a review of the most underappreciated building of CIV: the harbor

[GeoModder]

This talk about avoiding mercantilism in certain circumstances, I suppose that means simply running the basic economic choice, for free trade is offen not available yet?

[Arrian]

Yeah, it would mean sticking with the basic economic choice (decentralism?) until you get to free trade - saves you the anarchy (if not SPI) and the extra upkeep (MERC is medium upkeep IIRC).

-Arrian

[GeoModder]

Only too bad that we can't see the result of civic changes BEFORE they are implemented (at least not as far as I know). SMAC's Social Engineering was very convenient in that.

And yeah, I know I can reload but that takes quite some time on a slow(er) computer.

[DeepO]

Originally posted by GeoModder
Only too bad that we can't see the result of civic changes BEFORE they are implemented (at least not as far as I know). SMAC's Social Engineering was very convenient in that.

It's true. In the ideal world, you would get a break down on civic changes: what does it do to civics, maintenance, and upkeep costs, and what does it do to trade income. Just the civics stil leaves a lot to intuition, or expertise.

I'll mention it again as a suggestion for XPs. In the mean time, try to look at F2 every time before and after you switch civics, and try to see the relative costs and income change. After a few games, you can 'feel' from your F2 what benefit a civic will give you.

And remember: a civic which will cost you happiness also has a commerce effect: unhappy people aren't working cottages.

DeepO

[Arrian]

I agree - a good thing to put on the list for the XP(s).

-Arrian

[Hermann the Lombard]

I'm bumping this because of the "Whats the deal with harbors?" thread, since this has been worked over thoroughly here.

Now I have a question: has the state of the (harbor) art changed since December? Any revelations (or patch changes) that influence the perceived value of trade routes?

[make]

Originally posted by DeepO

Krill, I don't know it either. The reason is twofold: there is only a limited amount of free time I'm willing to spend on re-engineering stuff, and you might not see it but I'm spending a lot of that on combat. Something's cooking Further, trades is one of those aspects of the game that interests me, but simply to have some high-level ideas on, and not the find the last mathematical detail involved. Similar to forest growth, or barb spawn chances I guess...

I do have some observations, though. Size is definately a factor for trades. Distance and/or foreign cities as well. And I believe harbors and possibly culture/markets/libs is as well...

I'm not sure, but I think that trade is made in pairs. Or at least, the available (foreign) trade routes are distributed over your empire. You simply get a list, and your best city gets the first choice, gaining you most on the trade route. Your second next, etc.

I'm not sure, but I think this 'list' is global: when your emprire consists of many small cities, you won't have trade towards foreign capitals anywhere in your empire. Those trade routes are reserved for AI (capital) cities, which will all be bigger and better than yours.

Also, and this is even more speculation, I think determining which city is 'best' is not based solely on pop, but in general on commerce income you will gain by the trade route. I've seen it happen quite often that a city which builds a harbor does not simply get 50% more income from the same trade routes. No, it gets the 50% bonus, but also gets different trade routes which offer more in itself. In an empire with 2 large cities, your second city coastal (capital in-land), building a harbor early on in your second city will give the best AI capital up for trade there. Add currency, and you will have 2 trade routes to capitals in that city. Your own capital gets minor AI cities... this indicates to me, that when matching up your cities with all available trade route, commerce modifiers (and thus harbors) are taken into account.

Mind you, that could also be just a coincidence of the world situation changing at the exact time of harbor completion. At any rate, this changing-the-trade-routes thing happens a lot more early on in the game, later on more cities will have harbors and thus you won't 'jump'. You will only get the 50% increase from a harbor, nothing more. Not that it's not enough, it's already very powerful

DeepO

I have notices somthing similar. What puzzles me is that sometimes, when completing a harbor, the commerce from traderoutes go up 100% (from 1 to 2) and sometimes 500% (from 1 to 5) or something inbetween. Sometimes the 50% is correct, but there no telling in advance what will happen.

It seems very random and impossible to predict what happense with the traderoutes after the harbor is build, and that makes it impossible to make the decission about wether og not to build a harbor instead of some other building.

As a general rule, I always build a harbor in a city wich have 1 or more traderoutes wich give +2 or more commerce. Because then its garenteed that someting good will happen.

I havent payed attention to if the city sometimes get different traderoutes after completing a harbor, but that could explain some of the big boost I get sometimes.

[couerdelion]

I believe that there has been a lot of doubt cast over the effectiveness of harbours. I was even quite surprised to read this thread and find so few negative remarks which I guess might be attributable more to the earlier thinking on the effects of harbours.

From what I read, if multiple cities can trade with Athens but only one of your cities can benefit from the trade with Athens one can assume that the ranking of all cities needs only to get in the top # of global cities to benefit from the trade with Athens (presuming Athens to be the best trade route city for all other civs).

I can presume that this trading algorithm only works on cross-border trade because I already know that a large number of cities can trade with my capital.

At the simplest level the number above may be that only one city can get trade income from Athens so it is a simple ranking question of all foreign cities to see who gets it. I can then presume that the model might find the for the next “most desirable” and perform another ranking. The model would be more complex if more than one foreign city can get trade income from Athens.

Building harbours should, of course, come with another caveat if your smaller cities are the coastal ones. Prior to a harbour, your bigger cities were getting more trade income and we know that some of the gain from the coastal city’s harbour will be reflected in lower trade income from cities which it has leap-frogged in the pecking order for trade. The problem then arises that those other cities are likely to be the ones with the science and gold multipliers. If your +6 gold in harbour city has reduced gold in other cities by 4, you will lose out if your “aggregate multiplier” in those cities is more than 150% of the multiplier for the harbour city.

I really haven’t got a precise understanding of the effect of harbours but noticed this problem in my last game – where some of my second and third rate cities were coastal and I also had build the great lighthouse. I was very loath to build the harbours in those cities until I had a few science/gold multipliers ready to benefit from the extra commerce which other cities would lose.

[Hermann the Lombard]

Originally posted by couerdelion
I believe that there has been a lot of doubt cast over the effectiveness of harbours. I was even quite surprised to read this thread and find so few negative remarks which I guess might be attributable more to the earlier thinking on the effects of harbours.

Thanks for your clarification of more recent thinking.

I think there should be SOME mystery in how the formulas work (in other words we shouldn't be able to calculate everything to 6 decimal places), but they should generally work in a roughly intuitive manner...in this case building a harbor in city "Z" should help "Z" but shouldn't hurt "J" and "L". It's likely that the observed effects reflect an Undocumented Feature rather than a deliberate design choice.

[beancounter]

Each individual civ can have one trade route per foreign city. In my current game, there are 5 foreign cities that have a trade route to my capital.

This means the trade bonus from a harbour is not really that good. It steals a better route from one of my own cities.

[couerdelion]

Originally posted by beancounter
Each individual civ can have one trade route per foreign city. In my current game, there are 5 foreign cities that have a trade route to my capital.

This means the trade bonus from a harbour is not really that good. It steals a better route from one of my own cities.

Just out of interest, how do you know that they are getting trade income from the trade routes to your capital?

[make]

Last night I payed more attention to harbors.

I can verifiy that sometimes after building a harbor, the city gets new traderoutes to other cities. Thats is what is confusing me. The 50% increase in traderoute yield is actually very consistent after a closer look. Its just that sometimes different traderoutes appear after the harbor is build, and that confuses me.

Where the new traderoute come frome im not sure about, but if its stolen from one of my own citys with better multipliers, its not good.

[beancounter]

Originally posted by couerdelion


Just out of interest, how do you know that they are getting trade income from the trade routes to your capital?

Easy. Open one of your save games and enter the world builder. Plant a spy in every single foreign city. Then close the world builder and check the other cities. You will see that each civ that has a open borders agreement with you will have a trade route to your biggest city.

I believe the formula for trade income is determined by city size, so the harbour probably increases the size of your city for all trade calculations.

[couerdelion]

Originally posted by make
Last night I payed more attention to harbors.

I can verifiy that sometimes after building a harbor, the city gets new traderoutes to other cities. Thats is what is confusing me. The 50% increase in traderoute yield is actually very consistent after a closer look. Its just that sometimes different traderoutes appear after the harbor is build, and that confuses me.

Where the new traderoute come frome im not sure about, but if its stolen from one of my own citys with better multipliers, its not good.

The earlier messages seemed to suggest that harbours in larger cities did little to increase their trade income. This would contradict the statement that a 50% increase in trade yield is consistent. Other responses are practically confirming that you are not stealing trade from the AIs. All in all, harbours seem to have a mystery surrounding them. My biggest concern will be where the primary cities are not coastal. Particularly when running bureaucracy, your capital’s commerce multiplier (science/gold) can typically be 200%+ so for each commerce lost by the capital you will need to gain 2g in a city with no multipliers.

Since this period of the game will typically span many years, harbours could well be detrimental to the health of the civilisation for many years.

[cynyck]

Particularly when running bureaucracy, your capital’s commerce multiplier (science/gold) can typically be 200%+ so for each commerce lost by the capital you will need to gain 2g in a city with no multipliers.

If its 200%, then does 1 not become 3? Which means a loss of 1 with 200% multipliers is even worse, because you will need to gain 3g in a city with no multipliers. Or am I missing something?

Oh, and thanks for the wonderful insight into harbors.