This game explores such a captivating idea, well there's two ideas actually which I find particularly interesting about it:
The concept of computational reducibility. According to Stephen Wolfram, who is very prominently featured on HN, (he argues that) the most interesting systems are computationally irreducible, meaning the only way to find their state at step N is to simulate all steps 1 through N-1. Orbital mechanics is one of the rare counterexamples which says that: given a formula, you can compute a planet's position at any future time without simulating all intermediate positions.
This game deliberately positions itself on the opposite side of that line and uses computational reducibility as a design constraint rather than an obstacle.
The map generation is the other clever bit I find very interesting. An LFSR operating on a single byte is something you can compute with pencil and scratch paper and yet it produces maps with geographic clustering (lakes, mountain ranges) from the XOR correlations. You get interesting, realistic looking maps to play on without using a computer which I'll definitely steal for my next pen-and-paper game.
Suffice to say that yesterday I spend the next several hours obsessing over it then proceeded to write a rulebook for myself. In doing so, I ran into some confusion about the rules which I couldn't clarify from reading the post. jhylands, if you see this, please help me understand the rules a little bit:
1. Ore tier: hex or binary trailing zeros?
The rules say mountains have "an ore level equal to the number of trailing zeros in their hex value." But the worked example shows E0 as ore(5), which only works if you count trailing zeros in binary (11100000 has 5). E0 = 11100000 = 5 trailing binary zeros = ore(5). In hex, E0 has 1 trailing zero.
2. I couldn't see from the post how to get the first ore(0)?
A mine costs 1 ore(0) to build. A workshop also costs 1 ore(0). The player starts with no resources. Clearing forest yields wood, but no equivalent action exists for gathering stone from mountains. Where does the first ore come from?
3. Are pickup/drop/operate "actions" for the one-per-turn rule?
The rules say each agent performs one action per turn and list move, clear, build/upgrade, and operate. Picking up a resource, dropping one, and assigning an agent to a building are not listed. Are they free actions that don't consume the agent's turn, or do they count? I personally think it should be free since dedicating an agent to a building is a permanent assignment and not a per-turn action.
4. Should vehicle capacity be 2N or 2*N or 2(N+1) or ...?
The rules say vehicles carry "2N slots" where N is vehicle level. A level-0 vehicle would then have 0 slots, which makes it useless. Is the formula 2*(N+1), or 2^N, or something else? Probably 2\*(N+1) or 2^N would work here.
5. Can higher ore substitute in building costs?
The rules say "higher-grade ore can substitute one-to-one for lower" in the context of mines. Does this apply to all building costs that require ore? For instance, can ore(1) be used where ore(0) is needed for a workshop?
6. Smelter has "unlimited throughput per turn"
The rules say smelter throughput is "unlimited per turn." Does that mean a player can run multiple smelting batches in a single turn (consuming resources for all of them at once), or is it one batch per turn per smelter?
7. Can you put multiple buildings on the same square? The rules don't say you can't.
The rules don't mention any restriction on placing multiple buildings on the same cleared square. Is that intentional? Can a player build two houses on the same tile?
8. Small nitpick, LFSR worked example in the article is maybe incorrect.
The X-sequence for seed 0x01 is listed as 01, 03, 06, 0D, 1A. When I try it I get 01, 03, 06, 0D, 1B.
Tracing step(0x0D) by hand: 0D = 00001101, feedback = bit1 XOR bit2 XOR bit7 XOR bit8 = 1 XOR 0 XOR 0 XOR 0 = 1, result = (0D << 1) | 1 = 1B, not 1A. Was the feedback bit omitted by mistake?**
I had played around with some of the formulas to get everything to be 2^n but it it works better if L(0) is free and houses and mill/mine are 2^(n-1)
1) Ore tier is when the number is represented in binary.
2) I've made a mistake in my writeup here so I was changing the rules a lot as I played it to try and optimise but if you play with L(0) mill/mine being free then follow 2^(n-1) it works better.
3) Picking up and dropping are free, so the cost to move wood from x to y is just the travel cost. If operating, that is the only thing an agent can do. Assignment does not take time but the agent has to travel there.
4) I'll change it in the post but yes 2(n+1)
5) Yes higher ore can always substitute lower grade ore.
6) Yes unlimited batches.
7) I played it as only one building on a cleared square. The reason I'd keep this is because of the nature of houses means you're paying exponentially more for increasing density.
> simple and emergent
It's certainly interesting and all. But if this is your idea of a simple boardgame, what on earth do you consider a complex one?
It's more the simplest emergent thing I could come up with, over it being truly simple in the absolute.
Its a fair comment.
Thank you for making this!
This game explores such a captivating idea, well there's two ideas actually which I find particularly interesting about it:
The concept of computational reducibility. According to Stephen Wolfram, who is very prominently featured on HN, (he argues that) the most interesting systems are computationally irreducible, meaning the only way to find their state at step N is to simulate all steps 1 through N-1. Orbital mechanics is one of the rare counterexamples which says that: given a formula, you can compute a planet's position at any future time without simulating all intermediate positions. This game deliberately positions itself on the opposite side of that line and uses computational reducibility as a design constraint rather than an obstacle.
The map generation is the other clever bit I find very interesting. An LFSR operating on a single byte is something you can compute with pencil and scratch paper and yet it produces maps with geographic clustering (lakes, mountain ranges) from the XOR correlations. You get interesting, realistic looking maps to play on without using a computer which I'll definitely steal for my next pen-and-paper game.
Suffice to say that yesterday I spend the next several hours obsessing over it then proceeded to write a rulebook for myself. In doing so, I ran into some confusion about the rules which I couldn't clarify from reading the post. jhylands, if you see this, please help me understand the rules a little bit:
Thank you!
I had played around with some of the formulas to get everything to be 2^n but it it works better if L(0) is free and houses and mill/mine are 2^(n-1)
1) Ore tier is when the number is represented in binary. 2) I've made a mistake in my writeup here so I was changing the rules a lot as I played it to try and optimise but if you play with L(0) mill/mine being free then follow 2^(n-1) it works better.
3) Picking up and dropping are free, so the cost to move wood from x to y is just the travel cost. If operating, that is the only thing an agent can do. Assignment does not take time but the agent has to travel there.
4) I'll change it in the post but yes 2(n+1)
5) Yes higher ore can always substitute lower grade ore.
6) Yes unlimited batches.
7) I played it as only one building on a cleared square. The reason I'd keep this is because of the nature of houses means you're paying exponentially more for increasing density.