[Freeciv-Dev] Re: Dips. OOPS please read this one!

[Tony Stuckey <stuckey@xxxxxxxxxxxxxxxxx>]

On Tue, Feb 27, 2001 at 02:31:36PM -0500, Jeff Mallatt wrote:
> At 2001/02/27 14:17 , Thue wrote:
> >On Tuesday 27 February 2001 20:09, Jeff Mallatt wrote:
> >> At 2001/02/27 13:50 , Thue wrote:
> [snip]
> >> >Diplchance is 80 to begin with - 80 percent chance to defeat defending
> >> >diplomats! This should rather be around 20, no?
> >>
> >> From what I read when I implemented it, this (80% for diplchance) works out
> >> to be very close to the Civ2 behavior.
> >
> >But then in civ 2 diplomats are way overpowered!
> 
> Yes, in Civ2 they *are* very powerful.

        The Civ2 behavior is described on page 111 of the manual:

        "Diplomats and spies stationed in friendly cities have a chance to
thwart 'steal technology' attempts by enemy Diplomats and Spies.  Each
Diplomat has a 20 percent chance to do so per attempt.  Spies have a 40
percent chance; veteran Spies have a 60% chance of catching their fellow
envoys.  Getting caught ends the interloper's turn."

        Insert standard disclaimer about the code not matching the docs. :)

> >We should lower that value considerbly in freeciv mode.
> >We also need to raise the cost of inciting and bribing; they are way to
> cheap 
> >right now (as seen by the fact that many players builds diplomats
> exclusively 
> >for land warfare).
> 
> Some changes may be to the good.  Be careful with diplchance, however --
> from what I recall (without looking at the code) it's very sensitive.
> Lowering it too much could make diplomats completely useless.

        I'm not sure that the Freeciv equation is grounded in anything.
(gold+750)/(dist+2)*(unit_build_cost/10).  It depends on the unit build
cost and is lowered by distance from the home city, but those don't seem to
work in the same relation as Civ2.
        Civ1/2 was based mostly upon the unit type, lowered by distance from
the home city.  For instance, a Legion would cost between 41 and 105 gold to
bribe.  Settlers cost way more than other units -- the high would be like
1200.  I think we should consider a (2n + (n^2/20))/dist style equation
where n is the unit build cost.  (Similar to the rush-build equation.)
-- 
Anthony J. Stuckey                              stuckey@xxxxxxxxxxxxxxxxx
"And they said work hard, and die suddenly, because it's fun."
        -Robyn Hitchcock.