Gainy bo said:
To build a worker it costs your city one population point. It also costs up to 10 turns to build. Then you've got the maintenance costs (which vary depending on your goverment). That's hardly 'free'.
Ok, this is becoming a little bit off topic now, but I feel it is needed to make my point of view clear:
We have to differ between
a) fixed costs
b) variable costs
The costs to produce the worker (in shields) and the maintenance for the worker are fixed. That means, as soon as you have a worker, you have spent 10 shields (aka 2 gold) and you have to pay 1 gpt (or whatever your form of government dictates). Whatever you will do with that worker has absolutely no influence on these costs. He could do roading, mining or whatever, it will always cost 1 gpt (2gpt, 3 gpt).
The costs for the improvement would be the variable costs, but there are just none of them. This is best shown at the example of fortresses which are considerably massive fortifications. Do you have to pay anything for them apart from the fixed costs of the worker? No, you don't.
Do you save anything, if your worker doesn't build a fortress, but constructs a road? Again, no.
As soon, as the worker is there, all of it will just be the same.
Therefore, after you've made the
strategic decision to create a worker unit, everything coming after that is for free (from the point of view of variable costs).
Well, let's look at the result after 100 turns. You created a worker (+10 shields = 2 gold, -1pop [costs of opportunity not taken into account here], +100 * 1gpt) resulting in total costs of 102 gold.
This worker constructed 100(turns)/6(base construction time) = 16.6666 ~ 17 roads.
Allocating the total costs to the total of roads => 6 gold per road. This is the portion of fixed costs any road tile has to carry.
Now, although there
IS a difference between
costs and
revenues, still those roads gain you money. At the end, if you subtract costs from revenues, you will have either gain or loss.
What do we have in this example?
Let's assume only 50% of those roads are "worked" upon by city population.
As you start at 0 roads in turn 1(worker) and end at 17 roads in turn 102(worker), the average is 8 roads per turn (for 100 turns, as the 17th is near completion, but not completed yet).
As 50% of the roads are worked upon, the average is 4 road tiles per turn worked upon.
Or, to make it more obvious:
turn 100 - 6: 94 gold (gain per 1st road in 100 turns)
turn 100 - 12: 88 gold (gain per 2nd road in 100 turns)
turn 100 - 18: 82 gold (....)
.....
turn 100 - 96: 4 gold (....)
=====================
at turn 100: 392 gold revenue (784 * 50%)
392 (revenue) - 102 (costs) = 290 gain
Of course, the figures become different taking into consideration that you will have more workers who not only construct roads. And we didn't calculate the "loss" in population against the added value by quicker science leading to multipliers as markets and banks (this would be the costs of opportunity)
Nevertheless, at the end the costs of all workers will be hardly 1 per improved tile.
So, I hold my argument, that any improvement caused by workers is just for free. At least, it will be that hard to calculate the "real" costs of a given improvement, so that saying "nil" costs will be a valid approximation.
Gainy bo said:
As for the Jav Thrower, it still costs shields/time to build, and there are maintenance costs associated with it also. Again, not free.
With the military units who did capture the slave, above calculation becomes more complex, but it should be easy to identify that the "costs" of a given military units cannot be allocated to a given tile improvement as the main reason for having this military unit hardly was the improvement, but the protection of your city in combination with the happyness provided by the unit.
You might regard this as
indirect costs, as I will admit.
But even if you would try to allocate those costs to the improvement, this would be just balanced by now not only having nil variable costs but nil fixed costs (per worker) either.
To come to an end: I admit that above example holds some slight inaccuracies as there is only a 33% chance to gain a slave. And the the chance to win against a barb might be something around 66% ~ 22%.
OTOH, everything which relates to the mentioned "costs of opportunity" hasn't been put in, either.
