a multiplayer game of parenting and civilization building
You are not logged in.
Once you have crafted all the objects there are in the game, I doubt people continue to play it for super long after that.
Most games don't have crafting, there are other reasons to play.
I'm not building the tech tree with some high-level goal in mind, top-down. I'm instead looking at what tech we currently have and thinking: Where can I go from here? What's the next step? And most importantly, what would people who had this current tech be desperately trying to make next? Probably not glass or gunpowder.... they'd probably be trying to farm some more crops and domesticate some more animals.
"If I had asked people what they wanted, they would have said faster horses."
Technology depends not only on people's needs, but also on physics. You don't have that limitation since you're the one creating physics in the game.
Unless they're from Earth and they're remembering Earth technology, not inventing? In which case they'd be gunning for electricity.
I think having to build wells far from home is good, makes it necessary to preserve information about their location.
breezeknight, I think you're attacking a strawman. I haven't mentioned morality at all.
Humanity is supposed to simulate human empathy that makes one feel bad about killing someone or having a social fight with someone, regardless of morality or consequences.
Humanity is generated by all elderly people, so it helps the party that has more people that live to old age.
I agree with this. What I meant by "not up to the designer" is Jason, himself, monitoring specific players and making a judgement call on whether or not that person is a griefer.
I agree, that's not sustainable.
Joriom said that he kills most players by sabotage, not a knife. That is definitely griefing, but clearly not murder.
Oh, by "parties" I meant groups of players, so I intended to include (one kind of) murder as damage.
But it's true that there's a difference between property damage and homicide. Death is inherently irreversible, plus it directly affects the attacked party's ability to recover and retaliate.
So there are two kinds of terrorism: murder and vandalism. They're more specific than homicide and property damage: homicide includes lawful executions and combat; property damage includes sabotage as a warfare tactic.
What if we added this?
5) It is possible to recover from mismanaged resourcesNow we've got a new question (and one I like a lot), which is, "how do we implement (5)?"
(5) may already be possible in the game, but not if everyone in village is dead, which takes us back to murder. So, maybe murder is the appropriate topic for this thread, as opposed to griefing. Hm.
Recovering from vandalism can be implemented as making almost all actions reversible. The more numerous party wins if they spend exactly as much energy as the less numerous party attacking them.
Note that to revert an action, it's necessary to know about it in the first place. For example, finding hidden items is harder than hiding them; footprints could help with that.
Recovering from murder is a harder problem because a) respawns are free, b) death is irreversible, c) the attacker has the initiative and could manage to kill two. When the attacker is born in the same party she attacks, there's also d) resources wasted raising them.
One way to fix this is to make homicide resource-intensive, so that there's a necessary resource gathering step. And those actions are reversible.
For example, there could be a slightly VtMB-like humanity system:
Humanity meter that doesn't reset on death.
Fighting costs humanity.
Zero humanity makes one break down crying and starve to death. Or turn into a bear. Or both.
Shouting burns humanity to decrease someone else's humanity.
People over 50 regenerate humanity and can give it to others.
By the way, here's another kind of griefing. Raiding is using superior numbers to destroy already existing villages by spawning there or nearby. Defending should probably be easier than attacking, and that's why in the above list humanity regenerates only after 50, but that's just a half-baked sketch of a solution.
Perhaps we really should move to a separate thread.
Edit: clarified the last point of the list.
I would agree with you if we were talking about PVP FPS/TPS or RTS game. It's designer's job to balance things there. But OHOL is survival game. There are literally hundreds of ways you can fail in it. If you balance all this things in a way that cannot be exploitable by griefer, you'll get completely different game, definitely not survival one.
I see two ways to interpret this:
1. "Removing griefing from OHOL is hard, therefore it should be the community's job, not the designer's."
2. "Removing griefing from OHOL is impossible without ruining the game, therefore we shouldn't do it."
Am I reading this correctly? Which interpretation is correct (if any)?
If #1: How should the decisions be made? Why would they be better?
If #2: Why is it impossible?
I'm aware of this. That's why I'm asking for suggesstions and for pointing out flaws in my proposition, but then take decision to include or change anything on my own.
Uh, I was replying to aowen...
Some constraints:
- The primary goal of the solution is to give players a way to effectively counter reincarnating griefers. The consequences of being deemed a griefer must persist across lives.
- The solution should be resistant to a griefer exploiting them to generate false positives/otherwise affect a normal player.
- Automatic solutions (stuff that depends on the server determining if a player is a griefer) are too easy to exploit/maneuver around. As a dev, they're a nuisance to maintain. As a designer, you're essentially creating "rules" that define "good" and "bad" behavior (why would murder be necessarily "bad"?).
- - Since it's not automatic, and it (probably) shouldn't be up to the designer, the players must decide what counts as griefing.
Problem statements like this one are very useful.
While automatic solutions are most likely impossible, I disagree that determining what counts as griefing shouldn't be up to the designer. Griefing is a problem with the game's design, so it's the designer's job to fix it, like any other. And design by committee never works.
There are several problems here that are hard to disentangle, but it's still necessary to discuss them separately, and they need their own names, more specific than just "griefing".
Here's one: terrorism is the kind of griefing where the weaker party can cause disproportionate damage to the stronger party.
It's bad because in a highly competitive environment this would lead to everyone being stuck at level 0 forever. The real world is not a highly competitive environment, but it approximates one.
Expected a thread about everyone using their straw hat as if it was a cowboy hat expy
Gods are imaginary, but altars are not:
Altars are immovable, so their sprites can be procedurally generated on client based on their coordinates.
Altars can be decorated: they are containers, but objects inside cannot be removed and do not decay.
Unwanted children can be sacrificed in exchange for useless shiny trinkets.
Vote here: reddit link
Chance to use doesn't apply when increasing the number of uses?
Kinrany, not sure your math is right on the refill times.
Deep well contains 20 buckets of water, and refills one bucket ever 30 minutes. It refills completely in 10 hours.
Refill Rates:
Pond and shallow well, 2 water per hour.
Deep well, 20 water per hour
Oh, so deep wells measure water in buckets, like cisterns, all right.
Makes shallow wells even less useful, since one deep well is better than ten shallow wells :p
A term that might be useful: bootstrap technology -- tech that is only useful in the brief period of time when a better version is already available, but has higher overhead cost.
If you want to cut down a tree and you have stones, you could cut off a long, thick branch and make a big axe, or you could cut off a short branch, make a small axe, then use it to make the big axe faster. Then you throw away the small axe, because the big one is just better at everything.
Oh, and I found this: wiki.c2.com/?BootstrapTechnology
(Not the category I described above, but close and otherwise relevant to OHOL)
If anyone wants to make another tool for OHOL: it might be possible to automatically generate a Petri net or some other model based on crafting recipes, and then use static analysis tools to find the bottlenecks.
All water sources refill very slowly (one unit every 30 minutes), but dry sources aren't fatal. Ponds contain 20 water. Shallow wells contain 80 water but must be built on a dry pond. Deep wells contain 200 water in the form of 20 buckets of water, with 10 units per bucket, but they require a bucket to fetch water from them. Each bucket can be used to fill smaller water containers 10 times, and cisterns hold 9 buckets of water. Buckets can also go in carts now, so a cart can carry 40 water.
Ponds contain 20 water and refill completely in 10 hours.
Shallow wells contain 80 water and refill in 40 hours.
Deep wells contain 200 water and refill in 100 hours.
Unless my knowledge of the game is outdated, without buckets a cart can carry 12 water. With buckets and deep wells a cart carries 40 water.
(By the way, is it possible to use buckets with shallow wells? Because it should be.)
Is there actually any reason to build shallow wells?
Containing more water is better because you can get more at once: you spend less time moving water. But a full pond has more water than one player can carry without buckets.
Shallow wells seem to become obsolete the moment they become useful.
(Except if they're cheaper and can still be used with buckets, in which case it's probably better to build several shallow wells before upgrading them.)
What do we say to the god of death?
No
Jason, the main issue I see is you are refusing the idea of items affecting each other in the area. I think this is a key for further technology advancement and managing even current state. With area effects you can make things dynamic, self-managing and focus on farther development.
It's not a design choice, but a technological constraint.
It does not work.
Can we please have a magical tool that just outputs "Yes" or "No" randomly, based on the current tile's coordinates
Jason, I think you're confusing us and/or yourself by using "finite" and "infinite". Taken literally, many of your statements about finite and infinite resources are obviously wrong.
I don't really see why wastelands are a problem. Just don't spawn Eves there, the map is infinite. Have randomly placed Eve spawn points that can be deconstructed into resources, and pick the spawn closest to the center of the world.
I think salvaging should be like dungeoneering: find a dungeon, bring a cart full of food, leave with a cart full of artifacts.
Making it possible and desirable to build long roads would help with exploration a lot.
Roadside Picnic, but the aliens are actually your ancestors.
As a bonus, make broken tech decay into death traps.
I think the "Everything runs out" approach ignores time and doesn't work in practice. In the game, in the real world, or in any other sufficiently similar possible world. It's blind to the fact that, while there's entropy and human civilization is technically doomed, locally all economic systems grow all the time.
They still spend energy to sustain themselves, but they're mostly constrained by their ability to use resources. So the life of an economic system consists of two parts: the first part where it grows by increasing it's ability, and the second part where it dies because there's nothing to do but consume the available resources.
Maybe this is how successful villages should look like. Though the process of dying seems really boring.
It might be better to use something like a resource flow model. There are resources, and there are tools used to convert resources. Natural resources are sources, humans eating food to survive is a sink, and survival depends on the maximum flow between them being sufficiently high.
Except it's not a static network, so Petri nets might be more appropriate. Tools are resources too, flow capacity can be increased by making new tools. This improvement can't be always free though, or any system that is over the survival threshold would grow exponentially.
Using a tool has three costs:
the overhead cost paid once to create the tool
the marginal cost of the resources consumed in the conversion process
the fixed cost paid over time to maintain the tool.
Maintenance means that there are more sinks, and more energy is irreversibly lost over time.
Replacing "don't do X to Y, it's wasteful" rules with "when Y breaks, do X to restore it" could kill two birds with one stone: make in-game education more pleasant and intuitive, and make the economy easier to control.
P. S.
I love economics in multiplayer survival games, so if anyone knows of any places that have high quality discussions on this topic, I'd be really happy to be invited, please feel free to PM me.
Edit: broken link fix
But doesn't an infinite resource mean busy-work with no decisions to make?
I'm not sure if there's any connecting between the resource being infinite (in the "never runs out" sense) and busywork.
Note that if the village's progress is, due to stuff running out, a series of plateaus, each plateau will have it's own unique kind of busywork, but many characters will be born and die on the same plateau and never see anything change.
Resources being finite or infinite is village-scope gameplay, but busywork is character-scope gameplay.
There are a few things that do take time, mostly for thematic reasons. Making charcoal, growing plants and animals. It would be weird and confusing if you planted carrot seeds and instantly got carrots. But maybe this game should work like this. Waiting isn't interesting.
This.
You could also make actions cost food, and make the cost significantly higher than the cost of living. This way there'll be a trade-off between optimizing for food and optimizing for time, and wasting time will mostly harm the character, not the whole village.
That said, waiting can be interesting. The optimal strategy can be multitasking, doing several similar or different things in parallel.
I think it works well when the waiting times are long enough that multitasking makes sense in the first place, and also when the risks of losing progress because of multitasking are not too high. If you're smelting unobtanium, wasting half the time waiting is still better than smelting two batches in parallel and risking to lose one. Multitasking should be more efficient than waiting.
Waiting for a pond/well to refill is even less interesting than waiting for carrots to grow. But I guess I'm having similar thematic trouble here.... is this a world without rain? How can a well never refill?
Right, water isn't something you'd expect to be finite, since nothing actually destroys it. Unless you have a sci-fi reactor that uses it as fuel or something.
There's also something different here, because a well/pond refills over and over, and all without you doing anything. No resources are spent, no decisions are made. Growing carrots also takes time and waiting, but it's not automatic, and it uses resource along the way.
Ponds and wells could randomly become clogged. You'd need to clean them to make them refill, and you'd need better tools to maintain deeper wells.
Actually there needs to be only one thing that eventually runs out, right? If you have nothing to eat, it doesn't matter that you can turn air into oil.
I agree that a lot of the gameplay focuses on "not doing something," which is not very interesting. I never thought of it that way, but "don't pick the non-fruiting milkweed, don't drain the pond, don't drain the well, don't snare the non-family rabbit, don't forget to water the languishing bush" do kind of create a walking on eggshells feeling. I've drained a pond by accident, picked milkweed at the wrong time, etc, and I always feel like an idiot when I do this. I want players to teach each other things, but mostly that teaching should be how TO do something, not a long list of what not to do.
At least there should be loud, clear and immediate signal that you goofed up. Accidentally dropping your keys down the drain just sucks, accidentally erasing a production server is a disaster.
My friend says that only the top level tech should be infinite at any stage. So, for example, ponds are infinite at first, but then once I introduce wells, ponds become finite, and wells become infinite. Then I introduce infinite hand pump wells, and the old wells become finite. Then I introduce combustion engine pumps, and the hand pump wells become finite.
This!
And it only needs to be potentially infinite, not practically infinite. It's good enough if players will eventually accidentally kill themselves. Angry bees, pollution, grey goo, etc.
I think median number of deaths in a row would be more intuitive
The game seems to focus on two views of civilization:
subjective, where you see the civilization from inside, by living one life as a child and a parent,
and objective, outside view, where you observe the general dynamics and outcomes.
Family tree browser contributes to the second view, and I like that.
Your name is '; DROP TABLE players; '
Human society is literally the most complex thing we know about. And recreating a system is the ultimate test. I think we should be very, very, very pessimistic about our chances to make it work as expected.
After all, in real life, we are all single-hour players..... and yes, there were ups and downs along the way.
B-but Anthropic principle?
No matter how unlikely it is for a civilization like ours to succeed, we would still observe a civilization that did succeed, because we couldn't possibly have this conversation otherwise.