Why Mneme Splits Off Its Own Economics — and Why CE/Traveller Stays Exactly Where It Is
A development note from the World Generator, v1.3.141
The short version
The CE World Generator now has two economic engines, and you pick which one runs. The default is CE/Traveller Economics — the one most of you came here for, the one that has worked at the table since 1977 and still works. The toggle is Mneme Economics, which is a different beast with different assumptions, built for a different kind of campaign.
This post is about why both exist, why neither is “wrong,” and what’s actually under the hood now that v1.3.141 is out.
If you only run classic Traveller-style sandbox games, you can stop reading after the patch notes. The split is there so the tool doesn’t bother you with anything you don’t want.
What’s fixed in v1.3.141
A quick rundown for people tracking the generator:
The economic calculation has been rebuilt from the ground up and split cleanly between the two modes. Culture notes are reading credibly now — no major contradictions across rolls.
There’s still a tuning issue I want to flag honestly. The Annual Income / 52 × 3D6 formula for weekly trade is producing some wild results. I’ve got an X-class starport with a population of 234,104 generating ~76.5 million credits a week on baseline, and one roll spiked it to 917.9 million — about twelve times normal — with no ships in-system to move that volume. The 3D6 multiplier is acting more like a scaling factor than a variance band. I’m leaning toward replacing it with a plus-or-minus percentage swing around the baseline, with the floor never dropping below Annual Income / 52 × 3%. I’d like to hear what the CE/Traveller crowd thinks before I commit to that change — the current behaviour is technically RAW-adjacent, but it breaks immersion the moment a player asks “where did the money come from?”
Orbital mechanics: Sphere of Influence and Hill Sphere Radius enforcement kept getting violated despite the constraint tools. That’s now working — moons no longer cross into other planetary orbits, and any world that exceeds an orbital slot’s capacity is correctly demoted to a rogue. Next problem on the list is orbital spacing. I have a system right now with a gas giant at 8.9 million km, a dwarf at 16.4, a terrestrial at 23.9, another dwarf at 3.9, and a second gas giant at 46.3 — all packed into less space than Mercury-to-Venus. They’d play stellar pool with each other inside a million years. Fix incoming.
The 2D system maps at the bottom of the output are new. Best viewed on desktop. Brachistochrone trajectories are calculated as straight lines for now — I’ll iterate toward curves that respect SOI/HSR boundaries and add waiting-period optimisation, with the algorithm recommending either cheapest-delta-V or fastest-given-delta-V routes.
Coming next: name/place/faction generators as opt-in toggles, then auto-economics for factions, then auto-relationships between worlds. That last one is where the two economic models actually diverge, and that’s what the rest of this post is about.
The thing CE/Traveller does that I don’t want to lose
Classic Traveller economics — and CE faithfully carries this forward — is frozen in late-1970s/early-1980s economic assumptions, and I mean that as a compliment. It assumes that wages, productivity, and the cost of labour stay roughly stable across tech levels. A starport worker on a TL-7 world earns broadly comparable wages to one on a TL-12 world. A merchant captain’s monthly nut is what it is, regardless of whether the ship is built around vacuum tubes or fusion plants.
That assumption is not a bug. It’s what makes the entire trade game playable. Cargo runs work because prices are legible. Speculative trade works because arbitrage between worlds is computable on a piece of graph paper. The whole loop — buy low on world A, jump, sell high on world B, pay the mortgage, hire a gunner, repeat — depends on a stable pricing universe where the player can reason about margins without a spreadsheet that updates every tech level.
This is the same reason D&D’s gold piece doesn’t track inflation across centuries of in-world history. Some abstractions are load-bearing. Touch them and the table-facing game falls apart.
So CE/Traveller mode in the generator preserves all of this faithfully. If you toggle it on, you get the economics you’ve been playing with for forty-plus years. The numbers behave the way you expect. The trade tables still work. Nothing has been “improved” out from under you.
So what’s Mneme doing differently, and why?
Mneme is the SF setting I’m building alongside this tool, and it has a different problem to solve. It wants to handle megastructures — gigaton, teraton constructions that reshape star systems — and it wants those structures to make economic sense rather than appearing by GM fiat. That requirement breaks CE/Traveller’s frozen-wage assumption almost immediately, and here’s why.
If wages and productivity stay flat across tech levels, then a TL-15 civilisation has no economic foundation for building anything beyond what a TL-9 civilisation could build with more workers. Population scales linearly; output scales linearly. You never get the exponential productivity curve that megastructures actually require. Without exponential productivity per capita — the kind that comes from cascading automation, where TL-9 humans direct drones and agents, TL-10 humans direct robots that direct drones, TL-11 humans direct fleets of self-replicating constructors — you cannot terraform a planet, let alone wrap a star.
Mneme assumes that each tech level genuinely changes the life of the general population, because if super-science doesn’t translate into broad quality-of-life gains, no state and no population has any reason to fund it. Super-science that doesn’t reach the average citizen gets defunded by the average citizen. So in Mneme, salaries rise, productivity rises, and crucially, the relationship between tech levels gets weird.
That weirdness is where the four pillars I actually want to talk about come in.
Productivity, Friction, Arbitrage, Inefficiency
These are the four levers I’m using to make Mneme’s economics generate stakes rather than just numbers. They’re also, in my view, the four levers that make any setting feel like it has a real economy underneath it — including settings that use straight CE/Traveller rules. The difference is just whether the levers are explicit or whether they’re handled by referee judgement.
Productivity is the per-capita output curve across tech levels. In CE/Traveller, this is roughly flat by design. In Mneme, it’s exponential, and that exponential curve is what funds infrastructure that would otherwise be physically impossible. A TL-13 polity isn’t just “TL-7 but with better guns” — it’s a civilisation where one citizen’s labour, multiplied through automation, equals the output of a small TL-7 city. This is what makes the megastructures work, and it’s also what makes the political situation interesting, because…
Friction is what keeps that productivity from flowing freely. High-tech worlds want lower-tech immigrants because automation has bottlenecks too — there are always jobs that don’t economically justify a robot, or that need messy human judgement, or that the local citizenry won’t take. So the high-tech polity dangles full citizenship as bait. But citizenship pathways have friction: paperwork, residency requirements, language requirements, loyalty oaths, vetting against the immigrant’s home polity, quotas driven by domestic politics. The state’s interests and the immigrant’s interests are not aligned, even when both parties technically want the same outcome. PCs trying to move between tech levels — and most interesting PCs are trying to do exactly this — run head-first into this friction.
This is, not coincidentally, how the real world works. The US, China, and the European middle powers all need the rest of the world as a market and as a labour buffer, even as their domestic politics make immigration an open wound. Client states get pulled into proxy conflicts they didn’t pick. Buffer zones become flashpoints. None of this requires any of the parties to be evil; it just requires them to have legitimate, conflicting interests at different scales. That’s the texture I want at the table.
Arbitrage is what PCs do about all this. The classic Traveller speculative-trade loop is one form of arbitrage — buying cheap and selling dear across jump distance. Mneme adds tech-level arbitrage on top of that: moving labour, knowledge, contraband, refugees, ideas, and patents across the friction gradient between polities. A TL-9 engineer with TL-12 schematics is worth a great deal to someone, and is also a problem for several other someones. The PCs are the people who notice these gradients before the relevant authorities close them.
Inefficiency is the gap between how the economy should work on paper and how it actually works in practice. Every polity has these gaps. Corruption, regulatory capture, supply-chain vulnerabilities, smuggling routes that exist because legitimate routes are taxed into uselessness, security forces that are nominally loyal but practically for sale. In CE/Traveller, the referee inserts these by hand when the story needs them. In Mneme, I want them generated procedurally so that the players can find them by looking at the data — so that a clever player who reads the system map and the trade flows can identify an exploit the GM didn’t plan for. That’s the kind of agency I’m trying to protect.
Why this matters for player agency
Here’s the part I want to be most direct about, because it’s the actual design argument and not just worldbuilding flavour.
If a setting’s economics aren’t grounded in something — productivity, supply chains, labour flows, the actual cost of moving atoms around a star system — then everything in the setting is GM fiat. Space stations appear because the GM put them there. Fleets show up because the GM rolled on an encounter table. Security responds (or doesn’t) based on dramatic timing rather than actual coverage. This is fine for some games. Plenty of great campaigns run this way.
But it changes the stakes. If a player blows up a refinery and nothing economic follows from that — no shortages, no price spikes, no opportunistic competitors moving in, no security review of the surviving facilities — then blowing up the refinery was a cutscene, not a strategic act. Conversely, if a player notices that a system’s grain imports are being routed through a single chokepoint and proposes a venture to build an alternative, but the setting has no underlying logic that would reward that insight, then their cleverness was wasted.
What I want — and what Mneme economics is trying to deliver — is a setting where players who peek behind the curtain are rewarded rather than disappointed. Where the friction, arbitrage, and inefficiency are real enough that a strategic player can draft plans against them, and where their plans have consequences that propagate.
That requires an economic foundation. It requires productivity that scales with tech level so that the infrastructure can sustain itself. It requires friction that comes from somewhere identifiable, so players can work with it or against it. It requires the GM to not be the sole source of what exists in the setting.
CE/Traveller doesn’t try to do this, and shouldn’t. Its design goals are different and its design goals are correct for what it is. The frozen-1980s economy is a feature: it keeps the trade game playable, it keeps the math at the table, and it gives the GM a clean canvas to paint whatever stakes the campaign needs. If you’re running a Traveller game, that’s the right tool. The generator’s CE/Traveller mode will keep doing that, faithfully, indefinitely.
Mneme is just trying to do a different thing, and now it has its own toggle to do it in.
What’s next
Once the orbital spacing is fixed and the name/place/faction generators are wired in, the auto-economics layer ships next. In CE/Traveller mode it’ll generate trade figures and relationships using the rules as written. In Mneme mode it’ll generate productivity curves, friction gradients between neighbouring polities, and the proxy-conflict and buffer-state dynamics that fall out of those gradients.
You’ll be able to run either. You’ll be able to switch between them on the same generated system to see how the same astrography reads under two different economic models. That comparison is, honestly, half of why I built the toggle in the first place — I think looking at the same star system through both lenses tells you something useful about both.
Patch notes for v1.3.141 are above. Bug reports welcome, especially on the orbital spacing — if you generate a system that looks wrong, send it.
Leave a Reply
You must be logged in to post a comment.