[Freeciv-Dev] Re: Thoughts about corruption

[Gregor Zeitlinger <zeitling@xxxxxxxxxxxxxxxxxxxxxxx>]

On Sun, 25 Nov 2001, Reinier Post wrote:
> > 1) I think it's easier and more predictable to make it a function that
> > depends on the total number of cities only. 
> > 
> > 2) Which city has how much corruption could still depend on the distance
> > to the capital. Still easier would be to have the same amount of
> > corruption in each city, as in Communism (civ2 at least). Order of
> > founding could also be a possibility.
> 
> Don't 1 and 2 contradict each other?
Well, what I meant was that corruption for a city depends on the total
number of cities or the number of a city (in order: distance to capital or
history) rather than the acutal distance to a city. In two I was
describing that order.

> > 3) How about making gold income and sience production a continuous
> > function rather than a discrete one. This would eliminate the problem of a
> > 10 and 1 city producing both 1 sciene because of 90% corruption.
> 
> It doesn't have to be a float; more precision, e.g. turning it into a
> percentage, will be good enough.
Of course, but it doesn't cost much to have a float or not? It could be
displayed with one digit after the period.

> > 4a) (luxury is continuous) Convert the excess luxury (after making
> > everyone content) into science or gold.
> 
> Mmm, you explicitly set luxury to 30% and gold to 20% and
> still get excess luxury turned to gold?
Yes, that removes incentives to micromanage the exact luxury rate thus
speeding up gameplay.

> > 4b) Pool all luxury together and then distribute it among the cities to
> > make all people content, or, if sufficient, happy.
> 
> That makes the game easier, but shouldn't there be a challange in finding
> the right luxury rate?  (Not that I've ever taken up this challenge,
> except in games where I really had time to burn.  Using entertainers
> is much quicker.)
I've accepted that challange each and every time and the one time I played
multiplayer it drove my opponents crazy!
Seriously: Yes, this model would change gameplay and balance more than 4a
but save more time.

> > The idea is to eliminate losses from roundings, which I hate, because I
> > love efficiency :). Predictability is another point though.
> 
> I'm not sure what you mean.  Having an effective UI isn't the same as
> making everything easy for the player.
This doesn't have anything to do with UI. If you have the rounding issue,
I'll try to set the trade in a city so that the rouding works in my favor,
which makes me advance faster, but costs me more time.

> > I also though about the function to calculate corruption. I think it plays
> > a central role in determining the optimal number of cities. Is it a lineal
> > one right now? Does anybody know, or know where to find?
> 
>    http://www.freeciv.org/lxr/source/common/city.c?v=cvs#L2046
Thanks, its linear up to 36 units map_distance and no check on the number
of cities.

> > 5) I would suggest something like (1/n)*base_trade, where n is your nth
> > city. (ln(n)/n)*base_trade is the equivalent function if it applies to all
> > cities, which has the convinience that you dont have to worry about which
> > city has with level of corruption.
> 
> But that makes Courthouses superfluous.  I also like countingthe distance
> to the capital, because it is a factor in deciding where to place cities
> and the capital.
On the contrary, as the function produces alot of corruption courthouses
are more important.

About the distance: It favors smallpoxers as they are less hurt by
corruption as the distance is shorter. You're also hurt if you start on a
small island and have to sail further for new cites.

bye

-- 
Gregor Zeitlinger      
gregor@xxxxxxxxxxxxx