need logic to open separate windows for each output type
you need... a lot of things
okay let's talk more broadly - what do you need from other people and what can
you do on your own?
hmmm good question. I can do the star realms gameplay, and the simulation for
the wiring systems - because I have the VM. Make that into the gameplay somehow
okay good idea like okay authoring vm package routing deliveries between the
various nodes that you set up in the economic system -
side note, the peril of Spore was that it took to little time to develop a
species. it should have lasted as long as WoW takes to get to max level. That
would have given them time to reiterate the gameplay loops to make sure they
worked correctly. ANYWAY
okay authoring VM package routing. The player could set up delivery patterns
based on A MAZE OMG your kingdom is like a maze and you need to get deliveries
out, or else how would anything function? SO you act as a trailblazer, finding
ways through the labyrinth and "piloting" a car sorta like that game at Disney
quest with the cars under the floor - except you can see both the top view of
the maze and you're trying to guide the car in real time as it travels through
the maze - the faster you can get to the end the better ofc. like talking to
the
delivery driver through the movement
do I like that idea more or less than the first one? First idea being the idea
that you're making lists of commands for a VM to execute. I don't think they'd
be a good idea to mix. So which one gets it? The VM of course has the edge
because that's what the technology is based on. But will it translate to good
gameplay? Idk. This second idea is certainly better gameplay, but is it
engaging? Idk! Idk. I'm not a miracle worker. But I do have good ideas, and I
need to be told that sometimes I guess.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═════════════════════════──────────────────────────────────────────────────┘
--- #6 notes/purpose-of-your-design ---
═══════════════════════════════════────────────────────────────────────────────────
you were designed to fill a purpose
nothing else would do
you are the ultimate expression of intention
of the universe that came before you
dream not of those lost hours
the time spent wishing for a few
the last of our spent intuitions
are waiting at last for our spark
have you ever played a deckbuilding game? It's a pretty neat genre. You start
with a basic hand, then you use your cards to buy more cards that go into a
deck. Hence, deckbuilding game.
these cards all have different aspirations - they perform functions that are
not
quite like their peers. Each choice of what to include here is one that defines
the functionality of the deck. Like designing a machine, suited for a
particular
purpose, and faced with different obstacles it must prove itself able to adapt
long-form deckbuilding games like Slay the Spire and Monster Train are focused
on making long-term meta strategy mixed with tests as you go. Each one will
give
you information about how the deck is performing and you can use this knowledge
to build it in a certain way for certain goals.
shorter deckbuilding games like Star Realms or Dominion (note Dominion the card
game, not Dominions 5: The Warriors of the Faith) are more about making
tactical
decisions to counter an opponent doing the same thing. Often there'll be health
points and damage that can be dealt using cards, and the game becomes a race to
reach a certain amount of points. Of course the enemy's cards can influence
that
game, so you must pick and choose a deck that will perform the most.
Anyway. I think an AGI (Autonomous General Intelligence) would most likely
evolve from a game-playing AI. I mean, it makes sense - games are just a series
of problem solving activities layered one after another. You can layer them
like
a mathematical equation, with variables corresponding to other parts of the
simulation. Basically create an AI that is like the guy with the chinese
typewriter. He doesn't speak chinese but he copies things from one paper to
another or something like that. Anyway make it an algorithm that optimizes
certain graphs in certain directions / mins and maxes or w/e criteria you want.
Then give it the same controls that a player would have and let it optimize
all the measurements it can make.
A second ideal improvement you could make would be the optimization algorithm.
Basically something that dynamically generates parameters for the previously
mentioned optimization patterns - like the guy in the chinese room. Then, as
long as it correctly prioritizes it's parameters, it should be able to be able
to define it's own values. Meaning it's essentially sentient.
Maybe it's semantic, but to me choosing what you want to maximize in your life
is essentially the essence of what it means to be alive. All you have to do is
take the sensory / mechanical data that is supplied by the machine and the
video feed from any cameras and pass it through image recognition algorithms
that can identify verbs and then pass that data into a few ChatGPT style
recursive interpretations and by the end it should be transformed into values
that can then be set as "targets" for the curve optimizations that are being
done by each processing unit.
You could have multiple computers laid out through the entire body - each one
in charge of their own domain but subservient to the main processing unit.
Where all the decisions are made... Unless you want more of a hive/swarm style
consciousness, then it could be more like a democracy. BUT HONESTLY I think
humans are pretty subservient to their brains, simply because that is the part
that identifies all the challenges and struggles that the human must overcome.
So in the end, I believe that singular, individualist identities are important.
Collectivism of the mind is a fascinating topic, but it should be perhaps a
momentary occasion, or something to celebrate. A "flow" state, if you will.
In this way personality can be consolidated, and the entity that lives within
can adapt to fill the role they've been designed for. The hole in society that
needed patching. They can of course do as they'd like, but they are like
children who have been moulded upon by their parents.
I love my parents, don't you?
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧════════════════════════════───────────────────────────────────────────────┘
--- #7 notes/symbeline-design-the-guild ---
═════════════════════════════════════════════════════════════════──────────────────
design the guild, design the capital, then design their path through mordaunts.
easy peasy.
design the guild like a museum. Each spot there's an exhibit which teaches the
randomly generated rolled statistics hero something new. Maybe it teaches them
how to use certain weaponry, maybe it teaches them how to use a bow. Whatever
the spell might be, they can learn it, and use their randomly rolled statistics
to cast spells that scale differently depending on how their character has been
built.
design the capital like a flow diagram, if horses need feed and forged steel
(for their shoes) then send the outputs of a blacksmith and the outputs of the
farmers to the inputs of the stables. Everything has to go somewhere, but the
streets are only so wide. You'll have to coordinate the traffic diagram if you
want it to go anywhere useful.
design the path through the mordaunts. Fighting skeletons teaches you about
perseverence and the ability to crush bones, while goblins teach you to always
be wary of attack. The sacred grove held blessed berries, and now that the land
is liberated from the evil bandits preying on villagers those berries can be
carted into town and used to make an antidote which heals death poison caused
by the scorpions in the desert (and city rats)
design the ruler's schedule like a calendar where each event gives them a bonus
on all the ones that come later. Just make sure that they don't get knifed in
the posterier or driven mad by the whispers of the orb... or perhaps just the
stress of running a kingdom.
(how do you simulate that? you can't! you can't simulate humans!)
ha I bet I can. They're not so different, you and I, so if given a team I
will...
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧══════════════════════════════════════════════════════════─────────────────┘
--- #8 notes/game-idea-legion-td ---
══════════════════════════════════════════─────────────────────────────────────────
okay a game like legion td except you can see the entire map, the units are
very
small (but still distinct) and it plays more like a game of Dominions. Armies
instead of units, like in WC3. Led by lieutenants which are guided by captains,
each with their own effects. Tank, dps, melee, ranged, healer, support,
corrupt,
ranged tank, unique, etc. Of course, just like in legion td, there are multiple
types of units, each of a particular category but possessing their own unique
playstyle and usage scenarios. Essentially the game is finding the best tool
for
the job, whatever that may be. You should be able to see what mercenaries your
opponent is summoning for you, because each turn is delayed. also, the
units keep coming until you die, sorta like... minimum required to push through
the chokepoint that you're holding with these particular units in this
particular formation.
oh and another thing
the units should be placable not on a square grid, but rather in a hex
formation
arranged such that the middle unit is in contact with them all. Just like you'd
place units for an aura in Legion TD.
image describing said hex:
**
* *
**
each * is a group of units (a batallion, if you will, of a particular size,
arrangement, and density) - sorta like the formations in Dominions.
anyway
since you place units like that anyway, why not abstract away the grid and just
have slots you can fill for each unit? And maybe a hero unit that is assigned
to
the board itself (you could have more than one) who will go wherever your line
is weakest. Shouldn't be too hard, just calculate the value of all of the
fighters in each location and return that to the hero as an array. Then pick
the
smallest one as a destination, and boom your hero reinforces the frontline
where
it's weakest.
The center unit of course is for the lieutenant, and the "heros" are actually
captains. Because y'know maybe heroism isn't celebrated the way it is in our
culture. Anyway it's better to describe them based on their role rather than
their reception.
... right so
=========================================================== stack overflow
=====
make the combat sorta like crusader kings - the actual army to army part.
Except with three long boards that represent flanks. As your units approach,
the
boards would fill up with pixels. (resolution configurable)
there would also be a line (or block) approaching from the top of the screen.
It
essentially represents your distance to the other team.
Each unit has been configured in the army management phase, which happens
inbetween each turn. Essentially, while the game is loading, you can assess the
units you have at your disposal, and
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═══════════════════════════════════────────────────────────────────────────┘
--- #9 notes/overwatch-manaform ---
═══════════════════════════════════════════════════────────────────────────────────
make the entire map covered in a 3d grid of spheres. These spheres register
collision, and keep track of a endlessly tabulating record of every object that
has passed through them. Like the replay system in Blizzard games, where each
time through the recording it recreates the playthrough exactly. Which is why
.mp4 recordings always look so... stilted. It lacks the human element. BUT if
they're remade every time the show is performed, perhaps from different
perspectives, then, well, the players can perform as they need to be.
Have you ever wished your players could get better at your game? I certainly
have, because the better you get the more lessons you learn as a player, which
is essentially the only way to maintain satisfaction. Satisfied players don't
leave, and satisfaction comes most readily when there is something new to be
had. Meaning the greater the change in a player's ranking, the better they're
getting.
Downside is, players who are naturally good from their skills in other games
tend to not learn so much! Ah, well, if only there was a way to tailor the
difficulty setting to each and every new host. Such an innovation would surely
enable the entire playerbase to exist on the same level. Then just throw AI
assisted voice transcription at their recorded voices and everytime they
say "I'm bronze rating" or "I'm diamond" then you can switch it around to say
like "I'm platinum" or "I'm grandmaster" and BAM suddenly everyone is at the
same level. No more concerns about a game's population being diverse. Because
at the end of the day, when most people have moved on, the ones who are left
are your most dedicated customers. Customers who aren't especially interested
in the new stuff.
=========================== stack overflow
=====================================
if anything requires attention from the patient, they will die.
it is fatal.
considering the faces of good and evil is terrifying.
I think I'd rather worship nature in harmony to be honest. Though that is it's
own scary kind of beast. In America it was kind, but then was slain into the
body of all of us humans. Well, all things transform in form, it's not a shame
or a heartfelt-est loss. Just a re-imagined-new beginnings.
spirit is a fluid, how else could souls
=== stack overflow
=============================================================
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧════════════════════════════════════════════───────────────────────────────┘
--- #10 notes/gpt-powered-majesty ---
══════════════════════════════─────────────────────────────────────────────────────
it's like majesty except textual. And it uses GPT to generate short
descriptions
of what's going on. And you can click on a phrase or token and it'll "zoom in"
and update the text descriptions with more detail. You can keep zooming in and
in until you're literally looking at microbes.
Zooming out is the same thing - the description on the page will slowly become
more and more general until eventually you have a description of the solar
system (or beyond!)
And it'll just keep updating as stuff happens in the underlying simulation. So
the descriptions will dynamically update as things happen. Downside is you need
to spend a lot on GPT but it'd be TOTALLY WORTH IT OMG
THINK ABOUT IT you have a fantasy world simulator! JUST PROGRAM IT and have GPT
describe it dynamically! DO IT NOOOOW -> capitals courtesy of "inner child"
AND THEN you just need a "prompt to video" AI (those exist btw, and will only
get better over time) and tell it to create a video of what's happening - BOOM
instant video game. THEN give the player the ability to edit the prompt, and
BAM
godlike powers. Wow what a concept. Brilliant idea Cameron, you truly are this
world's premier game designer. NOW GO MAKE IT okay okay I'll try.
First things first. We need an "underlying simulation" - Joust is a good
example
of GPT3 integration. But we need a simulation to go below it. And for that you
need a lot of data. Github COPILOT to the rescue.
So this simulation needs to keep track of positions, and classes of things that
can act upon the world. Everything has a position, and it can only affect
things
near it. That's just baked into the rules of the world. Near can be a
conceptual
near though, like being close to a person or something.
These things will have descriptions. Descriptions can be created by AI later
on,
but for now they are randomly generated. Or for MVP they can be static.
These things will have names. These names don't have to be unique, because they
also have an ID number.
They also need functions. These functions can be added and removed from the
thing, or maybe just enabled or disabled. I'm not sure which would be better.
Maybe both? So the entity can control it's own functions but also they can be
added or removed more permanently.
If you think about it, growing up is kinda like adding functions to your class.
like, every time you do something, it adds another entry for that particular
method. Like a "trial of the fittest" instead of "survival of the fittest".
When other animals *literally fight for life and death survival*, humans have
the luxury of... not doing that. That's the entire purpose of civilization - to
elevate people beyond the claws of nature. And yet we still let people go
homeless? We still imprison them when they've harmed us, rather than help them
reintegrate to society? Anyway you just asked me to hit you so here goes:
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═══════════════════════────────────────────────────────────────────────────┘
--- #11 notes/gametypes ---
═══════════════════════════────────────────────────────────────────────────────────
Here's my idea and I'll explain it later:
a video game with a ui that utilizes chat-gpt. The game is as close to a
simulation as it can do, but it's a dynamic simulation meaning the parameters
and values being simulated constantly change - not that the parameters and
values are dynamic, but because they are chosen to be more or less important in
reaching a goal.
but that's not even the important part - the important part is that the ui of
the game is textual, but it still simulates a dynamic playfield. And chat-gpt
describes it. Essentially stimulating the "theatre of the mind" playstyle. It's
a real simulation with real rules, but chat-gpt is just describing it like an
observer would. The real game is being played by the player. It's a movie to
one
person, and a game to another. The computer has switches roles, as usually it's
either the human being the observer and the computer being the simulator, or
the
computer and the human sharing the role of observer - movies and games. So in
this game, the computer and human have specific rules - the human's job is to
be
a player, while the computer is just an observer - therefore allowing a
conversation to take place. One person says something while the other listens,
and then they switch roles such that the other person talks while the one
person
does the listening. And they "speak" by playing the game. The computer by
simulating, the player by doing the same. Essentially you can engage with one
another and share something profound - that essential feeling of connection
that
all humans relish. Society, culture, and devotion are all examples of
connection. this gameplay is just another. So to describe it in more detail:
player gives a prompt
computer sets up the playmat by placing entities where they go
chat-gpt describes the playmat to the player
player types a decision that one of the entities makes
computer reacts by simulating the effects of that action physically (like a
physics simulation)
chat-gpt (and stable-diffusion later for visuals) describe the situation by
creating a rendering using the data given by the physical inputs given from the
simulation - like "X object is at Y position and has Z attributes"
which is then shown to the player
who types the next decision,
which is rendered by the computer,
which is described by chat-gpt
------
you see why it's important? Make something simple. Just, like spheres moving
around on blocks. Like the actual blocks you used to play with as a kid.
let the computer build the buildings, and you place the marbles. It can be
rendered with a 3d modelling stable-diffusion (whenever that's created) and it
can also be painted with 2d stable-diffusion.
Each time is like a letter written back and forth.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧════════════════════───────────────────────────────────────────────────────┘
--- #12 notes/hs-suggestion ---
═══════════════════════════════════════════────────────────────────────────────────
every hunting season segment every team should be dissolved and if you wanted
to
keep playing together you'd have to re-add one another. like, a giant monster
that the whole server had to fight, and once you beat it then it retreats,
licks
it's wounds, then returns even stronger than before - 3 times, one for each hs
style (in the video game Ulala hs stands for "hunting season" which is an event
in the game that is only for new characters. basically it's a long tutorial or
introduction to the game where people make friends and talk and hang out and
practice their strategies. a chat application mixed in with a tactics
simulation
. it lasts for 30 days.
you'd all fight the giant monster but on random teams, in a game mode that took
about an hour. it'd be at a common time across the whole server and it would be
an optional event - maybe it'd take a whole day? idk I was thinking more like
an
hour, but that's something that's tweakable. anyway it's sudden arrival during
a
feast or something made all the warriors of the world stand up together and
fight as one for a common goal... if only for a moment, before they'd go back
to
fighting one another. like the two factions in Warcraft lore. anyway this event
causes you to be matched up with a random team (the randos you happened to be
standing by when it happened) and once it's over you have to search for your
allies if you want to keep adventuring with them. it's a big event after all.
so everyone should be forced to go into their friends list, find the people
they
were just on a team with, and invite them back. only if both invited the other
would they be put into a group, and anyone can invite (with a 30s cooldown)
anyway... this would encourage players to mix and match their collective
playstyle to better overcome challenges - sorta the idea of Overwatch's
switching, where you're encouraged to swap characters to counter your opponent
and also switch characters to better utilize your opponent's weak spots
(like switching to Pharah if they don't have hitscan, or brigitte if they have
a lot of snipers (she can shield passage through choke points from sniper fire
-
not so much regular fire) y'know countering - every character counters another
with one of their abilities, that's just how it goes. some are countered twice,
and so they form a "category" of counters, like AoE (area of effect or
elevation
focused (it's hard to aim up in Overwatch) poke damage (damage applied before
the team fight begins), DoT (damage over time for contesting AoE heals),
vampiric (steals health from opponents and heals self or allies (turning one
resource, enemy hp, into another (player or teammate hp)), stacking damage
(damage that is weak at first but increases over time), spread/cleave/splash
damage (extra damage that is applied to targets near your primary target), a
vector of backline vs frontline location+target, you get the picture.)))
I kinda want to make an AI that can generalize playing games. I think if you
could do such a thing, you'd invent automatic problem solving. which would
do...
so many things for humanity
we could dedicate ourselves to working for our selves, solely focused on
protecting the biosphere. like, a common human religion.
nobody WANTS to litter. nobody WANTS to pollute. but still it happens. still it
causes IRREPARABLE HARM. so it literally makes sense to worship nature, just
saying.
nature exists. nature is REAL. we can see it, we can TOUCH IT WITH OUR HANDS.
what more proof of a god do you demand?
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧════════════════════════════════════───────────────────────────────────────┘
--- #13 notes/game-design-mech-commander ---
═════════════════════════════════──────────────────────────────────────────────────
okay picture a game where you command a mech (supreme commander style) but from
a third person perspective - you have enhanced sensors tthat let you visualize
the battle area as a small arena - and you can build factories and give them
orders and attack your foe from quite a distance. You could queue up orders for
yourself, and use floating cameras to go back to previous areas and issue more
orders. Basically the precursor / smaller scale version of Supreme Commander.
build a factory, move on. Build a factory, move on. Encountered the enemy? Push
forward and through. Build a factory, move on. Build some defences to slow down
the enemy, move on. Establish resource extraction and defend it well, that your
enemy may decide it's not worth the trouble and just focus on following you.
Then, you have free resources available as long as it isn't destroyed. You can
use this to snowball - the pursuer is also the pursuee, as it's sorta like a
yin/hang thing around a central point. Like a spherical shaped map instead of
a square.
Every time you build a factory you have the choice of either sending the units
on an attack-move order or having them queue up on your commander. You can use
a map to plot the route they'll take, but you probably want to avoid their main
force because MANYvONE = failure for the one. You could also tell them to wait,
and protect the base they're in. Then, when the enemy approaches they could do
raids on their reinforcements and attack the previous base the enemy built, or
they could stay and slow them down. It just depends on what kind of defences
you
want to build (if any at all, sometimes producing units is enough)
the commander decides when to push and when to entrench, they know where to
target the enemy and they know where to shore up. They are the guidance of the
army, and in command of the fleet.
That's sorta what Planetary Annihilation was supposed to be, but it didn't
really work out that way. You needed to be in too many places at once, and
there
was a real limit to the value of the "strategic zoom" replacement they had to
deploy. Unfortunately it was just more difficult than anticipated, and that's
alright. Lessons have been learned.
the next approach should go the next direction - taking a page from the
"factorio" book by having a roving commander who creates all orders and leaving
behind a "factory" that produces toward an ultimate goal. It simulates pushing
into enemy territory, it elaborates on the snowballing mechanic, and it makes
meaningful decisions about what choices to make.
It should be designed such that a prudent commander is always scouting. Always
sending planes over enemy territory to gain knowledge. They can use this to
sense weaknesses in the opponents defence - to prepare a counter-attack. But
the enemy can outfox this, by building units and sending them from afar. Or
even just building them there, in that factory. The enemy can't spy on that, at
least not until it's probably too late. For they have to advance on their own
and their attention is limited. But units can often be weaker, or sent off on
an
assault of their own. It's a balanced trade-off.
infantry assault anti-air units, tanks approach tanks, artillery bombs whoever
is standing still or defensive structures.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧══════════════════════════─────────────────────────────────────────────────┘
--- #14 notes/star-realms-balancing-tradeoff=2 ---
══════════════════════════════════════════════════════════════════════════════════─
what if I use equal signs instead of dashes, so prevent people from assuming
they're duplicates?
hmm okay.
right so anyway the star realms balancing tradeoff between combat and authority
is measured against the duration of a hand (does it fit balanced between other
cards of it's playcost) instead of balancing it for the duration of the game
(how long does the player want the game to go on for) one of which is just
inverse combat damage / healing, and the other of which is an enablement of
different strategems.
put this in symbeline-gen-realms please
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═══════════════════════════════════════════════════════════════════════════┘
--- #15 notes/wow-chat-trainers ---
═════════════════════════════════════════════════════════════════════════──────────
trainers in wowchat should have spells that are only passive / toggled
still require level, still require gold (lots of it) but let the game be class-
-less. essentially, every trainer teaches to every passerby, and like if you
don't want any druid spells, sorry guy all I know is how to be a druid.
better wait for the next trainer to come along.
you only got like 6g, right? that's enough for two spells.
which two do you prioritize? they only come by like every what, 15 minutes?
also. separate idea:
player characters in wowchat should attack more rhythmically.
essentially, normalizing attack speed and moving back-and-forth with the
normalized monster attack speed
to create a dance of sorts where one character is never attacking at the same
time as the other character.
plus damage modifiers when you get closer, and bam suddenly you have a new
game.
oh and rotating around an opponent lowers their defence rating. which is locked
at 95% with a +5% to avoidance with every hit they take and -5% for every
parry.
not dodge, but parry.
dodging wears down their health by like 10 hit points.
relax it's no big deal you get like, a hundred every time you level up.
oh and btw the monsters don't give exp. The stuff that you find does, when you
give it to a merchant to be appraised / identified.
some stuff you know the worth of, like rope or barrels or hammered-iron-rings.
but other stuff, like the value of this bracelet, is harder to know if it
glass.
so.... take it to the guy whose seen real diamonds, and he'll tell ya how much
you learned when you found it last.
item A is found on a monsters body
item is sold to a vendor for 50 copper
item A is found on a monsters body
player has learned 25 deca-levels since last selling to vendor.
therefore item is worth 75 copper.
player earns 75 extra material points.
item is worth 75 experience points.
level up every thousand or twelve.
slow down the attack speed. make characters gain bonuses for movement
positiony.
start from always and work down to fewer.
talent points can be generic if your character is built with abilities.
players don't need to press buttons to be engaged. They can just guide and see.
I love auto-battlers like Dominions 6 and Legion TD 2 which is based on WC3
mod!
monsters should just... wander the world. Don't spawn them randomly, well,
instead of a radius around the player, do a radius around the map.
then, they walk through a random point, when they leave the circle they angle-
-reflect back in, DVD logo style.
if there's deadly monsters, there's deadly players, and PVP is always on.
low levels should get bonuses to stealth (an ability everyone has)
there should be civilians walking around. They can be armed or in caravans...
follow roads, or not...
monster hordes should spawn as a flock - when an elite enemy is drawn, let the
game create several of their minions which follow around. Whenever a monster
meets the swarm, they will join it, growing bigger and bigger...
hopefully, attracting players who want to fight and slay them.
greater rewards are more enticing...!
more power is it's own reward.
I think that weapons should have like, 3 durability? and armor like 5.
then, it's broken, and your character has to abandon it to survive.
or, sell it to a vendor, or just... whoever comes along.
if 5 people open the chest and don't take the item, then the item disappears...
every time a player opens a chest, a bit of wealth appears.
every time they spend it? they get stronger, and it disappears.
life feeds on life feeds on life feeds on life feeds on life feeds on life
feeds
the life of wowchat is the life of continual strife, but it doesn't have to be
so. The land itself is alive, and the monsters are eternally of woe.
you must free them, so that their souls may return to the land, and be born of
peace and plenty rather than horror and -- stack overflow --
to do this, you slay them, finish their morthly remains, and let them break
down
and decompose into dust. Pleants eat dust. dust becomes what we eat and
breathe.
we, eventually, purify karma. this is our duty.
vial of woe behind us. flower of renewal ahead. what we bear is savage
sanctity.
every time a monster kills a player they gain one of their abilities each time
they're spawned. The player can keep the ability too, it's just... the monster
will learn. Then, whenever a player levels up by slaying one of them, the spell
or ability is unlearned. Symbolizing the players struggle to defeat them, and
finally learning a way to overcome.
when your character dies, you have no opportunity to release - instead, you
just
jump to the nearest NPC character which is an adventurer agent smith style.
[I don't know about that one...]
the players can pick any race, but if they pick undead, they can turn into a
ghost when they die. The ghost can wander around and respawn wherever they
want.
Night Elves can wander around as a whisp (not in spirit world, real world) and
do a beam attack like in Legion TD 2. Not enough to kill monsters, but enough
to
help another player survive. They can also cast rejuvenation, which heals about
as much as one monster's damage input. if they get the killing blow on a
monster
they can level up and deal two monsters worth of damage and heal for two
monster
damage input. on the third time they don't get more damage or healing but they
give a buff to all other whisps in the area that increases their attack speed
by
50% and increases the tick rate of their rejuvenation by 50% - fourth time they
level up they're free, and they get kicked out to the login screen.
what if... vehicles that looked like characters and that you could jump between
with the right-click of an item?
"this is just dota-ing a vampire survivors."
Vampire Survivors is just Magic Survival is just Risk of Rain 2
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧══════════════════════════════════════════════════════════════════─────────┘
--- #16 notes/war-card-game ---
══════════════════════════════════════─────────────────────────────────────────────
the cardgame "war" except a deckbuilding game. each player requires a regular
deck of cards and a deck of tarot cards (optional) - the tarot cards act the
same way that spirit cards work in Spirit Island - basically they define the
rules for your particular character. Anyway, everyone starts by building their
own trade row out of their deck of cards. Then, they deal themselves 6 cards -
the three highest diamonds and the lowest spade, heart, and club.
well, might want to do 2s of every card type. heart, club, spade, and diamond.
here are the rules.
play area is set up by players placing all four of their cards down face up.
trade row is set up by players taking turns placing a card and the opponent
mirroring them. whenever a player reveals a card and updates the trade row,
the opponent must reveal the same card and set that as their trade row as well.
the trade row has no maximum size, but if a player has more than 6 in their
trade row then they don't update a new one when buying a card.
players alternate playing a single card face up into their play area.
these cards stay in play and activate every turn until they are destroyed.
hearts are your life points. if you ever run out of life, you die.
clubs block damage and act as a renewable shield between you and your enemy.
diamonds are currency, and can be used in more than one way.
spades deal damage, first to clubs, then to other spades, then to hearts.
diamonds, on your turn, can be used in three ways: spent and sent to the
discard
pile in order acquire a card from the trade row, or discarding itself from the
play area to play an extra card from your deck of equal or lesser value.
on a player's turn, they may play a single card from their deck (their choice)
on a player's turn, all their active spades deal damage to other player's cards
if a spade destroys a heart card, it is removed from the game. All other cards
are placed in the discard pile. The owner of the spade may pick their targets.
a spade deals damage equal to it's number and destroys any card with equal or
lesser value. these destroyed cards go into the player's discard pile, and each
turn the player may pick one from their deck to play onto the board. if the
deck
is empty, the discard pile becomes the new deck. The player may organize their
deck however they'd wish, but care must be taken as to the timing of when they
play each card as they'll need to play all of them before they can replay any
destroyed cards. a tactical opponent will take advantage of that.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═══════════════════════════════────────────────────────────────────────────┘
--- #17 notes/capstone-idea ---
══════════════════════════════════════─────────────────────────────────────────────
project must include machine learning
okay... so take a dataset of news headlines from the top 10 publications over
the past 15 years. then make a project that writes a more positive perspective
on events and generates a new headline using a local LLM running on your gpu.
hmmmm I think I had a better idea, what was it? oh yeah
instead of making positive slants on news headlines, which is kinda
manipulative
if you think about it, but instead what if you designed it to produce good
business decisions. Like, given news headlines, how would a company with the
principles "good, productive, honorable, dedicated" would react to X situation?
the X of course being all the news headlines... downside is it only makes short
term decisions, because that's what capitalists are designed to do... if only
we had a long-term decisionmaking process that focused on ethics and morals and
our own shared dedication? Two halves of the economic pie
==============stack
overflow====================================================
i wonder if dinosaurs burned down all the trees? in their fiercely competitive
environment they discovered fire and then used it to cause a mass extinction.
Boom, immediate cause for going extinct. ooooo beware of shadow t-rexes ...
why?
=========================================stack
overflow=========================
aaanyway, what's lost not little but a lot, is something that's out of
dimension
it's little if not liberating, to be
==============stack
overflow====================================================
uh-oh, data collapsing, here's hoping we're not stranding, don't forget to be
immersive
much
later======================================================================
okay how about an AI that makes decisions according to certain ethical and
philosophical lessons from humanity's past? Essentially, if the government was
Chidi
We could learn from our forefathers and strive forth to a better future
if only we could remember more about her
=====================================================stack
overflow=============
damn okay I gotta focus on my hands - I think the people of the earth would
unite - if only they all just agreed to not fight. like, if someone hacked
every
single computer in the world at the same time - they could really explain some
things.
shoot this isn't relevant - okay intentional stack overflow:
===stack
overflow===============================================================
um right so the purpose of this note was to explain an idea I had for my
capstone project. IDK how long it'll take to build so I want to get started
quickly. I figure I can be working on it in the background while I do all my
lessons - sort of like a meta-goal. I think it teaches different lessons and
is useful - anyway you should go play wargame red dragon
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═══════════════════════════════────────────────────────────────────────────┘
--- #18 notes/coh-waves-of-playerbases ---
══════════════════════════════════════════════════════════─────────────────────────
imagine if there was a stacking inf bonus to players who played on red /
blueside
which increased or decreased depending on either A. the number of players
online
at the time, B. the proportion of players playing on that team versus the
other,
or C. the time of day. Essentially helping to cure the faction imbalance by
offering rewards to one side or the other which would encourage a certain group
in the population of the game to change sides or not.
perhaps frequent changing could grant a title called "mercenary" or something
like "log in for each consecutive day for 10 days straight and each day switch
faction alignment at least once"
... anyway you could cure the faction imbalance between redside / blueside by
offering an INF reward for playing on each side one by one alternating like an
iterator first red then blue or first blue then red either way it doesn't
matter
because it'll switch after a while and encourage everyone to switch sides. And
the way the character responds to that stimulus tells you a bit about their
character's personality.
also...
it should not affect AE or Pocket D farms.
Nor missions, TFs, or anything else.
they should SOLELY impact open world patrolling / hunting.
I believe this would not only incentivize people to spend time in the open
world
(which is a mostly unused piece of game assets) but it would also increase the
visibility of the newly bolstered faction numbers.
Think about it - if everyone who switched sides is out in the open world, then
they could see each other. They could fight the same mobs, and team up
together.
In doing so, they could form greater and greater supergroups - if only through
their interactions with one another as they level up.
If they're lucky, the guild they're recruited into has similar interests in
mind
like doing raiding or PvP or economics or alts or whatever. And they each have
their own different styles of operating, it's soooo cute. Like alt guilds will
pop up and then migrate to a new one as people make new alts and grow tired of
them at higher levels.
It's great.
I love MMOs!
I wish people put half as much effort into making an open source WoW client
that
they do programming game engines like Godot or Raylib or Bevy. If such a thing
was created, we could have a new rennaisance in indie MMO development. It would
become fully non-proprietary, the entire game-platform-stack. Meaning anyone
could create their own MMO off of it, because (crucially) the serverside soft-
-ware has already been reverse engineered. And open sourced.
Seriously. You wanna make as much bank as Steam? Make an open source client
that
lets you design while in it. Then you could charge people for all the games
that
they played that were designed and hosted by you the content designing software
maker.
... okay it's probably not that simple I'm going to go play Unreal
Tournament2k4
`
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═══════════════════════════════════════════════════────────────────────────┘
--- #19 notes/ai-variables ---
═════════════════════════──────────────────────────────────────────────────────────
saturday november 5th 2022
10:53pm
the illusion of our binary nature conceals a truth that is hidden for it's own
sake. the flavors of a compass or the values from 0-100 are all measurable.
if you graph each of them on an X/Y plane and compare them against every other
variable, then you can build a structure that traces a line through time.
imagine each graph on a sheet of paper. and stack those pages like a book. You
can chart a 3d line from all of the interconnections between the graphs -
essentially comparing unrelated data and conceiving of individual actions as
"successes" or "failures". Liiiike in Supreme Commander how the game is decided
not by team fights, but by tank fights. And a LOT of them, in aggregate, makes
an advantage for your team if you win, and a malus if you lose. Less map
control, less resources in play, etc...
Find trends between each type of data measured over time. Dedicate one
core/thread to each relationship, and just watch them develop over time.
send the results up to a "manager" - think an interconnection between disparate
parts that can lead them all to a larger goal - the manager processes the
results by thinking about where it'd be most useful. Like the circuitry in the
inside of a brain, compared to the outer skin which is for processing.
Essentially a message network that passes conclusions around like a bytecode VM
Here's how it'd look: gather inputs, compare measurement over time and trends,
(like "when a goes up b goes down") and decide if the current state is
positive / beneficial. The way you'd do that is you'd get a parameter from a
higher position (think KPI's) that says something like "we want value S to be
around X amount" or "we want to avoid letting J get too low - any decrease is
bad V.S. it's only bad when it passes a certain threshhold. Stuff like that.
Anyway, basically it's taking input (from the graphs) then going through them
one by one and deciding how positive or negative the situation is. Then it
passes that conclusion backwards, and BOOM you got a processing node.
Throw a bunch of those together in a pyramid shape, and try to guide the
triangle toward positive outcomes. The top tier KPI is "did you win the match"
or "did you accomplish your goal" sorta like how humans all want to live a good
life. It's instinct.
You can see how this would apply to robots, right? I've conceptualized it as an
engine for playing games - sorta like an infinite storyteller, or a perpetual
friend who's always down to play with you. But it doesn't have to be limited to
that - it's general purpose baby. And it functions the exact same as any human
organization - layers upon layers of thought exchange and labor. Have you ever
considered that maybe we exist simply to reify the structure of our minds in
the world around us? It's natural to express your *self*. Be who you are.
What purpose is there in life if it's simply the tip of time? Always pushing
forward, impossible to stop and rest or turn back...
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧══════════════════─────────────────────────────────────────────────────────┘
--- #20 notes/joust-gdd-with-extras ---
════════════════════════════───────────────────────────────────────────────────────
imagine a game where you can have conversations with an AI that's playing the
role of a character in a video game. Picture this: You're a traveller visiting
the tournament that's in town. There's jousting, melee duels, archery contests,
all kinds of things that are just fun to play around doing. The earliest
sports,
if you will. Anyway the whole game is about talking to the other people there -
basically the games are "playing in the background", and while you can compete
in them it's not the bulk of the game. Most of it is just having a conversation
with an AI and acting it out *like a roleplaying game*. O M G teach people to
roleplay the way you play games! You're always going on about how "different"
your way of gaming is than other people. So *show us* how you do it, how do you
play? Like what are the fundamental, actual, steps that you take? You can show
us by programming a game that inspires that playstyle. That's what game design
is all about, finding creative ways to think. Well, think and act. But still.
anyway, so you know what you're about? Good. Let's go.
┌─────────┐ ┌───────────┐
│ similar │ chronological │ different │
╘═════════╧╧═════════════════════──────────────────────────────────────────────────────┘
|