Rapid Early expansion (REX) - An Essay

[MarkG]

Well first off how would one know that they will not be attacked?

it's called diplomacy and knowing if your enemy(everyone is your enemy in a civ game ) can do damage to you

i'm saying that AI doesnt act stupid in several occasions(all AI's do). i'm just saying that some times we expect the AI to be mr. nice guy....

[vmxa1]

Originally posted by MarkG

Well first off how would one know that they will not be attacked?

it's called diplomacy and knowing if your enemy(everyone is your enemy in a civ game ) can do damage to you

i'm saying that AI doesnt act stupid in several occasions(all AI's do). i'm just saying that some times we expect the AI to be mr. nice guy....

I have got to say I do not understand this response. True in a sense we will all be enemies, but I can not treat you like one on the expectation that we will be as there are repercusions to that (at least for the human player). I do not want them to be nice, send the settlers, but they should not be able to move forward when they agree to leave. They are not doing damage to me, I am doing damage to them.
Like I said, I do not understand either of those two response. Are you saying that what I discribed is rationale behavior? If you have warriors or spearmen would you send one with a settler into Roman lands, exspecially after the last one got killed?

[Lawrence of Arabia]

I had to post the edited version here because it is over the 20 000 character limit and if i divided it up, one part would have ended up here and it would have been weird

1. REX – A definition and overview

Rapid Early expansion (REX) is when you pump out settlers as fast as you can to try to get as much land as possible. It is essentially a controlled land grab. Depending on the map size and number of civs, REX may last for a few turns, or for many.

2. Why REX?

You will want to REX because that is exactly what the AI does. A successful REX means lots of land for you. More land means of course a bigger chance of getting resources and luxuries inside, so you will be self sufficient. This is crucial for the Industrial age, where trade embargoes dominate. More land and more cities than your rivals also translates into more commerce, production, population (this is one of the biggest things counted in your final score) and land (another big thing in score.) It is also important to have a big empire at the higher difficulty levels because the AI will have production/science/and growth advantages. While you may not be able to expand as quickly as the AI, you will still be able to expand rapidly.

These REXed cities will mean that you never ever have to worry about building more cities later on in the game because you already have them all ready to go. A successful REX is one where at the end of the game, you can say proudly to yourself that ‘I never once need to trade for a resource/luxury which I did not have inside my original borders’

3. How do I REX?
a. Your First city and Settler

The best spot to found your first city is usually where you start. If there is a hill nearby, you can put your worker or scout on top of it. If there is no better land within site, found your city there. If there is a river one square over, then move your settler one square. Moving it more than two will slow down your progress and may give the AI a big lead. Depending on how many shields you take in, and which civ traits you have, you will either want to build a warrior or a spearman for defense. For example, if you are Germany (Mil/Sci), and Berlin (pop 1) produces 4 shields, then you will be better off if you produce a Warrior and then a Spearman. You send the warrior out to explore, and then you fortify the spearman. Russia (Exp/Sci), would be better off building the Spearman first in this case, because it already has a scout which can explore. A city on a flood plain with more food will be able to produce a pop three city faster, but it will be lacking in the production department. In that case, build one or two warriors (depending whether you are expansionist or not) even if you can build a spearman or an impi. You will not be able to produce as many units as a city with more production. And thus it would take too long to assemble the shields necessary to build the spearmen/impi. However, if you play the game right, you will be able to get out a second city faster then if the city had been a production center. After producing your first settler, you will then want to look for a site which is the exact opposite of your first city. Ex. If your first city is a production city, then look for a site which has lots of food.

There is only one exception to the rule of always building a defensive unit first (ie warrior or spearman) If you have set the barbarian level to sedentary or none, then you can use your worker to explore and find the closest civ. This means that if you are militaristic, your first unit can be an archer and you can try a very early archer rush. At these lower barbarian levels, I have also seen AI settlers walking around without any escort. Take your archer and take the settler even if you are not yet ready to archer rush. It will permanently cripple the AI and you will have only one city to take anyway. Its also an easy way to either add more pop to your city (by adding them) or you can use them as workers. However this exception will probably not work in MP because if a human player sees a worker walking around, he will gobble it up and thank you for giving him an easy unit. If you don’t plan on doing an early archer rush, then building a defensive unit is still a good idea.

As soon as your city reaches size three, you want your settler to be ready ideally that same turn. If not, then next turn or the turns after that will suffice. If you find yourself reaching pop three and you still have 10 turns to go on the construction of the settler, then you blew the production-food analysis. There is some timing involved and you will need to do some calculations to get the optimal REX speed.

b. What to build next

After you first settler, you are faced with a difficult decision. You could build a city improvement, a military unit, a worker, or a wonder. If you are a militaristic civ (not including Japan), you may want to consider building an archer. The archer is the basis of the archer rush (duh!) An archer rush is meant to cripple the nearest (or nearest two) civs by taking their capital with around three archers. It is a low cost campaign that will net you a city, and will knock the AI civ(s) back on their heals. This rush can also net you a worker or two, and maybe a captured settler (which turns into two workers) This more aggressive approach can also give you a couple of techs for free. Most likely, you will get all of the AIs techs because they will be down to one city and will want to survive and make peace as soon as possible. The Babylonians UU is the Bowman (2.2.1) The extra defense means that it can hold its own against other archers which another civ might counter attack you with.. They are a good choice civ for the archer rush (Even though they are not militaristic) If your neighbors are the Greeks, then you will need to take two or three extra archers because they have the Hoplite (1.3.1) which totally outclasses your attacking units.

If you do not have Warrior Code, at the end of the construction of your first settler, do not attempt an archer rush: The enemy cities will be well reinforced by then. Instead, you should then build either a granary or another unit. The granary is a good choice for a first city with high production: not only will the city improvement be completed more quickly, but it will also cut in half the amount of food necessary to increase in size. This double whammy will mean fast growth and big production. If you have Masonry, City walls are also a good choice because they are cheap and don’t cost a dime to maintain. A city on flood plains, or an otherwise devoid of production will want to crank out a military unit. It could be a warrior or a spearman or a scout. This is because any improvement will usually take too long to build (except for city walls) and thus you will not be able to build it by the time you get to population three. Scouts are for scouting (of course) but the choice between a warrior and a spearman is less clear. A warrior would be the better choice in a growth city. If you are planning on doing a swordsman rush on the AI, then you may want to build up your warriors and then upgrade them. A spearman is useful for escorting the next settler. It can protect it against barbarians. At lower barbarian levels, don’t bother with the spearman because you wont need any protection. Build a warrior or a scout.

Wonder and workers should not be built until your reach found your third or fourth city. Note that I said found, and not captured. If you have captured a city using the archer rush, don’t count it in your total city analysis. This is because there will be all sorts of messy stuff like resistors and unhappy people. That captured city will need to do things on its own for a while (temples, units for martial law) Building a worker or a wonder in your first or second cities will slow down and may cripple REX.

c. Your second city

The placement of your second city is very crucial to your success. A good placement will speed up your REX and a bad placement will stall your empires expansion.

If you have discovered another civ and/or archer rushed their capital, then you should expand in that direction. The next city site should be within 3-6 tiles of your capital. Any further, and there will be gaps in your line which the AI will exploit (ie plop cities down in your territory) The second city site also needs to be an exact opposite of your first city: If the first city was a production center then make the second one a food center and vice versa. This is because these sites complement each other: A slower growing city can pump out more units, and a faster growing city can pump out more settlers. The second city need not be on a coast, Two production cities or two growth cities will result in either a lack of settlers or a lack of units. Look for luxuries and resources which will help you support larger populations later on.. Your first settler was most likely not guarded. That means that your new city will need to build either a warrior or a spearman to defend. If the city is now a food city, build warrior. If not, build spearman. When its complete, switch to settler. If you decide to archer rush, after the settler you will need to build an archer.

d. Third and Fourth cities

If you do not have a city on the coast, then your third city must have salt water lapping against its shores. This coastal city will let you build galleys later on and is an opening to a whole new world: the sea. Harbors in these cities will get your trade network up and running quickly. The first thing this city should build is a spearman. I usually have my third city build a wonder. In my case, it is always the Colossus. This means that the city will need a good production site. I try not to build wonders in my capital, because at the start of the game, it is the dominant force of your civ: building something like a wonder would slow down development. Since this city will become big, you will need to get some entertainment in here. Bringing over a worker to mine the tiles will vastly increase your chance of getting the wonder which you build. Unlike the second city which should be built in the direction of the AI civ, this one should be at the nearest coastline. Placement should also be with in 3-6 tiles of your capital.

The fourth city is usually your first bid for resources. By the time your crank out your fourth settler, you will have researched/traded for either Iron Working or The Wheel. If none of your cities have these resources within their borders, this city should have it within its immediate nine square border. If you have only one AI civ nearby, then you will want to also place this city in that direction. However, if you discover another one, then the fourth city should be built in that direction.

d. The Fifth city and beyond

Your first four cities are now your core cities. By now, your capital and maybe your second city should stop producing settlers and start building either workers or improvements. Workers can get your resources to all of your empire, making your swordsmen/horsemen rush that much quicker to assemble. Now you have two choices: you can plan on going to war to take over the whole continent or to take a large portion of it or you want to wait until later (middle, industrial, modern ages) before you unleash your hordes.

If you plan to do the former, you will want all of your cities to be in good sites. This is because they will be able to produce military units faster and will grow without much micromanagement. Let the AI build their cities in the desert. In a few turns, your horsemen/swordsmen will have conquered them. In this way, you do not waste population points by founding cities in locations where they won’t grow. These badlands are also characterized by resources. Since you are attacking with units which don’t need saltpeter, or oil, you will have no need to try to acquire them from plains, deserts, tundras. Just settle the most fertile lands and take the AI cities and acquire those lands before you research Gunpowder. Like your core cities, these will all want to build barracks, spearmen and maybe temples/libraries/walls if you have the time. Settlers are also an option and should be built every time the city gets to pop three. (again by using a method of calculating.)

The only exception to this rule is if you see a luxery good that you don’t yet have, and it is only 2 or 3 squares into unfavorable terrain. If this is the case, then you should put a city there on that site. If however the resources is deep in the Amazon, don’t bother. Another exception is if you start in the middle of the Sahara and you need to REX your way out of the predicament. You should slowly build cities in the bad lands in one direction until you get out of it to more fertile land.

If you don’t plan on going to war for quite some time (ie before the advent of Knights and/or cavalry) then you will need some cities in the badlands.

These cities are put in as a way to limit another civs expansion. When you are not planning on going to war, you will need to make sure that you have the needed resources inside your borders for future wars. In this case, there is not need to look for premium territory to place these cities in. You can put them in deserts, jungles, tundra, plains, wherever the AI civs are, you put them in that direction.. These cities will fill up the map with your color and will increase your chances of getting unknown resources within their limits. For these fringe cities, you will want to have temples and culture producing elements to expand your borders and keep the AI cities away. Later, some of these cities may become productive and net you lost of money and growth and producing. In the short term however, they are there to advance the cultural line of your civ and to make sure you are not dependant on other civs for resources.

Special Thanks to Thesus

e. Back to the First city

So, after you have completed the initial military unit, the settler, and maybe archers for your archer rush, you will need to build some more settlers. Rinse and repeat. When your fourth city gets built, stop building settlers in your capital. You can switch to either workers, or wonders or improvements. The workers will improve your land, connect resources, and increase production. I would not build a wonder in my capital city this early because it may stunt the growth of the entire civ. However, if your second city is in a very good site, then you may wish to start building a wonder. For me, city improvements, and units along with the odd worker are the way to go. You should also build city improvements which match your civ ability: Militaristic civs can build barracks, religious can build temples. A temple is good because it will let your city grow larger without getting unhappy people. If you are planning a swordsmen or horsemen rush, then barracks will give your troops extra training and make them veterans. Either way, you need a goal, and you need to follow up on that goal.

4. The End of REX

Ending REX is just as important as staring it. You mush now when the transition phase happens.

The end of REX can manifest itself in a few different forms. The first is when your run out of space to expand into. You have built your cities all of the way to the AIs borders, and you are surrounded on one or more sides by water. When this happens, or as it approaches, you need to have less and less cities building settlers and more and more building units, workers, and improvements. With the new territory that you have acquired, a decent road network and lots of resources linked up, this is the time to go and decisively cripple a nearby civ. If there are two civs nearby and you only archer rushed one of them, then your should swordsmen/horsemen rush the other one. From these wars you can gain valuable workers (to free up city production,) more land, more population, and leaders.

If for example, you have run out of room expanding in one direction, have over the forbidden palace city trigger for the map, but in the other there is loads of land waiting to be claimed, but not AI civ in site, you should not REX them. The last city that you build should build a settler and send him in that direction. Then switch production to something else. In this case, only your last city founded is building settlers, while every other city is gearing up for war/city improvements.

5. REX vs. Culture

The opposite of REX is to build a few cities, and then build up their culture. When the AI cities lap against yours, they will inevitably culture flip to your side. This method will initially have you will few cities, but high population. On higher difficulty levels, where there are less and less content citizens, this will mean lots of entertainment if you are not lucky enough to have a few luxuries in your border. This method is very isolationist and passive. It is also a not very effective way to expand because it is slow and you are not in control of when the AI cities may flip. Some people may argue that with a higher population, you will have a higher production and more gold, and will be a superior force to smaller less developed cities. Each individual city will have more of these things, but on a whole, more cities will have more production and gold. This means that you can build more things at the same time. A culture strategy will get you quicker production, but your will not be able to build as many different things simultaneously. At higher difficulty levels, the AI will also build up a killer lead. By the time you have assimilated a couple of his cities, he will have built up such a horde of units that you will not be able to get an advantage on him. By REX-ing and archer rushing, you will be able to cripple him by cutting off his land, and by taking his cities.

An even worst case is if you don’t REX and you don’t culture. This is the worst of possible cases. You don’t have any land, so the possibility of getting any resources is minimal. You don’t have any culture, so you won’t be able to assimilate cities. You will have most likely quite a few units which you might be able to use to acquire more cities. However, if it wasn’t an archer rush, then you will have a hard time later on. You might not have any Iron or horses, so you will not be able to produce any horsemen/swordsmen. The larger amount of military units that you may have can net you a few AI cities, but the collective production might of a larger AI empire will stall your advance, and may even cripple your game forever.

[Lawrence of Arabia]

6. REX on larger/smaller maps, and difficult settings

Depending on the map size, there will be a large degree of variation within the REX phase. These variations can alter the very strategy which you will choose for your victory.
a. Small Maps

Small maps (being defined as Tiny or Small) have a very short REX period. Typically, this will include the formation of 4-6 cities before you run into the AI in all directions. This means that your cities will need to have perfect placement. This shorter REX also means that building a granary is a waste of resources because by the time you have spent the shields and gotten to use it, the expansion period will be well over.

Archer rushes are very effective on these small maps because the civs cities are very close to each other. By knocking out a civ early, you have doubled the space you have to expand into and you have prevented the build up of a killer AI.

b. Large Maps

Large maps (being defined as Large or Huge) are a whole different beast to contend with. The distances are great, the resources spread out, and the number of potential killer AI is big. In order to REX successfully, you will need to do things a little big differently.

Large maps entail the need for granaries. This is because the distances are so vast and the AI so plentiful, that you will need to get your settlers out there as quick as possible. Since you will be founding lots of cities, the granaries cost will be double paid off by the end of the REX period. These grannies are also necessary to keep pace with the AI in terms of city founding. Cities with granaries on (shielded)(mined)grasslands with special resources (like cattle or wheat) are called settler factories. You should have at least one of these which will crank out settlers at an impossible rate.

Archer rushes are not very effective on these maps. This is because the distances between capitals is huge. In this case, the resources spent on making those military units can be better used on exploring units (scouts or warriors) These units will help you uncover the map quicker, get goody huts, and be able to found cities in premium sites. Because of these distances, horses and a good road network become much more important. To get to the far reaches of your empire quickly requires mobility. Both roads and horses have the ability to get there ‘the fustest with the mostest.’

Large maps will also have longer REX periods. Unless you are an expansionist civ, the chances are that you will not have found another civ by the time you get out your first settler. In this case, don’t worry about the direction of the nearest AI civ. Just place the city in an optimum position. On lower barbarian levels, you will not need to build as many defensive units because there is no risk of any attacks. These resources can be used instead on city improvements.

c. Higher Difficulty Levels

On Emperor and Deity levels, the AI gets huge bonuses. It is therefore very hard for you to keep up in anything except for commerce. In these cases, it is very important to optimize your production capabilities and not to waste a single shield/bushel.

To keep pace with the AI in city expansion, you will need two settler factories on Emperor, three on Deity. However, this also depends on which civ is nearby. Some civs build both a spearman and a settler before they expand. This plays into your hands if you build only settlers. It negates the effects of their bonus.

Special Thanks to Anglophile

7. When not to REX

Well, I can’t really think of a time when you wouldn’t want to REX. The only instance would be if you are stuck on an iceberg all alone, with raging barbarians, sixteen civs, on deity. Who would ever be crazy enough to do that? In every other case where you start in a bad spot (like desert, or jungle) you can REX your way out of the bad predicament to more fertile and productive lands. Archipelago maps are not suited for REX and it should not be tried in those instances.

8. The Dangers of REXing

When REXing, it is important to build more than just spearmen and settlers. If you only build those two units, you will fall pray to a better prepared civ. You need to remember that archers and swordsmen must be built, if you do not want to fall pray to an AI. This is doubly important in MP where a craft opponent will use his resources more wisely than the AI. To survive, you too must diversify what you build. However, if you crank out a few military units, and never have more than five cities building settlers at a time, you should be OK. If the AI or player does send a few units to try to catch you in REX mode, you will dispatch them easily. The loss of their units and their resources invested will set them back. You should have no problem overrunning them. Also remember that the AI expands using just Spearmen-Settler stacks. This means that they will be stretched thin, with not a lot of units in their cities, and no offensive units. Now is the time to strike, general.

Special Thanks to Arrian.

[Vlad Antlerkov]

On behalf of Lawrence, here's the whole thing in Word format (all 8 pages of it). Not much in the way of formatting.

[Lawrence of Arabia]

Hey thanks a lot Vlad.

[vmxa1]

Thanks Vlad and LoA, now I can read it at my leisure.

[jshelr]

Dumb question -- how do I find posts 32 and 33?

[vulture]

Originally posted by jshelr
Dumb question -- how do I find posts 32 and 33?

\

Dumb answer -- count them.

More sensible answer -- look for two long posts by Lawrence of Arabia some way into the thread, that look as though they could be an essay on the REX strategy.

If you haven't changed the default setting for displaying posts, there are 25 posts per page, so it will be near the start of the second page of the thread.

[jshelr]

Thanks for the help.

[Tolwyn]

anything changed because of c3c?

[Arrian]

Well, actually, I don't typically do what is suggested in the REX essay above.

I often go warrior-warrior-warrior-granary in my capital, and then pump settlers as best I can. If the city has +5 food, great, it can be set up for a 4-turn settler pump, but often it isn't +5 food capable, so I end up mixing in other builds inbetween settlers. Often archers for barb hunting.

As I expand, certain cities will get specialized... high food ones become the population pumps, and higher shield cities build barracks and get to work building an army (how much of one depends on many factors). I may devote 1 city to a wonder, though this is typically a smaller one like the Colossus or Great Lighthouse.

Changes in Conquests... well, the AI definitely seems to expand strangely, when compared to PTW or regular CivIII. It used to build out in rings spaced 4 tiles out from its capital. Now it seems much more haphazard, and wider. You may find AIs plunking down cities very close to you early on, even though land right next to their capital remains unsettled.

Other changes that effect REX that come to mind:

1) the reduction in power of the industrious trait (instead of being 100% faster, industrious workers are now only 50% faster. Additionally, slaves of industrial civs no longer work faster than other slaves. They all work half as fast as non-industrious workers).

2) forest chops have dropped from 10 turns to 4. Big change.

-Arrian

[Barinthus]

Wow, great essay. It's now part of my required readings for Civ 3

[Aqualung71]

I'm with Arrian....the REX theory overall is very sound, but implementation differs. I have learned to delay the building of my first settler until I have 2 warriors (happiness police!) and a granary in place, with a third (and maybe fourth) warrior scouting for contacts. I would never send a worker - you need him to build roads.

That way you get your pop level up to a critical mass of 4/5/6 and can spit out settlers more quickly.....of course it helps enormously to have bonus food resources.

[Hermann the Lombard]

Wonder and workers should not be built until your reach found your third or fourth city . . . Building a worker or a wonder in your first or second cities will slow down and may cripple REX.

OK, I saw Thriller's approach and AFAIK Arrian has never been a REXer...so I have another question about the "state of the art." Is it wise to sneak an early worker in there (i.e. warrior-worker-warrior-settler) to get the road net growing to expedite the settlers and the rushers or defenders? In the unusual case of AU501 I started with a worker (worker-warrior-warrior-settler) as many did to pop that hut with the culture boundary while I had no warriors in play.

Another exception is if you start in the middle of the Sahara and you need to REX your way out of the predicament. You should slowly build cities in the bad lands in one direction until you get out of it to more fertile land.

I've seen this before, but at the time I didn't think much about it. Why spend the resources to build those intervening cities in lousy terrain? Why not send the escorted settler to the fertile land? The slow approach seems the antithesis of rapid expansion.

My first thought was that the idea was to avoid having the AI plop cities down in the desert gap, but I seem to remember (Vel Part III?) a situation where the capital was at the end of a crummy peninsula, and the second city was built up at the neck of the peninsula for an eventual palace jump. Yes, the AI colonized the gap, but that city was picked up by culture flip.

My second thought is simple: "What if the AI beats me to the fertile ground?" In that case having a bunch of small REXed cities might produce enough units to TAKE that city. Am I on the right track?