11-12-2002 14:50									#1
	Why is the AI not keeping track of ZOC? Suffering from ads?

I do not understand the programming of the AI.

Time upon time the AI sends units past a mountain where I have stationed a defensive and a offensive unit.

On average one in four-five of his units gets damaged by the effects of the zoc by my offensive unit.

Instead of going another way or trying to kill the mountained units, the AI keeps sending troups by the mountain.

Suicide is painless or what?

This is a real flaw ( it existed in CivII also) and should not be so difficult to resolve.

Now I can make a strategy out of it, but that is not the point of the game. It feels like cheating.

Or is it?

	11-12-2002 15:03									#2
	Support Apolyton, pre-order Civilization IV

how did this exist in civ II? ZOC was handled in a completely different way.

	11-12-2002 15:14									#3
	Increase Your PM Length

quote: Originally posted by TacticalGrace how did this exist in civ II? ZOC was handled in a completely different way.

you are right, I do not know where I got it

sorry

	11-12-2002 21:52									#4
	Got spare money?

What annoys me is when you're moving units to attack a city or troops along railroads with 13 or so different paths to the place you're going, and the pathfinder takes the shortest root regardless of the posision of enemy troops and their ZOC. Though it rarely changes the ultimate outcome, you would think the pathfinder would try to keep your troops from taking damage because of enemy ZOC when there is a choice.

	11-12-2002 22:16									#5
	Avatar Enlargement: We've got the solution

And its the pathfinder that does exactly the same thing to the AI's troops which answers the question.

	12-12-2002 00:02									#6
	Remove this text

As if the pathfinding didn't take up enough time in a turn, especially in the end game. Adding a provision to steer clear of enemy units would increase it even further.

Then we'd also have people complaining that their troops always end up away from the forces their supposed to be attacking. I'd say live with it, plan your own moves accordingly, and consider the enemy's bad moves as taking orders from a stupid general.

	12-12-2002 09:57									#7
	Support Apolyton, buy Civilization: The Boardgame

Well, I see your point about the pathfinding taking up more time making the game slower, but I would rather have smarter & slower than faster & dumber. The small change I would be happy with is a check to see if there are alternate paths with the same movement points which do not pass by the enemy.

Of course, this wouldn't fix the AIs gambit through ZOCs to get to a city for example. You'd have to add some intellegence to the AI in his attacking decisions.

	12-12-2002 14:27									#8
	Support Apolyton, buy Galactic Civilizations

There are much fewer AI suicides now than in civ2. As for the ZOC problem. I simply don't know if the AI will take advantage of an safer path to it's targets if such path exist. In most of my games such paths do NOT exist, therefore a direct approach makes more sense.

	12-12-2002 19:50									#9
	Got spare money?

quote: Originally posted by ChrisiusMaximus And its the pathfinder that does exactly the same thing to the AI's troops which answers the question.

Yeah but it seems like the pathfinding is a lot worse in PTW. I never had a problem with it in Civ3, not once. But now it's become commonplace to see my units get snared in an infinite loop, jumping back and forth between two spots when they could have gone around the clusterf*** a million other ways.

	12-12-2002 20:15									#10
	Got spare money?

So I would not be bad if the pathfinder (for us and the AI) stops at the first encounter with a enemy unit.

Both could then redirect its units.

That should not be so difficult to reprogram.

It could also solve the problem when you use the pathfinder somewhere far away and it gives a movement rate of O whereas now sometimes you get blocked by friendly and enemy units making the correct movement rate 2 or 3.

I know if you move tile by tile that this last remark has no value.