[freeciv-ai] Re: Massive AI 5
[Top] [All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
On Tue, Nov 12, 2002 at 09:11:39PM -0800, Raahul Kumar wrote:
>
> --- "Anthony J. Stuckey" <astuckey@xxxxxxxx> wrote:
> > On Tue, Nov 12, 2002 at 06:56:43PM +0000, Per I. Mathisen wrote:
> > > I'm against sprinkling the AI code with such braindamage. The AI should be
> > > limited by a few, well-placed handicaps, not a hundred small pinpricks
> > > that complicates the code.
> >
> > Yes, and H_PREFER_INEFFICIENT_PRODUCTION can be one of them, although I
> > agree with you and not Rahuul.
>
> I was thinking of adding two handicaps. BAD_PRODUCTTION and changing the units
> allowed to the easy ai.
>
> I believe the easy AI should be denied musketeers, knights/horseman,howitzers,
> mech inf,rifleman,pikemen and musketeers, ironclads, and battleships. I am
> certain this will prevent the Easy AI from beating anyone. If someone still
> loses to the easy AI after this they are too stupid to help.
This is dumb. We want an easy ai, not one that has a completely different
ruleset that the humans.
The key to having an easy ai is making it SLOW. If you want to have the ai
be somewhat dumb in that it doesn't attack when it could, expand as fast as
it can, etc, fine, but this suggestion is simply over the top.
An easy way to make it slow is simply to switch from the tech it's
researching as it's getting close to making the discovery. Switch from the
wonder/building it's building right before it gets done. Do this behavior
with high (maybe 70%) probability. That way it takes "forever" to get the
techs to build ironclads or get republic, etc. This way, a neophyte gets to
see all the bells and whistles that the ai has, and if he waits long
enough, he will still be able to get killed, (by cool, high-tech units),
but he'll have to wait a very long time. I'd call it H_INDECISIVE.
-mike
[freeciv-ai] Re: Massive AI 5, Anthony J. Stuckey, 2002/11/12
|
|