ICS, rush... a few definitions

[Trias]

1. Infinite City Sprawl (ICS)

Having a double or triple amount of bases/cities from normal, but occupying the same amount of space on map. For example, if a regular game board of 10x10 tiles would be designed with 2 cities in mind, you'll have 6. This strategy relies heavily on "per city" bonuses and base tile yields while trying to negate the penalty for having that many cities. Civilization 5 has no ICS, because both the AI and human player are building cities at 4-hex (or more) distance from each other.

- filling every possible spot on map with your borders is not ICS (although it's possible with ICS)
- placing cities as fast as you can is not ICS (although ICS can embrace such strategy)

This seems like a somewhat degenerate definition of the term, as it has been used for years in the community. Originally, ICS refers to the ability to build as many cities as you want without being (effectively) checked by any game mechanism (other than map real estate). Generally, this is viewed as a problem with the game design.

 

[Windsor]

To me, ICS is all about distance between cities. CxxxCxxxC is ICS in Civ5. It does not mean that you settled those cities fast or early, only how closely you have placed the cities.

 

[Camikaze]

I think there is a cross over between REX and ICS. ICS doesn't necessitate close city spacing, just a continuance of REXing, which is a short term strategy.

 

[Hail]

I think there is a cross over between REX and ICS. ICS doesn't necessitate close city spacing, just a continuance of REXing, which is a short term strategy.

REX and ICS are related in the early game. however ICS does necessitate close city spacing.

the whole point of ICS is to build the maximum number of cities. as in a city has value in itself. if the city has little value in itself(the city tile's yield is low and/or few civic/policies grant bonuses on a per city basis), ICS strategy looses it's strength

 

[Camikaze]

You can REX to both pursue a dual strategy of ICS and landgrabbing, though. You want to deny your opponent a foothold, then you'll REX but still reap the benefits of ICSing as a result (closer spacing allows you to gain the maximum out of ICS, but there is minimal difference between the benefits of 10 cities spaced close together and 10 cities spread further apart).

 

[Hail]

No. ICS is a goner from Civ5 because the minimum and optimal distance between cities is the same.

gone? what?

increasing min distance between cities in civ5 has only direct one effect: decrease city density. the player still wants to have as many cities as possible

now REXing to take land and deny it to others has increased in importance. that's it. everything is the same. still, since 20 Sept. 2010, a new city is a net positive.


dude! you really have no clue, what you are talking about

You can REX to both pursue a dual strategy of ICS and landgrabbing, though. You want to deny your opponent a foothold, then you'll REX but still reap the benefits of ICSing as a result (closer spacing allows you to gain the maximum out of ICS,

uh?

REXing does not imply packing cities as close as possible.
one can REX to pack cities(ICS) or he/she can REX to take good spots(maybe farther than minimum city distance).

[...]

but there is minimal difference between the benefits of 10 cities spaced close together and 10 cities spread further apart).

just

 

[Camikaze]

uh?

REXing does not imply packing cities as close as possible.
one can REX to pack cities(ICS) or he/she can REX to take good spots(maybe farther than minimum city distance).

Exactly.

 

[Trias]

This seems like a somewhat degenerate definition of the term, as it has been used for years in the community. Originally, ICS refers to the ability to build as many cities as you want without being (effectively) checked by any game mechanism (other than map real estate). Generally, this is viewed as a problem with the game design.

To add.
If map real estate is the main constraint on ICS, then one can maximize the benefit of ICS by packing cities as closely as possible. Close city packing, however is not a defining feature of ICS.

 

[_hero_]

No. ICS is a goner from Civ5 because the minimum and optimal distance between cities is the same.

In Civ5 there's only REX. In other games, a REXxer can ICS or an ICS guy can REX, but its not neccessarily mutually inclusive.

It's more than just a function of distance. A REXxer would avoid putting a city in a spot where there's no valuable resources until all other spots have been filled. An ICSer would not. A REXxer only puts cities at optimum locations. An ICSer puts them in all locations.

For example: If there is a little batch of forest tundra on the corner of your empire, and ICSer would place a city there, and a REXxer would not.

ICS is still possible, it's just no longer holds any advantage over REX.

 

[Bibor]

gone? what?
increasing min distance between cities in civ5 has only direct one effect: decrease city density. the player still wants to have as many cities as possible
dude! you really have no clue, what you are talking about

Whoa, wait a second there. Think through before you post, man. Literally every strategy is based on getting all possible land/resources/bases/cities. Sure, some land or bases may be more important than others, but every single strategy sim is about that. Sure, there's always counterbalance like happiness in CIV5 or Efficiency in SMAC or new base expenses in CIV4, but in the end cities always end up generating more than they cost.

Having a minimum distance equal to basically city work-radius is definitely not ICS. Not because it wouldn't be a valid definition in itself, but because nobody places cities 6 hexes apart. The generic placement pre-patch was already a 3-hex gap, to a maximum of 4. Extreme cases on archipelago maps or a single terrain blocker are just that, extreme cases based on a very specific map.

If the generic placement equals ICS placement than there's no ICS placement.

 

[Zyxpsilon]

To me, ICS is all about distance between cities. CxxxCxxxC is ICS in Civ5. It does not mean that you settled those cities fast or early, only how closely you have placed the cities.

My argument exactly back in November+December when i was extensively testing Dale Kent's "Solution Mods" which lead us to fairly drastic conclusions on optimal tile accessing within any given grids. Tall or wide.

 

[Zyxpsilon]

REXing does not imply packing cities as close as possible.
one can REX to pack cities(ICS) or he/she can REX to take good spots(maybe farther than minimum city distance).

Yet, extrapolating a resulting inherent growth provides number of cities which aren't tightly small for having shared tiles to assign in one only.
Thus, distance packs a REX punch.

 

[Zyxpsilon]

The generic placement pre-patch was already a 3-hex gap, to a maximum of 4. Extreme cases on archipelago maps or a single terrain blocker are just that, extreme cases based on a very specific map.

It is NOT 4 in v1217, it is 3 away from the closest city. (OP, is Minimum distance between cities)

GlobalDefines.xml value as in MIN_CITY_RANGE=3

You are including the prospective city tile yet to be built on... which is misleading and inaccurate.

 

[Trias]

Whoa, wait a second there. Think through before you post, man. Literally every strategy is based on getting all possible land/resources/bases/cities. Sure, some land or bases may be more important than others, but every single strategy sim is about that. Sure, there's always counterbalance like happiness in CIV5 or Efficiency in SMAC or new base expenses in CIV4, but in the end cities always end up generating more than they cost.

Having a minimum distance equal to basically city work-radius is definitely not ICS. Not because it wouldn't be a valid definition in itself, but because nobody places cities 6 hexes apart. The generic placement pre-patch was already a 3-hex gap, to a maximum of 4. Extreme cases on archipelago maps or a single terrain blocker are just that, extreme cases based on a very specific map.

If the generic placement equals ICS placement than there's no ICS placement.

Sorry, but you are simply wrong. ICS is about the number of cities not the city density.

 

[PieceOfMind]

It seems, Bibor, you've opened up a new debate about the definition of ICS. Personally I too disagree with your definition and would describe in a way more like how Celevin did. ICS is more about placing cities in as great a number and as densely as possible. ICS after all is not called the "tight-packing of cities" strategy. It's just infinite sprawl of cities, which I think of as the name of the strategy in which you wan to build cities indefinitely, never wanting to stop, where each city usually creates a net benefit to your empire. Usually to maximise the benefits of ICS one would choose to find the most efficient packing of cities but it is not a necessity. Small adjustments to make cities take better advantage of the terrain does not remove its ICS name.

I wouldn't go so far to say that ICS has been eliminated, because ICS has never been simply about the fact that city fat crosses would overlap. Certainly the extent to which ICS could be exercised has been throttled back by the dist-3 settling rule change. It could also be a less attractive strategy as a result of the change (because the number of per-city benefits you can enjoy is now lower considering the number of cities you can pack into the same space is lower, assuming that an empire's expansion is more limited by the land available rather than the speed in which it can grow).

I completely agree on the distinction between ICS and REX though, and that the two are often confused in discussions. I made that mistake a few times as well.

 

[Bibor]

I wouldn't go so far to say that ICS has been eliminated, because ICS has never been simply about the fact that city fat crosses would overlap.

Quote from CIVIII guide for city placement. It also applies to SMAC. Ergo my logic dictates this definition should be valid for Civ4 and Civ5 as well. I see nothing wrong in calling it "Tight placement". After all, maybe a future game will allow for a real ICS again.

Spoiler :

I. Placement Patterns

Before we can even discuss the pros and cons of each regular placement pattern, lets go over the following terms.

C = city, x = tile in between cities

Optimal City Placement (OCP):

Usually CxxxxC.

Each city's goal is to not overlap, and use the full 21 tiles.

You need a hospital to grow past size 12 and use all tiles.

Loose Placement:

Usually CxxxC.

Cities try to gain around 14-16 tiles. Some overlap is fine.

You don't need a hospital, but it helps if you want to use all 14-16 tiles and get the benefits of metropolises (more unit support, etc).

Tight Placement:

Usually CxxC.

Cities gain around 6 tiles. Overlap is very common.

No hospital needed.

Infinite City Sprawl:

CxCxC.

Cities gain 4 tiles. Overlap is rampant. Cities will usually not grow past size 6 .

Civ3, SMAC:

City work area: 
--C--
Minimum gap:
-C

Civ4:

City work area: 
--C--
Minimum gap:
--C

Civ5 (old):

City work area: 
---C---
Minimum gap:
--C

Civ5 (new):

City work area: 
---C---
Minimum gap:
---C

 

[Drawmeus]

ICS isn't just about "every city is net benefit so build as many as you can" in a vacuum. It's about "every city is a net benefit even if it doesn't work any useful terrain, because the city tile itself is worth more than the costs."

You can tell it's an ICS strategy is when you answer yes to the following questions:
Would you fill a desert with no tile yields, no strategic importance and no resources with cities?

CONSEQUENCES of ICS are that you want cities as densely packed as possible (to maximize number of cities per land area) and you want more cities even when the terrain is otherwise valueless.

 

[PieceOfMind]

As I edited into my post, strategies on employing the most efficient packing of cities into a land area are used for ICS but are not themselves ICS. The article you've quoted, while substantiating your argument fairly convincingly, IMO just does not set the definition for what ICS is.

If ICS were intended simply to mean "tightest possible packing", then why wasn't it named something to better reflect that? And even if it were defined that way, whether the min settling distance was 1, 2, 3 or whatever, it wouldn't exclude the possibility of ICS.

In civ1 IIRC you could place cities literally next to each other (distance of 1). If ICS existed in civ1 (I wasn't really hanging around civfanatics at the time), did it involve people settling cities so densely as the rule allowed? My guess is no.

 

[Bibor]

In civ1 IIRC you could place cities literally next to each other (distance of 1). If ICS existed in civ1 (I wasn't really hanging around civfanatics at the time), did it involve people settling cities so densely as the rule allowed? My guess is no.

But in CIV3 and SMAC they did. That's the point. There was a very clearly defined strategy that used ICS with maximum overlap. Let me make a screenshot of a sample game I downloaded recently.

 

[PieceOfMind]

ICS isn't just about "every city is net benefit so build as many as you can" in a vacuum. It's about "every city is a net benefit even if it doesn't work any useful terrain, because the city tile itself is worth more than the costs."

You can tell it's an ICS strategy is when you answer yes to the following questions:
Would you fill a desert with no tile yields, no strategic importance and no resources with cities?

CONSEQUENCES of ICS are that you want cities as densely packed as possible (to maximize number of cities per land area) and you want more cities even when the terrain is otherwise valueless.

This is all true, but even when employing the strategy of ICS one can't ignore opportunity cost. For example, when you ask the question "would you fill desert areas with cities?", really you should ask "if there are no better places to settle, would you fill desert areas with cities?"

I'm sure you probably implied this (particularly in your last paragraph), but I'm just mentioning it nonetheless. But with this in mind, one could still call a strategy ICS if first the strategy is to grab all the best city sites and then fill in all the possible gaps between those cities. This is probably how ICS was most often practiced in civ4.