Forums

I’d agree with you, if the maps were large enough to encompass multiple regions (such as if there was a “Greece” or “Central Europe” map). However, based off of the size of the maps, it’s not much different than the trader walking from one town to another in a localized region.

lol, this video makes you realize what a vast improvement this is

Merchants would mostly keep to certain regions. Hardly anyone in the Mediterranean tried to travel to the source of tin for example. I think it's fine that a given Trader bounces between two Markets (representative of a region, not two literal markets, of course they went around a few). I think a problem is that the game makes the most profitable trading route the only one worth having, making things quite dull. To solve this, I think traders per route need to be limited (have slots like resources, varying with distance), and the gain be such that multiple Markets at both ends is not the most profitable configuration (that's why a per Market limit instead would be easily exploitable). Looking at the code, the trade gain goes with distance squared, with a damping for long distances (dividing by 1.25 at map diameter, I think). Ignoring the damping for now, the question would be: given a bunch of Markets with all routes full, where should a new Market be placed for maximum profit? I'll assume Trader slots will vary linearly with route length, to account for time needed to travel them (fixing Traders arriving to Markets per unit time). This problem is then a generalised Leibniz geometrical theorem (https://arxiv.org/pdf/0710.0342), since one wants to add up the square of the distances to all Markets. The solution is that profit is greatest the farthest away from the centroid of all original Markets (I don't think damping changes this result much, and not in an obvious way anyway). This is not nice, since Markets would need to be in a periphery, as expanded as possible (a circle empty inside would be ideal for a given centroid distance, obviously), instead of forming a more interesting network. I don't have a proposal for now, but I think this issue is worth exploring to reach a solution that is simple, intuitive, interesting and not exploitable.

Indeed, a lower threshold would be like trampling, a higher one charging/ramming. That's quite nice and flexible!

The damage is scaled with relative speed. And it has a threshold* you can set in the template (could be 0 too). It's called charge in the code but trampling could be somewhat descriptive too since it's any collision that would trigger the effects. *the threshold is only checking the charger own speed, so 0 for elephants makes the unit coming into contact to also count as hit. So it is somewhat a bit trampling.