On the second, it often comes in the form of npc price fluctuations
There are no npcs buying or selling anything.
The rest of it -- dynamic algorithms and changing some values after playtesting, sure. I'm just trying to get a general gross idea of feasibility before coding major things and making major choices. Like "Taxes vs. decaying machines" for instance, are fundamentally completely different solutions. One is embracing accumulation of junk, and just bringing newbies up to speed with it. The other is fighting accumulation of junk.
I need a basic concepts nailed down. I'm not talking about the level of how many units of this do blah blah yet, definitely not, without a working prototype.
Edit: New concept. Security-based limitationsFollowing the idea of "automatons as sheriffs," let's say the game has no built in hardcoded anything stopping you from stealing, for example. As in it won't physically prevent you in the form of a popup message that says "action canceled, owned by Joe" or whatever. Instead, if automatons catch you, they all turn on you and hunt you down or exact some other punishment.
This can limit amount of stuff people accumulate. You only get one automaton, and although you can build contraptions to extend its eyes and ears while you are offline (intruder alarm shindigs), these won't stretch for miles very easily. You can therefore only really secure one area while you're not there. Anywhere else is relegated to cruder things like booby traps probably, unless you can afford umpteen mile networks of telegraph cables etc. for your automaton to be satisfied with enough evidence to "convict" somebody from far away.... which you can't.
So if you try to run multiple sites all over, you're opening yourself to Robin Hoods. UNLESS perhaps you get some other players to agree to lend their automatons to guard other sites. But they're going to want you to pay them for that. Which also redistributes wealth, so fine!