Unit Abilities

[asleepathewheel]

Originally posted by GhengisFarb
Actually, I thougth the enslavement option was a cool idea.

I wish there was away to simply have it enslave what it captured with ships. So you could more accurately simulate the Spanish Main with a chance of commandeering Galleons and Frigates and not having them all shrunk into Privateers.

yeah, again firaxis has a good idea, but has a sloppy followthrough

(I know its been recommended to them to change this in the next patch, or make it editor-able, who knows if they do that or not.)

[Kuciwalker]

yeah, again firaxis has a good idea, but has a sloppy followthrough

Sloppy? Because the system doesn't do everything one could possibly hope, it's sloppy?

[asleepathewheel]

Originally posted by skywalker

yeah, again firaxis has a good idea, but has a sloppy followthrough

Sloppy? Because the system doesn't do everything one could possibly hope, it's sloppy?

yes.

its been a pretty big complaint by modders.

pretty farfetched turning galleons into privateers while on the high seas.

i can see units into slave workers.


Its a neat ability, but needs some more modability.

(And I thought I was a huge firaxis fanboy )

[Kuciwalker]

Again, because the system doesn't do everything one could possibly hope for, it's sloppy?

[asleepathewheel]

Originally posted by skywalker
Again, because the system doesn't do everything one could possibly hope for, it's sloppy?

yes

[asleepathewheel]

I'm just giving you crap, sw, sloppy was a bit of hyperbole.

what I should have said was that it was a bit short-sighted in implementation.

[lajzar]

Here's one that should be brought back, and its corrolary:

Exerts zoc - as per civ2. Air units don't have a zoc if implemented civ3 style. Sea units with guided missiles (modern navies basically) have a zoc. Land units with missiles (archers upwards) have a zoc. This is of course a flag to be set, not hard coded.

Ignores all zoc - what it says. Should probably apply to submarines, guerillas, and diplomat/spy units.

oh, here's another flag:

AA gun [n] - will automatically launch a bombard attack of strength N on any aircraft unit that passes through an adjacent tile. Pass through here includes a) the straight line drawn between two cities when the airports are used for transport, and b) the straight line drawn between the base site and the target of a bombard mission.

Paradrop [n] - the unit can, as its entire move, jump up to N tiles from the city or airfield it is in. This move may be intercepted by units with the AA gun flag.

[Switch]

Originally posted by lajzar
Here's one that should be brought back, and its corrolary:

Exerts zoc - as per civ2. Air units don't have a zoc if implemented civ3 style. Sea units with guided missiles (modern navies basically) have a zoc. Land units with missiles (archers upwards) have a zoc. This is of course a flag to be set, not hard coded.

Ignores all zoc - what it says. Should probably apply to submarines, guerillas, and diplomat/spy units.

I would have to disagree whole heartedly on this. Civ3's ZOC is much better than civ2's, IMO, and is actually more realistic. Say you're playing a giga earth map, 256x256 or whatever...I'm not sure how wide the tiles are, but each tile works out to be several hundred km...sometimes even the civ3 style is a bit much, but i think it works fine.

oh, here's another flag:

AA gun [n] - will automatically launch a bombard attack of strength N on any aircraft unit that passes through an adjacent tile. Pass through here includes a) the straight line drawn between two cities when the airports are used for transport, and b) the straight line drawn between the base site and the target of a bombard mission.

Paradrop [n] - the unit can, as its entire move, jump up to N tiles from the city or airfield it is in. This move may be intercepted by units with the AA gun flag.

Agree with you here. The way it works now, it's as if planes teleported . And paradropping was nice too, in SMAC. Maybe make it so that paratroopers can paradrop, ie a limited range from the city, and modern paratroopers can do strategic insertions, ie no range limit (like SMAC's orbital insertions).

[wrylachlan]

1) Air superiority on enemy cities to protect bombers

2) Abstract Ship to Sub combat similar to air missions. Some sort of "Depth Charge" mission.

3) Capture - different than enslave, it takes the unit not as a worker, but as whatever type of unit it is. The caveat is you can't use them except to move and to TRADE them back in a peace deal.

4) Civil engineering - have multiple sources of each resource, but only one automatically pops up on the map. You have to find the others.

5) Hidden numbers - if you use stack movement in certain terrain, it looks graphically like only one unit moves. So what looked like 1 Iroquis warrior, actually turns out to be 5 when they come out of the woods.

6)Specific bonuses and penalties versus certain other units i.e.pikemen get a percentage bonus versus mounted, but a percentage penalty versus foot.

7) Pillage-heal. Just what it sounds like.

8)Scout tracking
-----------------------------------------------------------------
And don't listen to skywalker, the idea of tying unit abilities to terrain types is

Archers getting better defense in forests/jungles = great
Certain units getting to treat certain terrain as roads = great
And I'll add -
Mounted units get an x% charging bonus on plains or grasslands.
Ranged units get an even better defense bonus than regular units when defending across a river or behind walls or fortresses.

[MrBaggins]

Originally posted by wrylachlan
*SNIP*
And don't listen to skywalker, the idea of tying unit abilities to terrain types is

Archers getting better defense in forests/jungles = great
Certain units getting to treat certain terrain as roads = great
And I'll add -
Mounted units get an x% charging bonus on plains or grasslands.
Ranged units get an even better defense bonus than regular units when defending across a river or behind walls or fortresses.

Do you have any clue as to how complex that would be for the AI to figure out? The defensive boni of terrain is enough... adding a bonus to specific units on specific terrain virtually guarantees that only a human would use it effectively.

[wrylachlan]

Originally posted by MrBaggins Do you have any clue as to how complex that would be for the AI to figure out? The defensive boni of terrain is enough... adding a bonus to specific units on specific terrain virtually guarantees that only a human would use it effectively.

Looking at your post in the other thread about ZOC I'm noticing a pattern here. Anything that is difficult for the AI to implement that you happen to like is OK. But if you don't like the idea, you shoot it down with "Do you know how difficult that is for the AI?". Well too bad. Its a good idea. I like it. And given enough processing power, and intelligent design, it could be implemented. After all, what's the good of having an AI that's great a checkers??? It's still checkers.

[MrBaggins]

I've been working on (producing mathematical models and algorithms for Civ AI) long enough to know whats expensive and difficult, and whats easy and cheap.

ZOC's are a bit more expensive than no ZOC's. You need a (2-state) blocking map, which simply alters the cost for your A* pathfinding. Its a bit more expensive, but not THAT much... simple to calculate and to use. This cost is even, in some circumstances, outweighed by the reduced number of entities (and hence AI cost,) and that forgets the game benefit to the player of reduced MM.

Terrain-based boni necessitate not only a costly db lookup for each of the units per terrain, or if you cache the information in the army record, a slightly less costly lookup.

This variable state modifier has to be calculated for opposing troop types on each of the 8 surrounding squares of a square, when a unit isn't directly attacking, and its within range of enemy units, to verify that its not moving to a more vulnerable position.

All in all... this is a very costly. I'm not saying we should be playing checkers... just that some ideas require unrealistic amounts of AI time. This idea is one of them.

Computers won't be that fast when Civ4 is released. The fastest computers now aren't exactly greased lightning now, with Civ3's AI.

[GhengisFarbâ„¢]

Originally posted by MrBaggins
Do you have any clue as to how complex that would be for the AI to figure out?

The AI can't figure out the concept of upgrading units and attacks Modern Armor with Archers.

If the only good reason to exclude an idea is because the AI can't grasp it, then I wouldn't be too worried about it.

[MrBaggins]

The simplication of air combat made that viable for the AI to use. (Of course even thats missing some vital functionality, like air superiority... but still.) The map is typically a pain in the ass. You need to avoid unecessary complications.

If the AI doesn't upgrade then its either a) not resourced correctly or b) just bad strategy (although it might be attacking a point that wasn't well defended when it made the plan.) Threat matching is pretty fundemental. Although you want to have some concept of ablative attacks, its a definite short coming, if you don't concentrate the attacks.

Upgrading is typically a tremendous benefit to the AI. Witness Updater2 in CTP2

As for including advantages that you can use, but the AI can't... that lessens the challenge of SP... it makes the AI easier to beat.. and just plain makes it look bad.

Witness the AI's weakness competing vs the crawler tactic in SMAC.

[wrylachlan]

Originally posted by MrBaggins Terrain-based boni necessitate not only a costly db lookup for each of the units per terrain, or if you cache the information in the army record, a slightly less costly lookup.

This variable state modifier has to be calculated for opposing troop types on each of the 8 surrounding squares of a square, when a unit isn't directly attacking, and its within range of enemy units, to verify that its not moving to a more vulnerable position.

I would humbly suggest that if people thought it was a good enough idea for game-play reasons, AI coders would come up with a clever way to implement it.

The movement bonuses are already essentially in the code since wheeled units have different movement penalties than foot soldiers. There's no reason that expanding that concept to allow other special cases wouldn't be a no brainer.

And as for the combat bonuses, there should already be some sort of code that takes into account "other powers". Otherwise an enslaving unit would be treated exactly like another unit of the same power, which is tactically incorrect since a loss to an enslaving unit is a "greater loss".

And the bigger issue, for me, is that Firaxis is capable of assessing the difficulty of programming the AI. By all accounts the Civ3 AI is significantly better than Civ2, and has gotten progressively better with each patch while adding increasingly complex game mechanics, such as zero-range bombardment.

By definition, this forum is for brainstorming what we would like the game mechanics to be if Firaxis thought it practical. If we come up with what we think is a great idea, but Firaxis can't implement it, so be it. As such your constant harping on the difficulty of the AI implementation doesn't really serve to encourage the spirit of brainstorming that we're working on.

It's better to come up with as creative ideas as possible, and leave the decision on the practical aspects of implementing it to Firaxis.