One Hour One Life Forums

I wanted to write my own alternative to gene score: an algorithm for determining the current influence of a player on their descendants.

So there's now a Rust crate for downloading and parsing life logs: https://crates.io/crates/ohol_public_data

It's probably not super intuitive, so if anyone decides to use it, feel free to open an issue. Though the only two functions you'll really need are get and url::build_url.

Signs are not cosmetic.

If there were real people in the game, they'd invent sticks with baskets with pieces of paper in them, or some other equivalent of a semi-permanently anchored piece of paper.

Jason repeatedly said that he doesn't want to have a situation like in Minecraft where everything is covered in tons of useless signs. But it seems like premature optimization to me. We don't know if this will ever be a problem (and I don't think it will, given that space is valuable in OHOL and that dismantling signs is possible), but if it will, it can be fixed later.

It should be all relatives, not just kids. The greater the distance between Alice and Bob in the family tree, the less genes they have in common. Exponentially so, even. But also the more other relatives Alice has that are as removed from her as Bob. Also exponentially, I think.

IMO the simplest way to calculate gene score is the sum of all living relatives divided by 2^(distance):

score(Alice) = living_players.map(relative => 1 / 2 ** distance(Alice, relative)).sum()

Edit: except this only accounts for the current life. The same should be calculated for all previous lives too, and summed up. (This does pass the sanity check where the total weight for any other relative is never greater than 1).

Only in my mind, haha. English-as-a-second-language problems, sorry.

Yay, I haven't been this excited since the genetic fitness update! I look forward to being able to travel east and explore the ruins :O

I hope in the future different cities will make different technological decisions based on their local circumstances. This will help with two things: make every life different by making every city unique, and make the ruins visually surprising and different from the player's home city.

How big is R?
I guess ideally the radius is just big enough that it's no longer economical to gather resources outside of it? Though that's naturally achieved by the well mechanics.

In the middle of every grid cell there will be a place for a rogue city with a new well.
Also with some coordination it will be possible to have 4x as many cities as Eves in a grid cell: cities can be placed with interval R instead of 2R.

What are some other games that came close?
Minecraft? EVE Online? Space station 13?

I think Dodge is saying that homes would indirectly help us find our family members. Home as a place you regularly visit, perhaps to store your personal stuff. A shelter is not enough: a hotel or a bus stop is a shelter, but not a home. If there was a need to have such a place, family members would naturally either share homes or build them close enough to be able to find each other.

Wouldn't make people care about their family, but it would make it easier to actually find them.

Re: figuring out the direction of the causal link, let's formulate a few hypotheses and do an experiment! We have a controlled environment, we can do science

Nitpick, but I think it's virtually impossible to force players to behave like that. We do have something like free will: the ability to ignore external motivation, be it economic pressure, hardwired evolutionary goals, or the rules of the game.

On the contrary, games are the only art form where it's possible to actually check the designer's ideas.
Like "lies, damned lies, and statistics": you can still lie with statistics, but it's harder that just making stuff up.

And he does. (And I'm daring to speak for him because I have my own desire to explore human behavior through games.)
He's not preaching that people should be doing something he wants them to do. He's trying to create a situation where they'll do it by choice.

As I understand it, genetic fitness is a way to incentivize players to do the correct thing. There's no point in comparing players to each other, right? So I think Elo score is not a good analogy.

Neither are fitness functions from genetic programming. A fitness function measures performance of a single agent, not a gene. That would be similar to measuring performance of a single life in OHOL.
A better analogy for measuring performance over multiple lives would be reward functions from reinforcement learning.

It's interesting that players with hetuw's mod played twice as many lives, proportionally: 7% players, 15% lives.

Does the client protocol change often?

Reminds me of "The Goal": a book about management that argues that virtually every enterprise (in-game civilization/game development) has a bottleneck that limits everything else. There could be hundreds of different little things to work on, but doing some of them would widen the bottleneck and therefore affect the whole system's output. The rest can pretty much be ignored forever.

Yay, this is interesting!

It'd be cool if the restriction was less artificial. It's true that one can learn only so many skills in a lifetime, but only because our time is limited. Over time humans got dramatically better at teaching each other. It'd feel better if learning was limited by time and resources.

I love that games like OHOL allow us to learn about real society, but there are two dangers:
1. Inconsistent rules and hardcoded, non-emergent outcomes that turn the game into a just-so story.
2. Consistent, but unrealistic rules that turn the game into a rigged thought experiment.

Another thing: I wonder if recipes would be a better unit of learning instead of items. For example, you could learn a random recipe associated with the tool every time you try to use a recipe you don't know.

That's probably impossible without super hardcore economic modeling.

It's also similar to the tutorial lands, the Doomed City book, and the superstructure in Pushing Ice.

I think it's super cool to shape the civilization by literally shaping the environment they live in.

The question is, what should the civilization be like in the game, and what shape would make people naturally follow that plan?

This is basically playing God, yay. Though so is the rest of game design, I guess?