Harbours everywhere?

[Krill]

Blake and DeepO did a lot of testing on this just after CIV came out in 2005. Someone who really was dedicated could goand look in the archives for the thread...but it might have been changed due to the change in mercantilism.

[Blake]

It still works the same way.

All you need to understand is that harbors are magnetic to trade routes, build a harbor in a city, and that city tends to pull in the best trade routes. You want the best trade routes going to the best cities, so build harbors in the best cities.

Generally, you get about the same amount of benefit, from building harbors in the best cities (until you run out of foirgen trade routes), as you do from building harbors in ALL cities. Obviously, it makes sense to only build the harbors in the best cities and save a load of hammers!

Generally speaking, if you need the health from the harbor, you need the health. So the "trade route shuffling" is not an overriding concern, the health is an overriding concern! But if you don't need the health then building harbors in the best cities only is a nice "cost-free" optimization.

[snoopy369]

Originally posted by wodan11
Question: when, exactly, does the game decide to switch routes?

It must be post-modifiers, or else the Harbor wouldn't make a whit of difference.

So, the 2nd city (the 1, in this example), would be compared to... let's pick a number out of a hat, and say +100% modifier... 4 + 100% = 8.

The 2nd city now has a harbor (+50%). The 1 becomes 1.5. (I believe with BtS the rounding is now done after all commerce is accumulated, instead of along the way.)

1.5 !> 8, therefore, the route would not be switched.

Wodan

The game determines post-modifiers indeed. When I say modifiers I mean harbor/etc; the 4 is post-population modifiers (as the value has no meaning outside of this - any value other than 1 has some modifier applied to it). There are two values - the raw value of that particular trade route, which is determined by civ, distance, pop size; and the modified one.

For example, you have trade route 1, which for city A is worth 4 gp ('unmodified' ie no harbor). City B values that for 3gp. City B, however, has a harbor, which gives it a 4.5gp value; so it takes it, replacing a 1gp value trade route, and city A gains that 1gp value trade route.

(Aside: it is my understanding that harbors give an incentive even on top of the .5* value, so that they can sometimes cause a literal drop in value.)

Now, you'd say, "City B has 4.5 and City A 1.0, so 5.5>5.0". My arguments to this are two:
1. 0.5gpt is not enough to pay the 100+ hammers for the harbor. That is 200 turns if you value 1gp=1hammer, and ignore the discount you should be giving because hammers now > hammers later.
2. City A is bigger, and probably older, and potentially has more multipliers - libraries, marketplaces, banks, etc. These do not affect the trade route calculation, but they certainly affect the net value. If City A has a +75% average bonus, and City B has a +25% average bonus, City A had 4.0 +3.25 = 7.25 net gp+beakers. City B with 4.5 now has 4.5+1.125=5.725 net gp+beakers. You just lost 1.5gpt. (Aside 2: I think the Bureaucracy bonus is also post-trade route calc; so your capital could easily be screwed out of commerce this way if it's landlocked.)

The moral of the story:
* Don't build a harbor in a small city; it's not going to help enough to pay for itself.
* Don't build a harbor before you build other gp/beaker bonus drivers (library/market/etc.)
* Don't build a harbor when your major cities are landlocked, particularly your capital under bureaucracy, unless your WS and Oxford cities are coastal.

[Krill]

[q=Snoopy]Don't build a harbor before you build other gp/beaker bonus drivers (library/market/etc.)[/q]

Exception; it is production city and you need the health.

[wodan11]

Thanks for the analysis, Snoopy.

BTW, your #1 is moot. Unless needed for health, the harbor is a dumb build in the first place. The argument that it is a waste of hammers is self-fulfilling.

Also, I think the moral of the story is that the game's algorithm to calculate routes needs improvement. If it's going to do it for me, it should do it right.

Wodan

[Jaybe]

Originally posted by wodan11
Also, I think the moral of the story is that the game's algorithm to calculate routes needs improvement. If it's going to do it for me, it should do it right.

Do it as documented, or
Document it as it is done?