[freeciv-ai] Re: Measure the AI ability to harvest resources

[Raimar Falke <rf13@xxxxxxxxxxxxxxxxx>]

On Mon, Nov 25, 2002 at 10:12:07AM +0000, Per I. Mathisen wrote:
> On Mon, 25 Nov 2002, Raimar Falke wrote:
> > > > > I haven't searched why the server AIs perform badly. It is quite
> > > > > possible that they built warriors or barracks.
> > > >
> > > > Looks like I know the errors of the server-AI.
> > >
> > > I believe your conclusion is premature. Check worker placements in those
> > > cities. Do the cities get enough food?
> >
> > Even if they don't get enough food this isn't an excuse to build
> > barracks.
> 
> I don't know if this is because it uses barracks as fallbacks (possible
> since it has so little tech == so few other things to build) or because it
> wants to build warriors later.
> 
> If it is the former, then barracks is indeed a bad choice, and
> capitalization might even be better. But the case of researchcost 100 is a
> bit odd. I'll look at it, though.

> What _should_ it be building instead of settlers, though? Perhaps military
> units?

The answer is easy: settlers. If it doesn't have enough food it has to
produce enough food.

> I forgot one additional AI problem:
>  - The AI doesn't understand that although the city may be short of food
> at its current size, there may be sufficient food for a settler after it
> has shrunk by one. The impact of this is probably not big once the AI
> learns to optimize cities for food since getting enough food is rarely a
> problem if you only know how...

        Raimar

-- 
 email: rf13@xxxxxxxxxxxxxxxxx
  Microsoft does have a year 2000 problem. I'm part of it. I'm running Linux.