#long lines of stations maximise profit from trains (i think?)

1 messages · Page 1 of 1 (latest)

proper helm
#

A basic question: how to get the most out of our train economy?

Today I looked at the train mechanic: ~~https://openfront.miraheze.org/wiki/Train~~ [correction:] https://openfront.wiki/Factory. If we have a large number of cities and factories to place, then I think the optimal placement is to have all our cities in one long line, as pictured.

Why? By using long lines and avoiding branching, each train that spawns is forced to pass through a large number of stations (cities and ports) in order to get to its destination - we don't get this benefit if a cluster of stations has lots of interconnectivity [edit:] *or if we have many disconnected clusters *.

Note: obviously it's better if that long line allows us to trade with allies rather than just ourselves, [edit:] *also, adding more factories improves our profit sublinearly in the number of factories, whereas adding more stations (cities or ports) to our long line improves our profit linearly. So in an idealised scenario where we have plenty of space and can afford to place a very large number of buildings, adding more stations to our long line is better than adding factories.

Can anyone tell me whether my analysis is correct?

In light of the conversation below, I have made edits to this original post!

image.png
umbral wharf
#

it's most likely not

#

the miraheze one is even less updated

#

and you should also know that trains lower in value after 5th stop

proper helm
umbral wharf
#

#💭︱general message

#

there's that too

#

#💭︱general message

#

and that

proper helm
#

Thanks. I've thought about the maths again and I think that under the correct mechanic that you point me too, the conclusion is the same

#

If you have a large number of cities to place, the average train will go past many more stations if you place all the cities on one line, which is better than having the average train stop a fewer stations, (even after considering the reduced advantage from the fact that you get 5k instead of 10k or 20k instead of 25k)

umbral wharf
#

you're analysis isn't fully wrong, but more factories is good because it means more trains

#

and stacking up to 10 factories I guess

#

and I'd say you gotta make sure the whole railway is connected too, right ?

proper helm
proper helm
umbral wharf
#

so a line (snake) would be more profitable than a loop (ouroboros) ?

proper helm
#

oh uh I hadn't thought about loops - I have been assuming the train has a predetermined destination and despawns once it reaches its destination(?); then adding loops would allow the train to reach its destination using fewer stations, so loops make things worse compared to one straight line.

umbral wharf
#

I'm guessing that without destination, loops would be op

#

and therefore, trains probably have a destination

#

or a designated amount of stops before disappearing

brave torrent
umbral wharf
umbral wharf
brave torrent
brave torrent
umbral wharf
#

Aw, I hate when things are hidden from the changelog :(

twin hill
# umbral wharf so a line (snake) would be more profitable than a loop (ouroboros) ?

I actually did a deep dive into the codebase a while ago on this exact point. Turns out single-line, non-looping train lines are in fact higher income-generating per unit time than loops, given the same number of stations and trains. This is because, mathematically speaking, the average path length is 1/3 in a line but 1/4 in a loop, so each train visits more stations per trip. Remember that trains despawn once they reach their destination, so categorically the worst setup is a hub-and-spoke approach, since you would be minimizing the number of stations trains pass through. As for stations (i.e., cities / ports) vs. factories (train spawners): Your marginal return on the number of factories starts to plateau once you reach 10 factories in the same cluster, whereas stations continue to scale linearly without plateauing. (By the way, train factories only affect the rate of train spawn, and the spawn rate does not scale with how tightly packed the stations are.) All that said, loops do have an advantage over lines in that they're more robust to destruction: you have to bomb them twice to break the train route. But if robustness is your concern, a network is of course even more robust than a loop.

spark geyser
#

I've always just done a network rather than a line or a loop. Sometimes I'll do a line to connect to a distant neighbour, but over time I try to expand it into a full network.

proper helm
heady vale
#

but what i wanna know is that when you got your effective line or lines of cities should you start placing cities in other places making less effective routes or should you just start stacking cities to not mess up the efficient lines but if you stack cities your not adding more trainstations and more trainstations should be better because more trains will choose your stations as destination

cursive jewel
#

so should factories overlap eachother?

#

like should my factory radius ninclude the other factory?

umbral wharf
#

not needed

cursive jewel
#

is it useless to have a city hit by two factories then?

umbral wharf
cursive jewel
#

ok thank you!

craggy stratus
#

wouldn't 1 factory and like 12 cities/ports in a straight line next to it do the same effect?

#

just as efficient

#

u just stack the factories

#

and u could also build sams to defend the whole line as its more compact

umbral wharf
#

Except than stacking factories is inefficient

craggy stratus
#

why

umbral wharf
# craggy stratus why

Because a level 60 factory will produce less trains than 60 separate factories, I think the most profitable you can get is like level 10 or something

craggy stratus
#

never heard about that

umbral wharf
craggy stratus