New Game: Wreckreation

New Game: Wreckreation

[Eric McDonald]

I have added a new game to the Xconq games library. It is called
"Wreckreation", and can be accessed through the list of incomplete games
in the GUI's. I originally intended it to be a simple lab for testing
the new wrecking behavior, but it kinda grew on me. ;-)

I believe it is fairly playable, though much could still be done to
improve it. I will likely extend this game further as I feel inclined to
do so.

Feedback is always welcome.

  Have fun wreckreating,
    Eric

Re: New Game: Wreckreation

[Hans Ronne]

>I have added a new game to the Xconq games library. It is called
>"Wreckreation", and can be accessed through the list of incomplete games
>in the GUI's. I originally intended it to be a simple lab for testing
>the new wrecking behavior, but it kinda grew on me. ;-)

Very nice game! Some initial feedback:

After a few turns, the sides waste a lot of time capturing and recapturing
corpses. We should perhaps add AI code that gives wrecked and/or inert
units a lower priority as targets, unless they can be used as bases
(ruins). Might be useful in other games as well.

The map display gets cluttered with all these corpses, and since they are
labelled with side emblems you get the impression (particularly at low
resolution) that a side is stronger than it actually is. Wouldn't it make
more sense to have corpses become independent rather than switch sides (as
opposed to zombies)?

The Good side doesn't seem to be building any new villages or temples,
unlike the Bad guys.

It should perhaps be easier to capture temples, orc-holes etc. Now they are
usually destroyed before they can be captured.

All bases grow in size by 1 each turn. This is rather fast. Not that it
matters since the unit size doesn't seem to be used for anything.

Hans

Re: New Game: Wreckreation

[Eric McDonald]

On Sun, 2004-06-06 at 02:47, Hans Ronne wrote:

> Very nice game! Some initial feedback:

Thanks.

> After a few turns, the sides waste a lot of time capturing and recapturing
> corpses. We should perhaps add AI code that gives wrecked and/or inert
> units a lower priority as targets, unless they can be used as bases
> (ruins). Might be useful in other games as well.

I think that the victim finder still needs improvement. I don't think it
currently considers whether an unit is ACP-less and ACP-dependent or
not, but I agree that it would probably be useful if it did. I may
actually get a chance to look into that today, since I also need to find
out why it is not going after the city types in the Roman game.

> The map display gets cluttered with all these corpses, and since they are
> labelled with side emblems you get the impression (particularly at low
> resolution) that a side is stronger than it actually is. Wouldn't it make
> more sense to have corpses become independent rather than switch sides (as
> opposed to zombies)?

Actually, I intend to let sides do some fancier things with corpses.
Maybe I should make them independent until I get around to implementing
what I would like to have done with them.

Ideally, what I would like is to have two types of corpses: Fresh
Corpses and Decayed Corpses. Fresh Corpses could be animated by Vampires
and Necromancers and thus turned into Zombies, and resurrected by higher
level Paladins or Priests and turned into their previous unit type.
Decayed Corpses would have to be brought back to a Necropolis or Vampire
Castle, and then changed into Skeleton Warriors, or brought back to a
Temple or Paladin Temple and resurrected.

But, in the interim, maybe you are right, and I should make Corpses
independent, and then have powerful units of various sides "kill" them
to change them into fixed wrecked types (Zombies, Fighters, Orcs).

> The Good side doesn't seem to be building any new villages or temples,
> unlike the Bad guys.

In theory, the Good Villages should be able to build new Paladin
Temples, and Paladins should be able to found new Villages by changing
type (though I still need to make them acquire a certain amount of a
"settlers" material first). Once Heroes (very high level Fighters) are
added to the game, I will probably give them the same ability to found
Villages as higher level Paladins.

Of course, the AI needs to be taught to use manual change-type. Until
then, I would recommend that a human plays the Good side (or whatever it
eventually gets renamed to).

> It should perhaps be easier to capture temples, orc-holes etc. Now they are
> usually destroyed before they can be captured.

Right. Because these cities and facilities can only belong on certain
sides, they will vanish if they are captured. I did initially have a
100% capture chance for them, but they were disappearing right and left.
But, I agree that maybe Orc Holes, Catacombs, and Villages should be
easier to destroy. However, if you have played the Bad side, then you
know that Vampires can make fairly quick work of these structures.

> All bases grow in size by 1 each turn. This is rather fast. Not that it
> matters since the unit size doesn't seem to be used for anything.

Not presently. I hope to allow Villages to upgrade to Towns to Cities,
Orc Holes to upgrade to Dense Orc Holes to Orc Cities, Catacombs to
Major Catacombs to Necropoles, based on size, in the future. May do
something with Paladin Temples et al. as well. (This would be similar to
way things work in Bellum II.)

  Thanks for the feedback,
    Eric

P.S. It would be fun to be able to have ZOC range increase with unit
size. Then, if I added a "holy ground" protection effect for Paladin
Temples, the ZOC radius against undead could increase as temple size
increased.

AI Build Tips

[Elijah Meeks]

Does anyone have any tips on how to make a unit a more
appealing choice for the AI to build?


	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 

Re: AI Build Tips

[Hans Ronne]

>Does anyone have any tips on how to make a unit a more
>appealing choice for the AI to build?

This depends on two things: the unit's plan and whether it is an advanced unit.

For non-advanced units, the build target is decided by
preferred_build_type, a huge and complicated function in ai.c. It does all
sort of calculations having to do with the need for explorers, transports
etc. The best way to learn how it works is to examine the code.

For advanced units, the choice is made by auto_pick_new_build_task in
run.c. However, this function just loads precomputed unit worths into a
utype vector and then picks a weighted random number. So the real choice is
made by the worth functions which live in ai.c.

There is one worth function for each plan type (offensive_worth etc), so
figuring out how they work is easier than for preferred_build_type. For
example, offensive_worth uses the following formula:

u_acp(u) * u_attack(u) * u_range(u) * u_speed(u)

while defensive_worth uses:

u_acp(u) * u_defend(u)

both in combat model 1. For combat model 0, the calculations are more
complicated since there are no absolute attack and defend values. The code
therefore compares the unit to all other unit types in the game and
calculates a relative attack or defend value.

So the short answer to your qestion is that a fast strong unit with many
acps will almost always be preferred over a slow weak unit with few acps.
Hope that helps.

Hans