Possible cheap strategy.

sucksforyou112

Superman
Joined
Jun 23, 2010
Messages
110
Location
USA
You know, i wonder if it would be possible to make a city, build roads everywhere, and then gift it to another civ to cripple their economy. Although a city grows one tile at a time, having a city grow normally, then build roads everywhere and gift it immediately to another nation to cripple their economy right before declaring on them might be a little cheap. What do ya think?
 
Is it worth a settler, territory, worker turns, money lost into roads while you are still preparing the town?

This is pretty much a suicide strategy. Nice idea tho. Maybe drop a nuke on your city (assuming you are already that far) to generate some fallout, too.
 
Roads aren't that expensive (1 gold per turn per hex), so you won't be able to crash someone's economy by gifting them expensive cities.

You'd also have to build the city, during which time you're hemorrhaging gold into it, so you'll end up doing more damage to yourself than the opponent.
 
Even in civ 4 the AI would not take crappy cities from you or cities that would incur large maintenance costs I assume thee same would apply ti civ 5
 
Even in civ 4 the AI would not take crappy cities from you or cities that would incur large maintenance costs I assume thee same would apply ti civ 5

Oooooooh yes it would! (See TheMeInTeam's YouTube Let's Plays for ample examples of this strategy.) It usually just had to be close enough to the enemy territory for them to see it as a "liberate" instead of "gift." Which actually can be surprisingly far away.

As for the OP though, I don't think it would be that great of a strategy for the reasons outline before me.
 
Roads aren't that expensive (1 gold per turn per hex), so you won't be able to crash someone's economy by gifting them expensive cities.

You'd also have to build the city, during which time you're hemorrhaging gold into it, so you'll end up doing more damage to yourself than the opponent.

Actually its 2 gold per turn.

...

And if someone wants to waste their time doing that then fine, ^>^. Their workers will probably be able to demolition the roads anyway so whats the point.
 
Actually its 2 gold per turn.

What? No it isn't. From the main site:

Road. Increases the movement speed of units moving along them. Connecting cities to your capital will provide a large amount of gold per turn. Costs 1 gold per turn to maintain. Requires The Wheel.

Oooooooh yes it would! (See TheMeInTeam's YouTube Let's Plays for ample examples of this strategy.) It usually just had to be close enough to the enemy territory for them to see it as a "liberate" instead of "gift." Which actually can be surprisingly far away.
But now that that cultural effect is gone, it's likely the AI won't just take whatever you offer it without analyzing it first.
 
Actually its 2 gold per turn.

...

And if someone wants to waste their time doing that then fine, ^>^. Their workers will probably be able to demolition the roads anyway so whats the point.

I'm guessing it you wait until railroads are available (which cost 2GPT) it would have even less impact...
 
Workers have the ability to remove roads. Whether the AI is clever enough to use it is unknown, but it's not hard to imagine that the AI has a function whereby it may remove roads if the costs are too high and the algorithm says it has too many.

Roads only cost 1 gold in maintenance per tile. If you found a new city, it's only got 6 tiles around it. Every tile beyond that you have to earn through culture or pay for with gold. You'd have to invest a HUGE amount of gold to build a city up to the point where it might be a drain on the enemy's resources. And if the enemy can just remove the extra roads and say, "thank you for the territory!"
 
Workers have the ability to remove roads. Whether the AI is clever enough to use it is unknown, but it's not hard to imagine that the AI has a function whereby it may remove roads if the costs are too high and the algorithm says it has too many.

I forsee the AI building roads, then tearing them down, then building roads.....
 
And there are screenshots that add credence to the idea that we might be able to simply burn cities down at will. (Long Overdue addition to the franchise IMO.)
 
I forsee the AI building roads, then tearing them down, then building roads.....
You mean like how it currently tears down Towns, replaces them with Workshops, then builds Cottages again?

Yes, I have little doubt. It's hard to build an AI with any kind of persistent memory of what it's done as opposed to what it's trying to do.
 
It's hard to build an AI with any kind of persistent memory of what it's done as opposed to what it's trying to do.

The memory is very, very easy. Planning is more difficult, but as we already know, the planning appears on all AI tiers, so I think it will be quite good in placing improvements.
 
The memory is very, very easy. Planning is more difficult, but as we already know, the planning appears on all AI tiers, so I think it will be quite good in placing improvements.
Constructing an expert system that evaluates a map and comes up with numerical ratings to determine what to do is fairly straightforward. Preventing the system from changing its mind every turn by remembering what it has been doing for the last n turns is much more difficult.
 
Constructing an expert system that evaluates a map and comes up with numerical ratings to determine what to do is fairly straightforward. Preventing the system from changing its mind every turn by remembering what it has been doing for the last n turns is much more difficult.

You don't need AI to reevaluate the whole strategy each turn. You should only consider new information and reevaluate relevant plans. If AI is built right, your own actions (like building improvements) will not cause reevaluation, as well as predictable events like city growth. All these should be a part of the plan.

Plus, it's easy to calculate the plan change cost and make decision if it's worth it.

P.S. Actually the most difficult part is "Constructing an expert system that evaluates a map and comes up with numerical ratings to determine what to do" :)
 
You don't need AI to reevaluate the whole strategy each turn. You should only consider new information and reevaluate relevant plans. If AI is built right, your own actions (like building improvements) will not cause reevaluation, as well as predictable events like city growth. All these should be a part of the plan.

Plus, it's easy to calculate the plan change cost and make decision if it's worth it.

P.S. Actually the most difficult part is "Constructing an expert system that evaluates a map and comes up with numerical ratings to determine what to do"
I'm guessing from your responses that you're not a professional software engineer. I am.

Programmatically, it's much easier to have the the AI reevaluate its strategy every turn; it's more CPU cycles, but no extra code. Having the AI form some kind of memory structure about what it's doing, and then somehow comparing this structure with new information is much more programmatically complex.

Assigning a numerical value to each terrain type and each improvement and then running a simple algorithm to get a "best" result is not very difficult, and doesn't require much computation. That's what the AI does in Civ IV every time you see the blue circle "recommendation" from the AI (and I would guess from what I've seen of AI civ behavior, they use this same algorithm). It's a really basic calculation, and though it's often not great, it's good enough for most game purposes. The problem comes in when the AI runs this calculation every turn, and game states change every turn, so that the game's blue circle recommendation changes from turn to turn, and you see the AI civilizations change their minds frequently. I have often seen AI workers replace one improvement with another, and then begin building the previous improvement as soon as the new one is finished. Giving the AI a "world state" image and memory of its goals for every tile that persists from turn to turn is a really complex programming job. I'm not saying it can't be done (it can), but it's not easy, and as far as I can tell they don't do it much at all in Civ IV.

I think you can see a little bit more persistent planning in the offensive military behaviors of the AI in Civ IV. They do assemble stacks and send them on missions, and you generally can't get them to consider peace until this operation has either finished its objective or been destroyed. But I've also seen enemy forces dance from one spot to another every turn when I move my defenders back and forth... their objectives seem clearly calculated on the disposition of my forces, and change every time I move them. I've kept enemy invading forces dancing on the same hill or forest tiles for untold ages.
 
Greg's response to this identical post from the 2K forums:

2K Greg said:
Well first of all, the receiving player has to accept a gift; it's not automatic (I am fairly sure about this.)

But even if that weren't the case, I would have a very hard time believing that the damage done to the giftee would be anywhere close to the damage done to your own empire during the many, many turns it would take to do what you said. The giftee could just have his workers head on over and remove the roads quicker than it took you to build them in the first place. Having a whole extra city that you're paying happiness and gold to maintain just to give another player a few-turn gold hit is probably not very efficient :)
 
Top Bottom