[Freeciv] Re: [ICS] yet another suggestion

[Mathias Uhlmann <Mathias.Uhlmann@xxxxxx>]

Hi!

As far as I understood, You suggest that the population distribution is nearly
Gaussian centred around big cities and
to punish players that don't have such a distribution. I have several
arguments against this and will simply list them.
  - you don't know the width of the distribution, in reality this width will
differ very strong
  - additional (in reality) you have the effect of overlapping distributions
(Saxony is one example of this, but it should also be visible in
    England, the Netherlands, Belgium, the Ruhrgebiet (Ruhr area or so)), in
every area that is settled very dense, you will never see the
    Gaussian distribution
  - the idea of a urban centre surounded by smaller cities and villages is in
(Free)Civ already included, the cities in the game are not like
    cities in reality but more like whole areas, with small towns and
agricultural areas around the big city
  - your method is very complicated, even if the first two arguments would not
apply, it is very hard for the player to implement a strategy
    your solution would need. I always try to grow all cities to the limit.
There are no Gaussian distributions there.
All in all I don't think this is applicable. I would rather favour the other
suggestion made recently, make settlers more expensive to build and
cheaper to upkeep.

Ciao.
Mathias.

Sylvain Boivin wrote:

> Hello all,
>   spended tons of hours playing this marvelous game. Freeciv is a great
> achivement.
>
> I have a suggestion related to the ICS problem. My suggestion is based on
> the fact that, in real life, when a civilization reach a given pop size,
> the
> spatial distribution of the pop tend to be the same. A big city, surrounded
> by smaller one.
>
> To reproduce it in Freeciv, the code should discourage "flat" pop
> distribution.
> It can be done by a set of rules like:
>   - for each city i of a given civ C, search the biggest city T of the
> same civ
>       in a given area A, that is a function T(C,i,A)
>          - compute the observed spatial distribution s of pop around T
>          - compute the ideal spatial distribution S of pop around T
>          - if s is below S (that is too flat), higher is the # of
> unhappy in city i
> It should do the job if:
>   1) the parameters of the ideal gaussian dist S are scaled with the
>        pop size of the civ C, in a given area A. That is, there is no
> penalty
>        to have some small town in a newly colonized area, but the penalty
>        is high when you have too many small town in the same area without
>        some big one.
>   2) the number of unhappy citizen generated by this mean should
>        remain below a given # in each area to always a minimal viability
>        of the area.
> With the right scale, for a flat civ with 10 cities in the same area A,
> adding a new
> city will create a city containing only unhappy citizen.
>
> Comments?
>
> Bye.
>
> --
> [Sylvain Boivin]   [http://dimsboiv.uqac.uquebec.ca/~sboivin/]
> [sboivin@xxxxxxxxxxxxxxx] Use Linux! [Yahoo: artoxx@xxxxxxxxx]