[Freeciv-Dev] Re: As if we needed another ics solution...

["Miguel Farah F." <miguel@xxxxxxxxxxxxx>]

 Mike Kaufman [01/03/2001 19:00] dijo/said:
>
>I haven't looked through the code to see how feasible this is, but...
>
>perhaps adding (as a server option of course) a production penalty might
>be a solution. 
>Ex.: up to a certain number of cities, a chariot, for example, costs X.
>If I would build Y% more cities, then a chariot would cost X+Y to build
>(or whatever algorithm you choose, you could also make it discrete
>increments)

I disagree on this proposal (as it is): it's not the same to have 50
cites, all size 1-2, than having 50 cities, most of them bigger than,
say, size 8.

Most proposals to eliminate ICS, IMHO, forget to pay attention to this
factor.

What I propose is to have first some sort of measure of ICS (an
"ICSmeter") to determine wether penalties of any kind sould be applied.

What I propose is this: we know that a city uses up 21 squares
(including the free city center). An ideal nation that has NO
overlapping cities of any kind, will control exactly 21*N squares
(N=number of cities).

ICSmeter will defined as this: 

            squares occupied by a nation
   I := 1 - ----------------------------
            21 x number of cities owned

(a square is considered occupied by a nation even if its owned by an
enemy - that can't be counted against the player)

Now, and ideal nation will have I=0%.

A nation that places 25 cities all in a square grid, all at distance 1
will have I=1-(165/(21x25))=68,6% and commits HEAVY ICS.


Considering the defects of the terrain, and conquered cities, an I=0%
will be unattainable, and therefore an index of, say, up to 15% should
be acceptable.

Nations with a bigger index should then begin suffering penalties.


-- 
MIGUEL FARAH              //   miguel@xxxxx
#include <disclaimer.h>   //   http://www.nn.cl/~miguel
<*>
"Trust me - I know what I'm doing."
- Sledge Hammer