RE: More revisions to exploring_worth

["Stanley Sutton" <sutton at t-surf dot com>]

If you think that through, you need to include the probability of seeing
the unit, too.  If the AI is encountering a lot of saucers, then it
needs to build defenders, but if it's encountering hovercars, it needs
saucers.  The value of the different type of units will vary with the
situations encountered.

If you calculated the probability of encountering the unit type and
multiplied it times the offensive/defensive worth of each unit against
the other units, you might get a better result for the mode 0 worth
calculation.  This would greatly increase the need for observors near
the enemy in order to have good informaiton to base the calculation on.



-----Original Message-----

From:	Lincoln Peters [mailto:peters2000@mindspring.com]

Sent:	Tue 23-Jul-02 15:00

To:	Hans Ronne

Cc:	Xconq mailing list

Subject:	RE: More revisions to exploring_worth


On Tue, 2002-07-23 at 11:11, Hans Ronne wrote:

> The multiplicative approach works fine as long as we have a way of

> assigning distinct values, such as speed, to individual units. I think
the

> latest version of Lincoln's exploration formula illustrates this. A
problem

> does arise when we want to compute attack and defense weights in
combat

> model 0 (the standard xconq combat model). The reason for this is that

> there are no absolute attack or defend values similar to speed. All we
have

> is the hit-chance table, which gives relative strenghts of different
units

> against each other. The AI code tries to handle this by summing up and

> averaging the hit chances against all units in the game, but I think
there

> is room for a lot of improvements here.



What if the offensive_worth and defensive_worth functions could be

re-written so that they could indicate that unit x is a good attacker or

defender, but only against unit y, or against everything except unit z? 

This would undoubtedly help the AI play games such as future.g where all

units have at least one unit they can easily kill and another unit that

can easily kill them (hovercar < saucer < defender < hovercar, for

example).



> 

> With combat model 1 it is much easier. You just use the unit's unique

> attack and defend values. Check out offensive_worth and
defensive_worth in

> ai.c to see what I mean.



Very true, but you lose the ability to do "paper-scissors-rock" things

like the one I just described from future.g.  I think that time.g also

does something like this with infantry, cavalry, and cannons, at least

in the Napoleonic sub-period.