[noob question] culture taken tiles

ddd123

Prince
Joined
Sep 26, 2008
Messages
387
maybe im missing something but too often ai doesnt take the best tile

whats the logic behind it? and is there a way to inlfuence it so the culture may take the real best tile available?

also would be cool to make it go TOWARDS best tiles if they are far but i understand this could be tricky

ty
 
The game doesn't like taking hills, forests or mountaints. It also doesn't like expanding across a river. Short of buying the tiles you don't want culture to spread to, I don't think there's anything you can do.
 
I was wondering what determined the costs of individual tiles. Hmm.
 
To be more specific, here's the info from the GlobalDefines

<Row Name="PLOT_INFLUENCE_DISTANCE_MULTIPLIER">
<Value>100</Value>
</Row>
<Row Name="PLOT_INFLUENCE_RING_COST">
<Value>100</Value>
</Row>
<Row Name="PLOT_INFLUENCE_WATER_COST">
<Value>25</Value>
</Row>
<Row Name="PLOT_INFLUENCE_IMPROVEMENT_COST">
<Value>-5</Value>
</Row>
<Row Name="PLOT_INFLUENCE_ROUTE_COST">
<Value>0</Value>
</Row>
<Row Name="PLOT_INFLUENCE_RESOURCE_COST">
<Value>-105</Value>
</Row>
<Row Name="PLOT_BUY_RESOURCE_COST">
<Value>-100</Value>
</Row>
<Row Name="PLOT_BUY_YIELD_COST">
<Value>10</Value>
</Row>
<Row Name="PLOT_INFLUENCE_YIELD_POINT_COST">
<Value>-1</Value>
</Row>
<Row Name="PLOT_INFLUENCE_NO_ADJACENT_OWNED_COST">
<Value>1000</Value>
</Row>

Hills and forests are more expensive because their "distance" is twice as large seeing they cost 2 moves.

So the game will always get all flat terrain in the second ring before claiming any hills or forests unless you have a resource there. It will also claim flat tiles in the third ring before hills in the second ring if they have a higher yield, for example because of being next to a river.

All in all, as I said before: Doesn't like hills or forests, doesn't like river crossings much, likes resources ;)
 
That looked very textual.. Can that just be altered by ourselves for single player games?.. I'd want it to make hills/forests a bit better too :)
 
Yes, you have to edit a few points, though, and best make a mod of it for yourself because otherwise you can't be sure the files will be overridden by vanilla.

A lot of it is in GlobalDefines.xml but other parts are in Civ5Terrains.xml and Civ5Features.xml

I looked a bit more into it lately and it's a bit complicated and messy. Basically, each tile has an influence cost distance multiplier. If you look at a tile, use its distance (the ring it's in) plus maximum influence cost from terrains, features and the hill modifier defined in GlobalDefines. Multiply that number by 100. Add 25 if the only adjacent tiles are over a river. Subtract the yield. Subtract -105 if the tile has any resource. All tiles with the lowest cost have a chance of being taken in the next culture expansion.

For example: A grassland tile in the third ring costs 2*100 - 2 = 198. A hill, forest or forested hill tile in the second ring costs (1 + 1)*100 - 2 = 198. So both can be chosen. You can reduce the cost of forested tiles by chopping the forests if the underlying terrain is not a hill. Ocean and coast tiles also cost more influence, putting them on the same level as hills. That is, unless they are on the other side of a river which adds a cost of 25.

Cost for land buying is calculated a bit differently, and in particular has an increased cost if you want to buy a tile that doesn't have the lowest influence cost.
 
This is good information.

I know I could probably find this information if I looked really hard, but I can't find this on the current bug list. Is there a reason I'm missing why the game devs can't make the next tile to be culturally acquired by a city manually determined? Seems simple enough and a bug in its current state, especially with how bad the logic currently is to determine.
 
This is good information.

I know I could probably find this information if I looked really hard, but I can't find this on the current bug list. Is there a reason I'm missing why the game devs can't make the next tile to be culturally acquired by a city manually determined? Seems simple enough and a bug in its current state, especially with how bad the logic currently is to determine.

I'd call it much needed feature, not a bug.

It's something I wanted in Civ4 several years ago: culture boundaries that wrap to the terrain. It makes the simple ring expansion in previous games look extremely cartoonish.
http://forums.civfanatics.com/showthread.php?t=167781

Actually, looking back at that post, it does make me think that it would be nice to have a few more ways to steer the growth beyond chopping forests. I wonder if "PLOT_INFLUENCE_ROUTE_COST=0" was an attempt to make culture more likely to expand along roads, but it either didn't get tested or there was no middle ground between too-easy-to-exploit and had-no-effect.
 
I'd call it much needed feature, not a bug.

It's something I wanted in Civ4 several years ago: culture boundaries that wrap to the terrain. It makes the simple ring expansion in previous games look extremely cartoonish.
http://forums.civfanatics.com/showthread.php?t=167781

Actually, looking back at that post, it does make me think that it would be nice to have a few more ways to steer the growth beyond chopping forests. I wonder if "PLOT_INFLUENCE_ROUTE_COST=0" was an attempt to make culture more likely to expand along roads, but it either didn't get tested or there was no middle ground between too-easy-to-exploit and had-no-effect.

I think the latter. The problem is that the algorithm is bad. What it does is always take one of the tiles with the lowest cost. What it should do is assign weights to tiles and take the one with a lower cost more likely than one with a higher cost. This would still keep a more realistic looking border expansion in grassland but also give you some hills.

Even better would be an algorithm that takes into account what the city needs most, and takes that. If the city has a low production in the currently claimed tiles, it will try to take hills or forests. If it has low food, it will try to take a river grassland tile.

I don't like choosing manually. If you want to choose, you already have the plot buying.
 
I think the latter. The problem is that the algorithm is bad. What it does is always take one of the tiles with the lowest cost. What it should do is assign weights to tiles and take the one with a lower cost more likely than one with a higher cost. This would still keep a more realistic looking border expansion in grassland but also give you some hills.

A non-deterministic algorithm would be more fun, but it also means that every now and then, it will pick an empty marsh instead of cows. If that happens on the first border expansion of the game, you'd better be able to buy the cow tile for about 20 gold, or it will suddenly be not fun.
 
A non-deterministic algorithm would be more fun, but it also means that every now and then, it will pick an empty marsh instead of cows. If that happens on the first border expansion of the game, you'd better be able to buy the cow tile for about 20 gold, or it will suddenly be not fun.

True enough. So an exception for resources should be made, which should be picked before anything else in their ring.
 
Ok, replacing this with something less...polarizing, having read the rest of the thread.

It's representative of one of my major problems with cultural border expansion in this game. To an extent much greater than Civ IV, each city is treated as its own little empire. Terrain-matching borders make sense if there's an opponent on the other side of that border. Internally, it's very strange to avoid choosing a wooded hill on a river - one of the best tiles in the game, basically forever, just because it's hard to move on.

Civ IV went out of its way to make your empire cohesive, giving you tiles between cities if there would otherwise be a one-square break between them. Civ V goes out of its way to make your empire into a collection of city states by actively avoiding exactly the sort of terrain that you want to settle near (hills, forests, rivers).

Yay, another edit: You could argue that it's like the borders between states (the Mississippi river is a great example, separating many states), but each city doesn't have its own distinct borders anyway. Once you own both sides of the river, all the annoyance of the governor refusing to spread across it in the first place was for naught.
 
In the early game you want to watch your automatic tile acquisition in your cities like a hawk (the highlighted purple ones obviously). You need to immediately quantify the field of play and determine HOW you want your city to progress before you can just go buying tiles and the like. One you determine how your city will grow you can then buy tiles appropriately. In the early game, it absolutely makes sense to buy a fish tile, a cow tile, or an oasis tile because your city will grow faster, you'll likely use that tile for the life of the city, and you'll better use your resources. It doesn't make sense to buy a forest tile to chop, a mountain tile to mine (because mines suck), or a luxury tile for something you won't have researched for another 20 turns! I see that last one all the time, what a waste of money. The only time you should buy a bunch of tiles before your city has grown is if you need a strategic resource 2 tiles away (like horse or iron early) or you plop down a city in contested territory on top of an enemy city and you need to culture critical tiles before the opponent can, effectively using your city to stop their city's growth. Obviously though, this will lead to war whether you do it to a computer or an AI.
 
You can buy tiles like gold, silver on a river or gem tiles if your city can work it immediately. In quick speed, if it's your first buying, it cost 35 gold.

But if you manage to get 2 more gold per turn for next 17 turns, well you just bought a tile which have been paid by himself and got the extra happiness faster if you bring a worker.

And like swordspider said, it worth to buy tiles to boost hammer production by growing city faster. 1 hammer=7 gold on quick speed. In that sense, if you buy a tile for 40 gold and if you build 1 hammer faster for 6 turns you begin to make hammer ''profits'' after that time. But not useful if your happiness is low. Better in the early stages when it's critical. You want to ''snowball'' the fastest possible.

@swordspider : ''because mines suck'' what? Don't worth to buy ok but you want to work them ASAP.
 
I don't see why the people wouldn't just expand in the direction I, their great lord and master, would dictate.

I would certainly prefer the river tile to the open tile, and generally prefer the mountain or forest.

If you will not let me pick my tile, how about it picking the tile most appropriate to the city's current setting? So, if I have it set to "production focus", pick the hill. If I have it set to "gold focus" pick the river tile, etc.
 
Hills and forests are more expensive because their "distance" is twice as large seeing they cost 2 moves.

So the game will always get all flat terrain in the second ring before claiming any hills or forests unless you have a resource there. It will also claim flat tiles in the third ring before hills in the second ring if they have a higher yield, for example because of being next to a river.

All in all, as I said before: Doesn't like hills or forests, doesn't like river crossings much, likes resources ;)

is there a way to force the tiles building roads?

if the movement is responsable for the choice it could be...
 
I don't see why the people wouldn't just expand in the direction I, their great lord and master, would dictate.

I would certainly prefer the river tile to the open tile, and generally prefer the mountain or forest.

If you will not let me pick my tile, how about it picking the tile most appropriate to the city's current setting? So, if I have it set to "production focus", pick the hill. If I have it set to "gold focus" pick the river tile, etc.
This would be essentially the same but with more micromanagement, so I wouldn't do that under any circumstances

is there a way to force the tiles building roads?

if the movement is responsable for the choice it could be...

No, the cost is independent of movement.
 
I've noticed that the city seems to ignore sea resources, preferring flat and useless tundra to to whales just out of reach. Does anyone else see this? I don't see why the resource modifier shouldn't apply in the water as well.
 
Back
Top Bottom