This here text is a round up of the conversations and activities I'm aware of with the Nth Society game project, 2nd edition.
Check out this link for an introduction to the project.
The first round up is available here.
While no actual design documents have been created, there has been plenty of brainstorming and conversation. I'll discuss that at the end of the post.
I hope some of these discussions crytalize into design specs of some sort.
Lua is the scripting language of choice, but 
Why another scripting language? As far as I understand it, it will help us to write in a platform independent way, using a structure borrowed from Prolog which is far better suited to game design - rule and query based programming. Games are explicitly about rules, and by generalizing data access to queries we are able to integrate that well into a distributed database.
Syntactically the language will be easy to learn and straightforward to script in, which is important as I want to facilitate the maximum variety of coding levels in world design - right up from complete beginners with an aptitude.
Meanwhile I have begun to spec out a network protocol influenced in part by Hashgraph. My personal shortlist of things to do looks something like this:
Recently there have been some really interesting discussions on the Discord server I run for the community. I'm going to try and summarize some of them as best I can.
Note that I'm not naming names just for simplicity but if you want to be attributed for the ideas you had let me know.
Perhaps this was the main topic. Players having a stake in the game means that they are invested in the game, have some skin in it (i.e. something to lose) and so are bound to care more and act more realistically, as this models the real life situation of our lives being meaningful, partially on account of imminent death.
Can you measure this stake reliably? Partially that could be replicated by insisting on some monetary investment (perhaps similar to Steem Power lock in on the Steem platform) and in time spent (though that is prey to the Sunk Cost fallacy). There are also the in-game items of possessions and possibly property (if strctures emerge for its protection).
Old age and death definitely have an impact on stake. It's clear that players have to be able to die, but it has been roundly questioned that players should have to die from old age. The case was made that players who know they have to die at some point will hedge their bets against their new life, and it turns the game cosmic order into a "reincarnation society", that is a society which operates on the sure belief that a soul (read: real person player controller) will come back as a new person (i.e. starting a new game).
If as you age you become more susceptible to disease, etc. until you are sure to die by say 100 years old, then there would be an incentive to kill yourself after a certain optimal age and attempt to resume your life, in a sort of weird magical rebirth cycle. This is clearly not realistic, but it is also disrupts the realistic staking in any one particular player life.
There were several ways around this that more or less came down to hacks: requiring a certain number of deaths after a certain age to get to better quality servers (a very disliked and challenged idea), punishing suicide (again, heavily challenged), a financial cost to rebirth that was balanced with cost to death (we had this originally, and it's a little better), and some others. Nothing conclusive was shown to make this viable, but the conversation ended with talk of balancing death to new life costs, so perhaps it can continue from there next time.
So removing old age inevitable death seems popular, both because it allows for more consistent gameplay and might increase life stake as players sunk cost, relationships and so on goes on.
Death is something quite weird in most games, just do a little mental survey and you will see that it is never realistic. This means that it's very atypical to treat death realistically in games, whereas you often find at least some other things modelled realistically.
It should be noted that this will come at the expense of simulating what the effects of personal old age is like in society. But that was not really discussed, so maybe for next time.
It was suggested before that time should slow down over the course of a life time, so that childhood is short, and each phase from then on gets longer.
This was remarked on, in a couple of ways. One is that if it takes too long to do things, like a build a house, then no one will do them, or the time cost will be way off balance. Yet on the other hand it should take time to do anything difficult, as this adds to the difficulty, and thus the worthiness of the effort. To this is another area of balance.
Another point is that if the simulation is too close to real time (one day in game is one day in real life, or similar), then it doesn't model anything we can't already do in the real world. Or to put another day, nothing is predicted in a complex model that goes no faster than the thing it is modelling.
It's clear that time will have to be faster than actual real world time, but how to balance this is not clear.
Also discussed in another conversation relatedly is offline time, and how it will be too difficult if too much happens when gamers are reasonably logged off, i.e. to do the important things in life such as work, sleep, eat and do their life! This is a really underdeveloped problem area, but there was the suggestion that maybe certain servers are only active during certain hours, which would kind of lock them into a time zone. This deals with the scenario where raiders in a different timezone come in and destroy everything while everyone else is asleep.
The idea of skills will be necessary to the game, processes that you learn to the ability to apply to the world to get some special outcome. For example, learning to build a wall well. Maybe every character can put stones on top of each other, but to know how to do each part of building a stone wall or wooden fence might take a number of learned skills.
This is a challenge for a game which intends to be realistic because the easiest way to implement a skills framework is to break the realism and give on screen hints, tutorials, skills progress tracking, skills relationship hierarchy (what you need to know to learn the next skills, like Civ tech tree), and even knowing what skills you know.
A good middle ground was found: to have obvious feedback in getting better at the actual processes, but to have no explicit skills feedback. So for example, the more you run, the faster you can run. The increase in ability is noticeable. This leaves the evidence for skills in the doing, so you will not readily be able to know that someone can do something without witnessing it first hand, trusting them or trusting someone else, or even a record of their skills. That is to say, the game will not step in and relay naturally hidden information as cosmic arbiter.
This still does not solve the problem though of things which require a kind of recipe. Such as making cement, starting a fire, or ... making a edible stew for example. Perhaps people should have to discover these things using their real knowledge of such things. I'm not sure and it wasn't fully discussed.
This all leads into the idea of an emergent economy: it was claimed that an economy cannot develop without specialization of skills, to the point of trades or even professions. This was not really resolved but it's an interesting point to consider.
There were a lot of reservations to these ideas though, mainly on usability and fun grounds. It should not be overly annoying or complicated, and players should not feel like they are fumbling around in the dark too much. So the level of difficulty and obfuscation should be generally satisfying, not prohibitive. It is a game after all.
In most games with the possibility of violence, it is the most prevalent means to succeed. People don't get to know other people readily and attacking is the first method of interaction.
While violence is quite common in real life, the over incentivization to violence is not realistic.
This point wasn't discussed that much but it is important enough to deserve to be noted so it can be expanded on.
Some method needs to be found to allow for a variety of naturally varying traits and aptitudes to be set when a player character is born. These would feed into starting levels for the various skills, as well as aptitude, i.e. the speed of improvement, and the upper bound on increase in ability.
While these are actually hidden from the player, as many as possible should be physically represented, i.e. strength should correspond to muscle size, and so on. However it should only be done when it makes sense. At the end of the day, the difference in aptitude will need to actually be tested, as anyone does in real life, by trial and error when getting to know oneself.
Hopefully this is mostly fair to all involved! Let me know if you think I should amend anything.
As always, the more discussion the better. Please join is on the Discord server or get involved right here in the comments section.
