You should invest as much as possible in scavenging (proof)

[Clavious]

I've recently started to pay attention to scavenging. And what has surprised me is that the time it takes to scavenge is concave in carrying capacity. As unit costs are linear in resources and recruitment time, and scavenging is linear in haul, the benefits of scavenging is convex. At the same time, building time of resource buildings is convex, and resources for building relative to the increase in production are approximately linear (or at least they seem just by looking at the building tables).

These facts make unit production for scavenging more effective in terms of resource yield than upgrading resource buildings.

Please see attachment for documentation and assumptions.

 

[--G--]

In other news, the sun is shining

 

[Hektik]

 

[DeletedUser]

"If you .... neglect unit training and building upgrade time"

What if you don't? I'm skeptical that "You should invest as much as possible into scavenging."

You can't scavenge with units that you don't have, and you can't make units instantly. Probably there's some points where it's more effective to buy mine upgrades instead of extending a recruitment queue.

The cost of farm upgrades to support the units seems to be neglected too.