How do biter reinforcements work?

[ehs]

In my current game, a ribbon deathworld currently at blue tech, wading through enemy nests is becoming quite the slog. The same map structure that makes it easier to defend my base makes it easier for the biters to defend theirs. The biggest problem has been getting close enough to reliably hit the nests through the biters, or to poison worms. Reinforcements spawn in huge numbers, the map shape makes circling around them impossible, and it's easy to get stuck if I drive close enough to ensure a hit. Even a rocket-fueled tank can only plow through so many biters before grinding to a halt.

While wading through one wave of said reinforcements, I realized I don't know how they actually work and the mechanics might be tactically relevant. How fast does a nest spawn biters, and under what conditions? Do biters spawn without limit during combat, or do they only respawn on death? Will nests "respawn" biters that wander too far away, even if they're not dead?

If new biters don't spawn until the first ones die, then perhaps I can bait most of them away, hit them with slow grenades, then leave them alive and out of the way while I go clear the worms and nests.

[mmmPI]

The main cause of spawn is pollution which i suppose is "concentrated" in ribbon world making each nest spawn faster. ( plus it's deathworld ).

There is a lot of explanations related to the evolution factor and the spawning on the wiki https://wiki.factorio.com/Enemies

You can try to lure them away from their nest but some will continue spawning in the meantime.

You can "counter attack" just after they sent a wave to crash onto your static defense, that's when they'd be the less numerous but that sound risky for me, i usually use turret creep when i struggle.

[Tertius]

mmmPI, you're wrong. According to the wiki, this is (part of) the biter mechanics:

• nests spawn biters and spitters freely. Source
• pollution is used for collecting and sending existing biters and spitters to attack. Not to spawn them. Source

From my observation attacking biters near nests, killed biters are replaced infinitely by newly spawned biters 1-5 seconds after they have been killed, without ever exhausting the invisible biter source. Each nest seems to have some fixed number of biters it maintains. May be 4-10, may be dependent on the evolution, don't know exactly.

So this is my conclusion for strategy:

• Killing biters serves no purpose other than defending some currently attacked entity, because they are replaced immediately and infinitely
• To get rid of biters, you need to destroy their nests first, then kill the remaining biters

Since the respawn time is so fast, it's not a viable tactics to first kill biters to thin them out, then go for the nest. Instead, you need to not kill them but instead distract them to have them far from the nest battle. If you kill biters while attacking the nest, the newly spawned biters will attack you immediately, so they are not really thinned out and you just wasted ammo and time that could be used for attacking the nest.

Pollution is the cause for biters being sent to the source of the pollution and attack it. It's not the source of the biters spawning.

[mmmPI]

mmmPI, you're wrong. According to the wiki, this is (part of) the biter mechanics:
Pollution is the cause for biters being sent to the source of the pollution and attack it. It's not the source of the biters spawning.

I wasn't precise enough, you are correct, i didn't read "pollution" in the original post so i felt compelled to mention it, but it is not the direct cause for biter spawning in the world, it is responsible for attaching already spawned biters to an attack wave.

From my observation attacking biters near nests, killed biters are replaced infinitely by newly spawned biters 1-5 seconds after they have been killed, without ever exhausting the invisible biter source. Each nest seems to have some fixed number of biters it maintains. May be 4-10, may be dependent on the evolution, don't know exactly.

I have observed the same thing, if you play railworld and peaceful, that is about the number of biters that is attached to each camp, they never get send to an attack wave, and their number is not 0 nor growing unlimited. I can only suppose it's a minimum compared to deathworld.

But killing them one by one is an "endless" task, one has to go for the nests.

It is still possible to lure away some waves when doing so, using cliffs or forests, though maybe not in ribbon world, which would allow to "go for the nest" while their guards are stuck out of range, you should not need such things in late game though.

It is also still meaningfull to "counter attack" , or at least not attack into an incoming attack wave but just after, because those would be extra defense and make it harder to go for the nests.

[ehs]

I tried some tests -- these are on a deathworld with a current evolution factor of 0.8439:

• Single nests spawned up to 7 biters and stopped.
• Respawning all 7 after wiping them out took ~32s, or ~4.5s/spawn (eyeballed with a stopwatch, not very accurate).
• Leading biters several chunks away did not seem to trigger new spawns. But...
• I tried leading biters away from the main hives, locking them all down with slow-grenade spam, then raiding the nests. It didn't work as expected; I was always met by a new wave of spawns, though not usually as large.
• Also, radar shows far fewer biters-per-nest in the giant hives than an isolated nest.

After that I went digging in the game's data files and things started to make sense. I found these values in enemies.lua for "biter-spawner". Irrelevant keys omitted, comments in original:

Code: Select all

max_count_of_owned_units = 7,
max_friends_around_to_spawn = 5,
-- With zero evolution the spawn rate is 6 seconds, with max evolution it is 2.5 seconds
spawning_cooldown = {360, 150},
spawning_radius = 10,
spawning_spacing = 3,
autoplace = enemy_autoplace.enemy_spawner_autoplace(0),
call_for_help_radius = 50,

So yeah, spawn rate based on evolution with a per-nest cap. But the interesting part is max_friends_around_to_spawn; if that's what I think it is (a cap on total population in some small area around the hive, independent of the hive's individual cap), then it explains why the giant hives seem underpopulated relative to single nests (they hit the local-overpopulation cap before the per-nest spawn cap). It also explains why baiting the first wave of biters elsewhere doesn't work -- once they leave the hive to follow me, the corresponding nests are free to spawn more because they weren't at their individual caps to begin with.

So back to square one. I'm nearly out of my initial copper, with no alternate source, so I can't tech up to nukes or spiders; but a fair number of weapons don't seem to require much/any copper to sustain, so my run isn't dead yet. At the moment I'm mostly using explosive tank shells and poison grenades to clear land, with land mines and flamers to defend it. Behemoth worms and vast hives are making tank play increasingly difficult though. I wonder if I can turret creep with the flamers. Might be hard into b-worm clusters, given their extreme range. We'll see.

[Qon]

Land mines used in offense is the answer.

[mmmPI]

It also explains why baiting the first wave of biters elsewhere doesn't work -- once they leave the hive to follow me, the corresponding nests are free to spawn more because they weren't at their individual caps to begin with.

Thanks for sharing your discoveries !

For the quoted sentence i think the scale matters a lot, i was wondering how wide is your world ? I wonder how far can they follow you in tiles before they are far enough away from the nests that some new one will respawn.

Not in ribbon world, i have observed while fighting clusters of 10-50 nests that you can "aggro" the "guards" by circling around the cluster and having them form a moving pursuing mass, which you keep moving behind you while shooting at the nests, and once nests are dead, then you can kill the pursuing mass, but if you try the "inch by inch" approach, then it doesn't work because they keep respawning, that's what i meant by "you have to go for the nests", and also this isn't required with artillery, but that's yellow science/late game .

I had played some ribbon worlds, and i remember a vanilla one 32 tiles wide, where i did turret creep with laser and flammer, but it was not deathworld i don't think i had to face behemoth worm pre-artillery, i suppose if you stop your turret creep you can make sure there is no behemoth worm in range of shooting at your defense, and also not enough room for new nest that would spawn/settle not in the range of your defense.

For the B-worm cluster, i would use many legs and shield, and sortie a few time surrounded by defender robots, to throw some poison capsule at the worm, i think they need 3 or 4 consecutive this or just put more turret in the creep.

[ehs]

The ribbon width is whatever the default is, I think 125 tiles. I couldn't tell exactly how far the biters had to roam before new ones spawned. I was too busy not dying. My guess would be 10 tiles, from the spawning_radius in the data file.

Circling large hives is how I normally whittle them down in the mid-game, but it doesn't work here; most hives extend the full width of the ribbon. I am not sure I have enough copper to get the exoskeleton/shields/reactor trifecta going. I have ~20k still in the ground, but that has to last until I find a copper mine. My tentative plan is to try flamer creep and mines up to the bworm range limit, then see if I can safely venture just far enough ahead to poison them. No-man's land will probably end up perpetually on fire, but I have tons of oil, and the tank seems to endure fire fairly well.

[edit]: It occurs that this may be one of the few situations where discharge defense is worth it...it could keep the biters off me just long enough to poison worms and retreat. I'll have to try that.

[mmmPI]

The ribbon width is whatever the default is, I think 125 tiles. I couldn't tell exactly how far the biters had to roam before new ones spawned. I was too busy not dying. My guess would be 10 tiles, from the spawning_radius in the data file.

[edit]: It occurs that this may be one of the few situations where discharge defense is worth it...it could keep the biters off me just long enough to poison worms and retreat. I'll have to try that.

Totally understandable discharge defense are re-usable so it spare the copper, now to me the situation where i use discharge defense are very risky, if it saves me, against behemoth worms, it's because i got stuck in acid pool, and then by luck i happen to (have one installed and) use it fast enough, i would immediatly go out of danger and save the game lol, i could not rely on repeating such situations expecting favorable outcome repeatdly, so i would save and reload when killed :/

I checked and yep default is 128 so it's not wide enough so as to circle around, but it's still wide enough that you get massive waves coming from both sides when you attack in the middle. It may then be easier to progress on 1 side first so as to deal with less density and only from 1 side and then using flammers sideways. That would not be a possibility for narrower ribbon. The behemoth worm range is 48 tile, so attacking in the middle first is "worse" i suppose, but yeah the real deal is the lack of copper landmines do not cost any copper, neither does flamme turrets , but using both at the same time is dangerous, gotta pick one for offense. I would creep with flammer, endure the losses, clear some land say 300 tiles by pasting blueprint of a wall 1 or 2 tile further than previous , and then go back a little say 50 tiles to avoid stabilizing the wall in range of behemoth, this to gain "250 tiles" x 125 width. The distance to find copper is less than for narrow ribbon, but the amount of worm is the same i suppose, though you have more of them hitting you at the same time, artillery should really help hopefully finding 1 patch of copper will be enough.

Also there are 2 directions to search for right ? maybe one of them has near copper, and the one 10 time further away ? did you allow yourself to cheat and have a look ?