OpenFront Factory Economy Commit: When a Train-Based Gold Engine Pays Off

Direct answer: commit to factories only when a connected rail loop pays

A factory-based gold engine is worth committing to only when the trains it spawns land paying stops at a controlled City or Port. A Factory building by itself earns nothing: the Factory Execution has no direct gold output, and all train income comes from the TrainStationExecution spawn loop, so a station with no connected rail reaches no station and produces no payout. If you can connect at least one controlled City or Port within 110 tiles and split the path into rail segments under about 155.56 tiles, the commit is justified; otherwise keep your gold in Ports, which pay on a schedule and never need a rail loop. The full question is not whether a single train pays, but whether a cluster of factories, a connected graph, and your map position together clear the shared cost ladder in time for the game to matter.

The short rule that replaces the “are factories useless” feeling is this: treat factories as a bet you place after the rail, not before it. First prove you have a paying destination and a connected network, then scale the factory count only when a completed paying stop, not a spawned train, is the thing you are short of. A spawned train is not income; a completed settlement at a valid City or Port, with the gold actually added to your total, is. That distinction is the whole decision, and the rest of this page walks the numbers that tell you when the bet wins. The reason the question lands at the ten-minute mark is that by then the early scramble has settled: the first few Ports are online, the first Cities are contested, and you have a surplus of gold with nowhere obviously to put it, so the 125,000 Factory looks like the natural next purchase and the “is it useless” doubt hits exactly when the build would be cheapest to regret. That timing is the trap: the moment you feel the surplus and the uncertainty is the moment the commit is least justified, because you have not yet seen a paying destination, a connected rail, or a destination rate that makes the math work. The honest read of the ten-minute feeling is that it is not evidence about factories at all; it is evidence that you have not yet built the three things the factory needs, and the fix is to build them, not to abandon the concept. See the payback math in the Port versus Factory guide and the rail geometry rules in the train network guide before you spend the first 125,000 gold on the build.

The decision framework: four thresholds before you spend

Decide the commit by crossing four independent thresholds, and fail any one of them and you stay port-only for now. The first threshold is destination control: you must hold or be able to seize at least one City or Port that a train can settle at, because a train that reaches nothing adds zero gold no matter how many factories you own. The second threshold is connectivity: the rail from your factory cluster to that destination must split into segments each under roughly 155.56 tiles, the diagonal cap of 110 tiles times 1.4142, because a segment longer than that cannot be walked by the station search. The third threshold is capital: the first Factory costs 125,000 gold, so you need enough surplus to fund the build plus a buffer without starving the army or the economy that keeps you alive while the train network comes online. The fourth threshold is time: the Factory builds in 20 ticks while a Port needs 50, but the Factory only starts paying once a train completes a paying stop, which needs a spawned train, a travel time, and a settlement, so the real payback window is much longer than the 20-tick build.

Treat these as a gate, not a checklist you skip through. A player who has destination and connectivity but not capital should frontload the Port instead, because the Port’s 50-tick build plus its scheduled trade income usually beats a half-funded factory cluster that spawns trains that go nowhere. A player who has capital and destination but not connectivity should not rush the first Factory; should first build the rail, because the rail is the thing that turns a factory from a 125,000 gold sink into a 35,000 gold per stop source. The framework keeps the decision honest: the commit is the moment all four lines are green, and the moment you know which line is red, you know which structure to build next. The building timing guide covers when each individual structure becomes the right next purchase, and the frontload port and factory investment guide covers the specific early-window version of this exact call. The capital threshold deserves one more concrete number because it is where the decision actually gets made. A 125,000 gold Factory is not a round-number purchase; it is roughly the cost of a full early army reset plus a buffer, so funding it means choosing to take the army down a level for the window while the rail and the destination come online. That is a survivability decision, not just an economy decision, and it is the reason the threshold exists: a player who funds the 125,000 and then gets pushed back to the rail while the factory builds has spent the buffer that was supposed to keep the economy alive. The honest way to check the capital line is to ask whether you can survive the 20-tick build plus the next 60 to 120 ticks of the train network coming online at one army level down, and if the answer is no, the Port is the purchase because its scheduled income does not require you to be exposed for the build window.

Why the factory count is the whole lever

The single most important mechanic for this decision is how the spawn odds scale with the factory count. In the v0.34.21 source, the TrainStationExecution spawn odds are computed as (numPlayerFactories plus 10) times 15, and the random.chance call uses that number as a 1-in-N odds denominator, so each tick the station has a 1 over (factories plus 10) times 15 chance to spawn a train. With zero factories that is 1 in 150, with 10 factories it is 1 in 225, and with 20 factories it is 1 in 450. The curve is deliberately sublinear, which is why the community keeps asking whether a second factory is worth it: the first ten factories move the odds from 1 in 150 to 1 in 225, and the next ten move them only from 1 in 225 to 1 in 450, a modest relative change in the spawn rate.

Here is the subtle part that the casual reading misses, and it is the reason the “are factories useless” thread is partly right and partly wrong. The spawn odds get worse as you add factories, but the train count you can run simultaneously is the lever that actually drives throughput. Each factory adds spawn capacity, and the existing-train throttle in the execution means you need multiple free slots before the next spawn can fire. The net effect, worked out in the community’s in-depth train economy write-up, is that going from 10 to 20 factories roughly doubles the number of concurrent trains while only halving the per-tick spawn probability, so the expected trains per second stays about the same, and the per-stop payout is what moves. A 10-factory setup running a single train at 35,000 gold per stop is replaced by a 20-factory setup running two trains at the same per-stop payout, which is the 33 percent throughput gain the community documents, not the 100 percent you would expect if spawn odds scaled linearly. The implication for the commit: the first ten factories are the bet with the cleanest payoff, and beyond ten you are buying concurrency, not speed. The practical read of that is to treat the first 125,000 gold Factory as the commit and the next several as scaling, with a very different decision standard for each. The commit is a one-time bet on the whole network coming online and pays back over the life of the game; the scaling purchases are incremental throughput buys that only make sense once the base cluster is already producing paying stops. A player who skips the commit and jumps straight to scaling, buying factories ten to twenty before the rail and the destination are proven, is paying the 125,000 cost for a 33 percent throughput gain on a network that is not yet producing, which is the worst possible place to spend that gold. The honest order is: one factory to commit, observe a completed paying stop, then scale in ten-factory increments only when the throughput, not the connectivity, is the bottleneck. That order keeps the 125,000 cost where the payoff is cleanest and keeps the scaling spend where the 33 percent gain is real.

The build-time asymmetry: 20 ticks versus 50 ticks

The second structural fact that shapes the commit is the build-time gap between the two buildings, and it is bigger than the cost gap. In the v0.34.21 source, the Factory build time is 20 ticks while the Port build time is 50 ticks, a 30-tick gap, and the Factory cost ladder and the Port cost ladder are the same shared 125,000 to 1,000,000 curve, so the first Factory and the first Port cost the same 125,000 gold but the Factory finishes 30 ticks earlier. That asymmetry is the entire reason the “commit early to the factory” camp exists: for the same 125,000 gold you get the factory’s 20-tick build plus its train income, or the port’s 50-tick build plus its trade income, and the factory is online 30 ticks sooner.

But the build-time advantage only matters if the factory has something to do when it comes online. A Factory that finishes its 20-tick build with no connected rail and no controlled destination inside 110 tiles is not 30 ticks ahead of a Port; it is 30 ticks into a 125,000 gold build that produces nothing, while the Port that started 30 ticks later is still building but will start paying on its own schedule the moment it finishes. The build-time gap is a real advantage only when it lands on top of an already-connected network and an already-controlled destination. Read it as a multiplier on the commit, not a substitute for it: the 20-tick build shortens the payback window, but it cannot shorten the time you need to connect the rail and hold the destination, which is the part of the commit that has to exist before the build starts. If you are deciding between the two at 125,000 gold with no rail yet, the Port is the safer purchase, and the Factory is the safer purchase the moment the rail and the destination are both in place. The 30-tick gap is also worth reading against the game clock. A 50-tick Port build is half a minute of real time at the standard tick rate, and a 20-tick Factory build is under half that, so the gap is not a game-changer in absolute terms; what it does is compress the window in which the factory can start producing before the next objective or the next pressure wave lands. In a fast map where objectives rotate every two or three minutes, the 30-tick compression is the difference between a factory that produces for most of the game and one that produces for the last third, and that is exactly the difference between a commit that wins the economy war and one that merely participates in it. In a slow map with long objectives, the 30-tick gap is noise and the build-time argument for the factory collapses to nothing, leaving only the per-stop payout and the route geometry to justify the spend.

The revenue ladder: 10k, 25k, 35k per stop, then the decay

The payout side of the commit is a clean three-tier ladder that the v0.34.21 source lays out in the trainGold function. A train that settles at a City or Port you own pays 10,000 gold per stop; a train that settles at a controlled allied City or Port pays 25,000; and a train that settles at a controlled team or other allied destination pays 35,000, which is the rate you are optimizing for in a team game where the destination is your team’s controlled hub. Every stop beyond the ninth then costs 5,000 gold, floored so the stop never pays less than 5,000, so a long route that visits 15 cities before the destination pays 10,000 minus 6 times 5,000, which is the 5,000 floor, and a route that visits 9 cities pays the full 35,000 with no decay.

This decay is the single most expensive detail in the whole commit, because it is the reason a longer “efficient” route is often worse than a shorter one. The wiki’s “first 10 cities are free” claim is an off-by-one against the source: the penalty is 5,000 times max(0, citiesVisited minus 9), so the eleventh city visited is the first one that costs 5,000, not the tenth. A train that visits 10 cities before the destination pays 35,000 minus 1 times 5,000, which is 30,000, not 35,000. The practical rule is to keep the train’s route to nine or fewer cities before the paying stop if you are chasing the full payout, and to accept the decay when the route geometry forces you past nine. The decay also caps the value of a single factory cluster: if your only paying destination is eleven cities away along the only viable rail line, your per-stop payout is 30,000, and no amount of factory scaling fixes the 5,000 per extra city you are paying on every train. The commit decision has to account for the decay at the route level, not just the factory level, because the route determines the payout and the factory count determines the throughput.

Network geometry: 110-tile search, 155.56-tile cap, and the 10-tick cooldown

The third structural fact is the geometry of the rail network, and it is the part that most often kills the commit in practice. The station search range is 110 tiles, the minimum link is 15 tiles, and the diagonal rail cap is 110 tiles times 1.4142, which is about 155.56 tiles, so any rail segment longer than 155.56 tiles cannot be connected by the station search and the train will not walk it. In practice that means a factory cluster and a destination more than 110 tiles apart need a chain of relay stations, and each relay station is its own build with its own 125,000 gold cost on the shared ladder. The geometry is also where the 10-tick cooldown lives: the TrainStationExecution enforces a minimum of 10 ticks between two spawns, so even at the theoretical maximum spawn rate the station cannot fire faster than one train per 10 ticks, and at 10 factories the 1 in 225 odds mean the expected gap between spawns is about 22.5 ticks, well above the 10-tick floor, so the cooldown is not the binding constraint at realistic factory counts.

The geometry is the reason the “are factories useless” feeling is so common in the early game: a new player builds the first 125,000 gold Factory, the 20-tick build finishes, a train spawns, and the train walks its rail, reaches the 110-tile edge of the network, and stops, because the destination is 140 tiles away and the rail segment between the last relay station and the destination is 155.56 tiles long and the station search cannot bridge it. The train settles at nothing, pays zero, and the player correctly concludes the factory is useless, when the actual failure is a 30-tile gap in the rail geometry that costs one more relay station to close. The commit decision has to include the relay station cost, because a factory cluster that needs two relay stations to reach its destination is a 375,000 gold commitment, not a 125,000 gold one, and the breakeven math changes accordingly. The train network guide covers the full relay chain and hub design; this section only needs the cap and the cooldown to frame the commit cost. The 15-tile minimum link is the other half of the geometry that the commit has to respect, and it is the reason a cluster of three factories 200 tiles from the destination is not a 200-tile problem but a chain of links each between 15 and 155.56 tiles. The minimum link is a floor, not a target: a 15-tile link is valid but expensive in build cost per tile covered, and a 155-tile link is valid and cheap per tile, so the optimal relay chain is the one that uses the longest links the station search can bridge without exceeding the 155.56 diagonal cap. A player who builds short 30-tile links all the way to the destination is paying the relay-station cost five times as often as one who builds 150-tile links, and that difference is a 625,000 gold commitment versus a 125,000 one for the same distance. The commit decision has to read the geometry at the link level, not the cluster level, because the cluster is the thing you see and the links are the thing you pay for.

Breakeven math: when the cluster stops losing gold

The breakeven question is the one that actually decides the commit, and it has three parts: the up-front cost, the per-stop payout, and the payback window. The up-front cost of a single factory cluster that reaches its destination with one relay station is 250,000 gold, the 125,000 first Factory plus the 125,000 first relay station on the shared ladder, and the cluster’s per-stop payout at a team-controlled destination with a clean nine-city-or-fewer route is 35,000 gold. The payback window is the number of paying stops the cluster needs to recover the 250,000, which is 250,000 divided by 35,000, which is about 7.1 stops, so the cluster is in the black after the eighth paying stop. That is the clean number: a connected, well-routed, team-destination factory cluster recovers its cost in about eight paying stops, and every stop after that is margin.

The number gets ugly fast when the route decays or the destination is self-controlled. A self-controlled destination pays 10,000 per stop, so the same 250,000 cluster needs 25 stops to break even, and a route that visits 12 cities before the destination pays 35,000 minus 3 times 5,000, which is 20,000, so the cluster needs 12.5 stops, 13 in practice. The commit is worth making when the clean number, about eight paying stops at the team rate, is shorter than the time the game will actually run, and not worth making when the decayed number, 13 to 25 stops, is longer than the window you have left. The Port versus Factory guide works the same breakeven from the Port side, and the two guides together give you the full comparison: the Port breaks even on its scheduled trade income with no rail dependency, and the factory cluster breaks even on its paying stops with a rail and a destination dependency, and the commit is the moment the factory’s dependency is cheaper to satisfy than the Port’s schedule is to wait for. The breakeven number has a fourth input that the simple division hides, which is the time value of the gold you have to hold while the cluster pays back. The 250,000 up-front cost is not a single tick of gold; it is 250,000 gold tied up for the 8-stop payback window, and during that window the same 250,000 could have bought a Port that is already paying on its own schedule, or an army level that is already winning the objective that decides the game. The honest breakeven is not just “8 paying stops” but “8 paying stops before the game ends and before the tied-up gold has a higher-value use.” A player who commits at the 20-minute mark of a 30-minute game has 8 stops of payback window, which is tight; a player who commits at the 10-minute mark has the same 8 stops plus a full 10 minutes of margin, which is comfortable. The breakeven math has to be read against the clock, not just against the stops, because the stops are the cost and the clock is the window in which the cost is recoverable.

Failure modes and counters: where the commit goes wrong

The commit fails in a small number of predictable ways, and each has a counter. The first failure is the disconnected factory, where the build finishes with no rail or no destination inside 110 tiles, and the counter is to never build the Factory before the rail segment that reaches the destination is funded, because the rail is the thing that converts the 125,000 gold sink into a 35,000 per stop source. The second failure is the long-route decay, where the route visits 11 or more cities before the destination and the per-stop payout drops from 35,000 to 25,000 or lower, and the counter is to reroute through a shorter chain of cities or accept the lower payout and scale the factory count to compensate for the lower per-stop rate. The third failure is the self-destination trap, where the only controlled destination is your own City at the 10,000 rate, and the counter is to either commit to the allied or team destination at the 25,000 or 35,000 rate, which requires a team coordination decision, or to stay port-only because the 10,000 rate makes the factory breakeven 25 stops, which is rarely worth the rail investment.

The fourth failure is the scaling past ten with no concurrency gain, where the player adds factories from ten to twenty expecting a 100 percent throughput gain and gets 33 percent instead, because the spawn odds halve while the train count doubles, and the counter is to read the spawn rate as 1 over (factories plus 10) times 15 and the throughput as train count times per-stop payout, not spawn rate times payout, so you know the honest number before you spend the second 125,000 gold. The fifth failure is the relay-station surprise, where the geometry needs two relay stations and the real up-front cost is 375,000 instead of 125,000, and the counter is to count the relay stations before the commit, because the breakeven math changes from 8 stops at 250,000 to 11 stops at 375,000, and that gap is the difference between a commit that wins and a commit that drains your economy for the whole game. Every one of these failures is visible in the game state before you spend the gold, and the commit is the moment you have checked all five. The sixth and most subtle failure is the destination rate mismatch, where the player assumes the 35,000 team rate but the destination is actually a 25,000 allied hub or a 10,000 self-controlled City, and the counter is to read the destination’s control state and team affiliation before the commit, because the rate is set by who controls the destination, not by who you want it to be. A player who commits at the 35,000 assumption and discovers the destination is only 25,000 has a breakeven that stretches from 8 to 10 stops, which is a 25 percent longer payback window, and a player who commits at the 25,000 assumption and discovers the destination is only 10,000 has a breakeven that stretches from 8 to 25 stops, which is a 212 percent longer payback window and almost never worth the rail investment. The rate is the single most important number in the whole commit, and it is the one most often assumed instead of checked, because the control state of the destination is visible in the game state and the rate follows from it, but the assumption is made before the check.

Mode and map adjustments: where the commit is right and where it is not

The commit’s value changes a lot by mode and by map, and the wrong call in the wrong context is worse than no call at all. In a team game with a shared team destination, the 35,000 per stop rate makes the factory commit the default: the team destination is controlled, the route is usually short because the team hub is near the contested front, and the 8-stop breakeven is short enough that the commit wins the economy war by the time the third objective starts. In a free-for-all with no team destination, the rate drops to 25,000 at a controlled allied destination or 10,000 at your own City, and the breakeven stretches to 10 or 25 stops, which often makes the Port the safer purchase because the Port’s trade income does not depend on a rail loop reaching a specific destination. The map matters just as much: a map with dense City placement near the front line makes the route short and the decay small, while a map with sparse Cities and long rail lines pushes the route past nine cities and the payout down to 20,000 or 15,000 per stop, which changes the breakeven from 8 stops to 15 or 20.

The practical adjustment is to read the map’s City density before the commit, not after. If the front line has three or more Cities within 90 tiles of each other, the route is short, the decay is small, and the factory commit is the play. If the front line has one City per 150 tiles of rail, the route is long, the decay is heavy, and the Port is the play. The mode adjustment is the same: in a team game with a hub, commit to the factory; in a free-for-all with no hub, stay port-only unless you can seize a City at the 35,000 or 25,000 rate. The map strategy guide covers the City density read in more depth, and the team economy space guide covers the team-destination coordination that makes the 35,000 rate available in the first place. The commit is not a single number, it is a function of the mode, the map, and the destination rate, and the right call is the one that matches the context you are actually in. One more context that changes the call is the opponent’s own factory commit. If the enemy has already committed to a factory cluster and is winning the economy war, the defensive read is to contest the destination they are settling at, because a train that settles at a lost City pays nothing and the enemy’s entire cluster becomes a 125,000 gold build with no income; that is a cheaper way to neutralize their economy than matching their factory count, which would put you in the 33 percent throughput race they have already started. If the enemy is port-only, the offensive read is to commit to the factory yourself, because the port-only economy is the one that does not have a rail loop to protect, and the contest over the destination is the thing that decides the whole game. The commit is therefore also a read of what the opponent has committed to, and the context of the whole board, not just your own factory count and your own route geometry.

Build checklist and what to watch in the game state

The commit, when all the thresholds are green, follows a specific build order and a specific watch list. The build order is: first, fund the rail segment that reaches the destination, including every relay station the geometry needs, because the rail is the thing that makes the factory earn; second, fund the first Factory at 125,000 on the shared ladder, which builds in 20 ticks; third, observe the first spawned train and watch it complete a paying stop at the destination, which is the moment the 125,000 sink becomes a 35,000 per stop source; fourth, fund the second Factory only when a completed paying stop, not a spawned train, is the bottleneck, which is the moment the first cluster’s throughput is the thing you are short of. The watch list is equally specific: watch the route’s city count to keep the per-stop payout at 35,000 and not the decayed rate, watch the rail segment lengths to keep every link under 155.56 tiles, watch the destination control to make sure the City or Port the train settles at is still controlled and at the rate you are paying for, and watch the spawn odds as you add factories to make sure the 1 over (factories plus 10) times 15 curve is the number you expect.

The exit signals are the mirror image of the commit signals, and they tell you when to stop scaling. Stop adding factories when the spawn odds drop below the point where the train count you can run is no longer the bottleneck, which is the moment the per-stop payout is the thing that limits you, not the throughput; stop scaling the cluster when the next 125,000 gold would buy a relay station that extends the route past nine cities and triggers the decay, because the 5,000 per extra city costs more than the additional throughput the relay station adds; and stop the whole commit when the destination control flips, because a train that settles at a lost City pays nothing and the entire cluster becomes a 125,000 gold build with no income. The commit is a bet with a clean entry, a clean watch list, and a clean exit, and the numbers on this page are the ones that tell you which of the three you are in. The in-game gold rates guide covers the broader gold budget that the factory commit has to fit into, and the train network guide covers the relay chain and hub design that the build order depends on.