Factorio: Engineering Addiction#

Transport options#

Walking is the baseline every other option is measured against — each vehicle buys speed, cargo space, or both, at a different cost.

• the car trades a fuel slot for speed well above walking pace, with a small cargo hold — cheap and early, favoured for outrunning trouble
• the tank is slower and less maneuverable than walking, but carries a submachine gun, a flamethrower and a cannon — durability over mobility
• the train (next week) trades route flexibility for speed and cargo capacity over long, fixed distances
• the spidertron takes no fuel — its speed comes from exoskeleton equipment in its own grid, and it crosses terrain wheels can’t

Obstacle removal#

Cliffs and water aren’t obstacles to route around forever — cliff explosives and landfill remove them permanently, at a deliberate resource cost.

• cliff explosives clear a cliff face outright, opening a route or building site that was previously blocked
• landfill converts water tiles to buildable land, at the cost of the landfill material itself
• both are one-way: once placed, the terrain change is permanent, which makes it a genuine design decision, not a temporary fix

Biters, worms and nests#

Enemy presence grows along an evolution factor from 0 to 1, driven by three independent things: time passing, cumulative pollution produced, and nests destroyed. All of it is driven by total pollution produced, not by how much of the cloud you’ve managed to contain.

• each source has a fixed weight per unit: time adds 0.000004/second, destroying a spawner adds 0.002 (same for every spawner type — not worms), pollution adds 0.0000009 per unit produced
• the three weighted totals are summed, then squashed into the 0-1 range by evolution_factor = total / (1 + total) — this is why evolution climbs fast early and crawls near 1.0, not a hand-tuned late-game curve

Evolution’s diminishing marginal effect#

• the same action’s marginal effect shrinks as evolution rises: any increase is multiplied by (1 - evolution_factor)², so destroying a nest at evolution 0 adds the full 0.002, but the same kill at evolution 0.5 adds only a quarter of that, 0.0005

Pollution’s second job, and nest expansion#

Evolution isn’t pollution’s only effect, and nests aren’t purely passive between attacks — both run on their own separate schedules.

• pollution has a second job: pollution reaching a nest triggers actual attacks, on its own schedule — containing the cloud reduces attack frequency even though it does nothing for evolution already accrued
• nests also expand on their own timer, claiming territory outward until they reach an occupied area — left alone, expansion fills every explored area

Evolution weights, in practice#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

Three independent tracks feeding one shared total — raising evolution from any one of them raises it for all of them, since the game only ever sees the combined number.

Biters, worms and nests, in practice#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

The curve flattens, it never resets — pollution already produced doesn’t un-produce itself, which is why the practical lever is producing less of it going forward, not cleaning up what’s already out there.

Reaching the same evolution factor, three ways#

Any of the three sources alone can drive evolution to the same value — the numbers below are what each one takes on its own, holding the other two at zero.

evolution   pollution alone   time alone   destroyed spawners alone
   10%          123k             7.7 hrs           56
   50%         1.111M           69.4 hrs           500
   90%           10M           625.0 hrs          4,500

Reading the three-ways table#

• reaching 50% evolution from time alone takes under three days of a world simply existing, running or not — the clock never stops
• reaching it from spawners alone means 500 kills, at a shrinking marginal gain per kill the closer you get — the last hundred cost far more than the first hundred

Evolution, in practice — by time#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

Same flattening shape as the pollution-only curve — every source feeds the same squashing formula.

Evolution, in practice — by spawners#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

Every source produces the same diminishing-returns curve — this graph and the time-based one on the previous slide are shaped identically on purpose.

Nest expansion#

Nests don’t just sit still producing pollution — they claim new territory on their own schedule.

• roughly every 60 to 4 minutes (the interval shrinks as evolution rises), a nest sends 5-20 units to found a new nest — the target chunk is picked within a 7-chunk radius by a score that only counts how many player structures and enemy spawners are nearby, never what those structures cost or how big they are
• that scoring is exactly why a single pipe run through open ground is an unusually cheap deterrent — it counts the same as a wall for suppressing expansion into that chunk, at a fraction of the resource cost

Nest attacks#

Attacks run on a second, separate schedule, triggered by pollution rather than by a timer alone.

• once pollution reaches a nest, it musters a group and launches an attack roughly every 1 to 10 minutes
• if some of the mustered units haven’t arrived at the rendezvous point yet, the attack waits up to an additional 2 minutes for stragglers before going anyway

Biter pathing#

Biters heading to attack a target take the shortest path that accounts for terrain, but not for player-built obstacles like walls — what happens next depends on how much detour those obstacles cost.

• if there’s a clear route around an obstacle that isn’t much longer, the group takes it and never touches the obstacle at all
• if there’s no clear route around, or going around would mean a large detour, the group attacks whatever is blocking the direct path instead

The maze tactic, and breaking off early#

• this is exploitable on purpose: a maze of walls with no clean way around forces an attacking group through a chosen corridor, tower-defense style — the tactic only works because the detour-cost rule is real, not because biters are simply attracted to walls
• a biter that comes within range of a military structure or unit while travelling breaks off early and attacks that instead, using the same shortest-path logic against the new, closer target

Turret creep#

Turret creep is an offensive tactic: advance on a nest by placing turrets just ahead of the last ones, retreating after each placement, until the turrets are close enough to kill it.

• a gun turret’s 18-tile range already beats every biter’s melee reach and a behemoth spitter’s 16-tile spit — a turret line wins that engagement once placed; creep exists only because turrets must be built within the player’s own reach, not because they’re outranged

Turret creep: the worm exception#

• worms are static, don’t chase, but outrange the turret that’s meant to counter them — small worms reach 25 tiles, medium 30, big 38, behemoth 48, all well past 18. A worm gets to shoot first and keep shooting throughout the whole creep-in; the turret doesn’t return fire until it’s already inside the worm’s range, not the other way round

Turret creep: damage math#

• piercing rounds do 8 damage/shot at 10 shots/second — 80 DPS per turret, enough to drop a freshly-spawned 350 HP spawner in under five seconds. The same math against a max-evolution 3,500 HP spawner takes ten times as long, which is the real reason creep stalls once behemoths show up: the spawner’s health scaled with evolution, the turret’s damage didn’t
• the placement limit forces a run-in, place, retreat cycle — turrets can only be placed within reach, so the line “creeps” forward rather than being planned from range

Turret creep, in practice#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

This is the “before”: no wall, no fixed line, just a nest cluster and a player about to place the first turret within reach of it — everything that follows is run-in, place, retreat, repeated.

Worms#

Worms are static — never chasing, expanding, or leaving their spawn tile — but everything else about them is built to punish getting close.

• worm tier is gated by evolution during expansion (map-generated worms ignore this): medium needs evolution ≥0.3, big ≥0.5, behemoth ≥0.9 — the further from the starting area, the stronger the worms map generation places regardless of current evolution
• range climbs sharply by tier: small 25 tiles, medium 30, big 38, behemoth 48 — every tier already clears the gun turret’s 18-tile range
• big and behemoth worms resist up to 70% of fire damage, and behemoth worms resist 80% of laser damage — a flamethrower or laser line built for biters and spitters underperforms badly against a worm nest by comparison

Worms, in practice#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

It never advances and never retreats — every worm engagement is decided by whether you can close 25-48 tiles of open range before it decides you’re in range of it.

Firepower and kiting#

• walls only block melee contact, not projectile or spit damage — big and behemoth biters reach 1.5 tiles regardless of a single wall layer, and a spitter’s 13-16 tile spit range makes a wall irrelevant to it entirely
• a spitter’s acid puddle outlasts the hit that made it: every tier’s puddle sits for 32 seconds, doing 7.2 acid/second for the small tier up to 360/second for the behemoth tier, and slows anything standing in it by 30-60% for 2 seconds per contact — the puddle, not just the projectile, is what makes standing still in a spitter fight expensive

Kiting and laser upkeep#

• kiting uses a vehicle’s speed to stay outside an enemy’s attack range while still doing damage — useful against spitters and worms, whose ranged damage otherwise ignores your speed
• laser turrets trade ammo logistics for power draw: no supply chain to feed, but 1.2 MW while firing against 24 kW idle means a laser wall needs power headroom sized for a real attack, not just its average draw

Firepower and kiting, in practice#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

This is what a finished, static line looks like mid-attack — the wall stops nothing on its own, every bit of stopping power is coming from the turrets behind it.

Static defences and flamethrowers#

A defensive line meant to sit in one place and hold indefinitely has a different job than turret creep’s advance — and a different weapon choice follows from it.

• flamethrower turrets deal damage in a spreading area rather than to one target at a time, so a mixed wave of small, medium and big biters dies faster per shot than the same wave against single-target gun or laser fire
• their fire also lingers on the ground briefly, keeping damage applied to anything still standing in it — a static line gets to spend that DPS on crowd density it doesn’t have to aim

Flamethrowers: trade-offs#

• the trade-off is fuel logistics and range: flamethrower turrets need a piped fuel supply and have a minimum range, so they pair with — not replace — gun or laser turrets covering what’s too close to burn safely
• fire is a poor answer to worms specifically: big and behemoth worms resist up to 70% of fire damage (see the Worms slide), so a flamethrower line sized for a biter wave burns through nowhere near as fast against a nest defended by higher-tier worms

Static defences, in practice#

Source: Factorio Wiki, CC BY-NC-SA 3.0.

The fire keeps burning ground it already touched — that lingering area damage, not raw per-shot power, is what a static line is actually buying with the fuel cost.

Other tools#

Not every threat is a straight firepower fight — poison capsules and combat robots solve problems that turrets and guns don’t.

• poison capsules deal heavy damage to worms specifically — spawners aren’t affected by poison damage at all, so this is a worm tool, not a nest-clearing one — with far more range than a grenade, the direct answer to a worm a player can’t safely walk up to
• combat robots (defender, distractor and destroyer, in rising cost order) follow the player or hold position and fight automatically, useful for soaking or distributing damage during an attack run on a nest rather than tanking every hit personally
• neither replaces firepower — both exist to change where the damage lands, buying survivability during an approach a straight gunfight wouldn’t allow

This week’s practical#

Goal: destroy a biter nest with this week's tools, opening up access
Test: use at least one non-turret tool (vehicle, capsule, or terrain change)
Pass: the targeted nest is destroyed, not just suppressed; pollution and
      evolution state at the time of the attack is noted

Suppressing a nest behind a wall isn’t the goal here — the nest has to actually come down, and by a route that uses more than gun turrets alone.