Current state of merchant agents?

[Mark_Everson]

Hi colorrr:

As you say the documentation is a bit out of date... this thread, and perhaps some mentions in the econ thread are the best indication of where things stand.

Originally posted by colorrr
The merchant operates from his province to other provinces, which could be the same civ or another civ. Another civ would give more potential income. Another race/religion/cultre could also give boosts.

For now profits are only by what the good is, and what the local economy wants to pay. There are no magical effects as in civ where trading with another civ gives better results. That said, other civs are more likely to have critically needed resources, and so give more profit. Also there will be trade effects on tech that Will be greater if trading with different civs and cultures.

When the merchant starts out he does not have much impact on the civ economy as his net worth is not that much. For every trade route, the total gain gets divided in percentage between various factors. That means some part of the gain goes to the merchant himself (raising his net worth and making him more powerful), some part go as taxes to augment the civ economy (variable by the player setting taxes), some part goes to hire guards, workers, some goes to corruption/black market (depending upon civ type/racial traits/social level etc), some to warehouse maintenance. When the merchant starts out he could only have one route, but as his net worth raises by the money generated per turn, he'll get a nice pool of savings. He can try to set up (controlled by AI) another route. Setting up a new route could also be a failure, making that merchant suffer a setback.

There is also no arbitrary bonus needed for numbers of trade routes since their Value is clear from the underlying detailed local economy. This is one of the advantages of the detailed economic simulation, very few arbitrary rules are needed. But as you say a merchant will pay taxes, pay for ship, wagons guards, and the lot.

The merchant could have reference to social class/race taking some of the traits of these into account in different calculus. F.instance a merchant from an aggresive race would be more prone to set up risky, long but profitable routes instead of shorter, safer but low value ones. Maybe some cultures/races trade better than others, some don't trade at all.

This is possible in the long run. But I want to get the basics going before we get too far into deep details like this.

I think the merchants should be abstract, not moving on the map, neither should it be possible to see the path. Each merchant should just list his trade routes in a simple from (with x goods) to ( with x goods).

Not moving on the map I agree with! But I think it should be possible to see the paths. I think some players would like to see, on a small map what their whole "trade web" looks like.

Thanks for your input!

[ogj20]

Could I suggest that merchants should calculate their trade routes using the pathing ideas discussed earlier, but that to avoid needing to store each of them, they could just increase the hidden 'trade' property of a square by an amount related to the value of their trade. This trade variable would decay over time, but merchants would update there paths every x turns or so, with x possibly related to some aspect of their social profile, like traditionalism versus innovation, so it would tend to settle to a particular level.

High values of 'trade' through a square should increase the probability of other merchants using that square as part of their trade route, a bit like roads would increase it by reducing the movement costs. This would give a tendancy for trade routes to clump together, creating major trading routes. This would model how merchants tended to try and follow established paths, since then they could either group together to reduce the dangers of banditry, or take advantage of the businesses that cropped up along trading routes to supply and support them, or even just use the fact that some other fool has gone that way before, so you know it can be done.

Which leads to another point, as trade flows through a square it should contribute to the creation of roads, even if only the laying down of very basic roads, like dirt paths, representing the creation of known routes to places, and the slow development of them as nire and more people use them. That would help reduce MM since merchants would automatically find suitable paths for roads, and start the process of building them, even if the player doesn't intervene.

[ogj20]

I forgot to say, that way you get to display the trading web to players, just by providing an overlay view that highlights squares with comparitively high trade levels passing through.

You could still store the routes of player controlled merchants explicitly, but is seems silly to store all the routes of all traders that way, since the player cannot directly influence most merchants individually, just by influencing the ease or difficulty of trading through certain squares, e.g by improving roads or by station troops in the way.

It would probably also be worth explicitly storing the supply routes of TF's so that the player could try and cut them, or at least force them to use unusual and costly routes.

But for the majority of background traders, knowing the aggregate level of trade would do just fine I think.

p.s. This got me thinking, is there an equivalent of building roads for water squares, e.g. the level of roads in a square presumably go something like

Terra incognita
no roads
dirt track
road
highway
rail
interstate/motorway/autobahn
something like 'mag tube' etc.

similarly water transport improvement could go:

Here be dragons
uncharted waters
charted water
navigation route
shipping lane
something like 'Soliton wave path'

That and the idea of trade being easier through high trade squares and trade slowly self-generating transport improvements would help to generate main trade routes across oceans were otherwise you would end up with maybe tens of equivalent cost paths from one side to the other.

[Mark_Everson]

Originally posted by ogj20
Could I suggest that merchants should calculate their trade routes using the pathing ideas discussed earlier, but that to avoid needing to store each of them, they could just increase the hidden 'trade' property of a square by an amount related to the value of their trade.

Hi ogj20:

Some interesting suggestions. I think the overall trade web indicator is a pretty good idea. Not so sure about trade flows generating roads. Usually its the other way around IMO. We certainly do need all we can manage to reduce MM. How about that a civ-wide option might be to spend x% of road-building funds where trade density is highest?