OpenFront Factories and Rail Income

A factory is just a dispatcher. The money comes from the cities and ports a train passes. Linking factories to each other earns exactly nothing.

And for a fixed amount of gold, spreading levels across several factories beats stacking them into one. The numbers are below.

1. Only cities and ports are paying stops

A train pays out when it reaches a building — except at a factory, where nothing happens at all. The factory stop handler is an empty function in the code.

So “link your factories together” is half right. A factory is the key that opens the rail network; the cities and ports on that network are what generate gold.

2. Who owns the stop is a 3.5× swing

And there is a rule on top of that which almost nobody knows:

If the stop belongs to somebody else, you get paid AND they get paid the same amount.

One pass through an ally’s city creates 35,000 for you and 35,000 for them, simultaneously. A pass through your own city creates 10,000, for you only.

So a route through foreign cities is 3.5× better for you, and creates 7× the gold on the board. That single line is why connecting rails to an ally is so strong.

3. Decay starts at the 11th stop

The count is of stops this train has visited on this journey. The first 10 pay full; from the 11th the payout drops by 5,000 each, bottoming out at 5,000.

The important part: this is not “how many cities you own.” It is one train’s journey log, and it resets to zero for every new train. With 50 cities, a train that visits 8 pays full on all 8.

You can watch this happen. Every time a train reaches a stop a figure floats up — +10.0K — and on a long route you can see it turn into +5.0K toward the end.

So a very long route is actively worse. A 20-stop route through your own cities pays 5,000 on the last ten. Two 10-stop routes beat it.

4. Build more factories, do not stack levels

⚠️ “Stack levels into one factory” does not work here. The denominator of the dispatch roll is not your factory count but the sum of their levels. The game’s unitCount is commented “Count of units owned”, but its body is total += unit.level() — read it as a count and you reach the opposite conclusion.

A factory’s chance to dispatch on a check:

attempts per check  = that factory's level
chance per attempt  = 1 / ((your total factory LEVELS + 10) × 15 ÷ saturation)

saturation = set by the number of trains on the whole map.
             empty map 1.45 → 300 trains 0.80 → 500 trains 0.50

The denominator sums levels, not buildings. So splitting levels across factories leaves the denominator unchanged.

What decides it is a separate ceiling:

One station dispatches at most one train, then waits 10 ticks.

A level-6 factory makes six attempts but stops at the first success — the rest are wasted. Six level-1 factories each check independently, so up to six trains can leave at once.

Upgrading costs the same as building new, so the only variable is how you split the same gold.

At a level total of 10 it is one factory 100% versus ten factories 144%.

These figures assume few trains on the map. The dispatch roll counts every train on the map, so as a game matures every factory slows down together — at 300 trains it is 57% of the empty-map rate, at 500 it is 36%. The ordering between setups holds.

This is exactly the port rule. A per-check ceiling of one means more checks wins.

But do not build factories forever either

The denominator is your level total, so every new factory drags down the odds of every existing one. Spreading still wins, but each additional site is worth less than the last.

The tenth factory does 29% of the first one’s work — and it costs more, since the price is pinned at 1,000,000 from the fourth building on.

So the real test is coverage, not dispatch rate. A factory only reaches 110 tiles. One that pulls in cities and ports not already on the network raises the income of every train, which dwarfs the dispatch effect. One dropped into an area already covered is buying 30% of a factory for a full 1,000,000.

• Brings in new stops → build it
• Area already covered → do not. Another city is the better buy
• Nowhere left to place one → only then raise a level

5. Trains do not choose the profitable direction

Do not count on it.

• The destination is uniform random among the cities and ports on the network
• The route is the shortest path. It does not favour the side with more buildings

So “the train heads toward the money” is not a thing. But you can invert it:

Since the destination is random, you want as few low-value stops on the network as possible. A network full of allied cities pays 35,000 wherever the train decides to go.

And the shortest-path rule is the reason to pack cities tightly. With cities dense between two points, the shortest path threads through all of them.

spread out   factory ──────── city ──────── city      2 stops
packed       factory ─ city ─ city ─ city ─ city      4 stops

More cities over the same distance means more money per train. There is no decay until the 11th stop, so aim to fill those ten.

6. Three placement rules

• Closer than 15 tiles and no rail is built. Packing too tightly prevents the connection entirely. And clicking within 15 tiles of the same building type upgrades it instead of building a new one
• No duplicate rail is laid between stops already closely linked. Clustering does not multiply routes
• Rail does cross narrow water. Do not write off a city on the far bank of a river

What rail can and cannot cross

Track may run over water tiles, with one restriction: from a shoreline water tile it can step out exactly one more. Narrow rivers and straits connect; open sea does not.

• A city across a river usually connects
• An island does not. To use an island’s buildings as stops, that island needs its own factory
• Water sections cost six times as much to route through, so an overland detour wins whenever one exists

The only way to check is to build. Once the factory is up, either the track is drawn or it is not. There is no way to measure it in advance.

Ports follow the same rule

One rule covers both:

A building produces at most one unit per check. So add buildings, not levels.

Levels only raise the odds of that single unit appearing. Factor in that a nuke deletes a structure whole regardless of level, and spreading wins twice.

→ Where Should You Build Ports?

Summary

Further reading

• All Six Buildings — factory prices and the shared counter
• Where Should You Build Ports? — a port is a stop and a trade hub
• Team Play — why you link rails across the team
• Building Cost Calculator — cumulative factory cost