Faster Early Expansion

Excilus

Chieftain
Joined
Nov 26, 2001
Messages
25
All or most of you are civ2 veterans, and you are all scraching your heads as to how the ai expands so quickly. Even on regent, with even odds, the ai manages to get culture buildings, spearmen to go along with thier settlers, wonders, and out expands you!

Even when you try, for instance, to produce a settler as soon as you get the chance, in all your cities, you are beaten. What's the problem?

The sad fact is that you've all been doing it the wrong way. In order to build a settler you need 3 pop points. It takes 20 food to go from 1->2, and 40 food to go from 2->3, or 3->4, or 4->5, or 5->6. This means you grow *faster* with more population, as long as you have the terrain to suit it. Often, the build order for most of your towns, and espacially your capital, is warrior, settler, settler... when you are trying for fast expansion. This should not be done.

Behold, the build order: (Note, if you are expansionistic build the scouts, otherwise ignore them. A second or thrid scout really helps in finding goodie huts, city positions, luxuries, rival civs)

Captial Note: If you go the peacful route, don't even build a barracks in your captial. Your captial should be getting as many wonders as possible, and should be built near a river since you can't make settlers while making a wonder.

Warrior(First should be scouting), Scout, Granary, Settler, Warrior(Fortify), Settler, Worker(Build Mines), Scout, Settler, Spearman(Fortify, send warrior scouting), Warrior, Settler, Worker(Build Roads), Wonder.

First Couple Near Captial Cities(Main Cities):

Warrior, Granary, Settler, Worker(Roads), Warrior, Settler, Spearman, Settler, Spearman Temple, Worker.

Border Cities: Note: If you have an extra warrior/spearman(You should), stick them in the border city and skip building the spearman.

Temple, Walls, Spearman, Worker(Mines or roads), Granary, Worker, Library(Or, you don't have it yet, skip to next), Spearman, Worker, Settler(For either any uncolonized bits of land, or reserve for a galley).

There are also a number of key wonders for the early era:

First and foremost is the lighthouse, getting this may mean a bunch of footholds on continents or islands that no one else could get for ages.

Second is the great library, having this means you can specialize up a tech tree, or have loads of money for republic or monarchy.

Third is the pyramids, because this drastically shaves time off the border city build order(You won't have time to finish this for your main cities)

Finally the Colossus, because it is quick to build and quite effective for a very long time. Super Science Cities are MUCH MORE POWERFUL if you can get them in civ3, becuase everything else is corrupted to all hell(Normally your SSC is your capital, or near it, or has the forbidden palace).
 
Thats great - very useful thanks. i tend to neglect granary's. The forbiden palace is sooo underestimated, but it can turn a huge corrupt city into a powerhouse.:goodjob:
 
One way to expand quickly early in the game is to hurry production of your settler. Under depotism your are able to do this without any extra cost.

When a city reaches three, chose to build a settler and then choose to hurry production. this will get you a settler almost immediately without have to wait xx turns for the city to produce it.

The drawback is that you lose population points sometimes 2 somtimes 1. But you are going to lose at least one when you build a settler anyway. The other drawback is that it can make your citizens unhappy. But early in the game it is better to pump out settlers and grab as much land as you can and worry about happiness later.
 
One way to expand quickly early in the game is to hurry production of your settler. Under depotism your are able to do this without any extra cost.

When a city reaches three, chose to build a settler and then choose to hurry production. this will get you a settler almost immediately without have to wait xx turns for the city to produce it.

The drawback is that you lose population points sometimes 2 somtimes 1. But you are going to lose at least one when you build a settler anyway. The other drawback is that it can make your citizens unhappy. But early in the game it is better to pump out settlers and grab as much land as you can and worry about happiness later.
 
Does the hurry settler thing only work on easier difficulty levels? I lose the population as soon as I hurry production, and when I don't have the population to produce the settler next turn, the city just has to wait for me to grow to 3. Probably how it should be though, the method you described sounds like an exploit.
 
It doesn't matter. It's not the fastest way to expand anyway. Like I said in my post, you grow faster, in addition to producing more and getting more commerce, with higher pop. Even when doing the despot rush strat, you want a city at size 4-5 first.
 
Good on you for posting your ideas if you think that they will help others. I agree in general with much of what you say but I think you may be over-emphasising the necessity for that particular build order.

There have been a number of posts before this where people put forward their “best build order”, but I can’t help thinking that it’s a rather inflexible approach. If a new player is having trouble matching the AI’s build rate then I think they would be better spending some more time trying to grasp the mechanics of optimum city building, rather than just adopting someone else’s formula.

There are a lot of factors affecting growth speed – civ specific bonuses, city square bonuses, the terrain you started on, etc. It seems to me that the big issue to come to grips with is learning how to balance your shield production against your food harvest without wasting either.

As you say, a classic newbie error ican be to set production for settler, settler, settler, and then fail to notice that the ‘turns left’ number often races to 1 and then stops while the city population catches up to the necessary 3. This situation does waste potential production and lets the AI get ahead. At least it “usually” does – but depending on your starting terrain it can actually work out OK. (see example in post below).

Of course, not all situations pan out that way. Some starts give you a good crop of shields and a more modest food potential, and some will be the other way round. I think that you need to be able to alter your build order to suit what you've got to work with, and what your overall goals are.

You also list Granary as your second thing to build (if you don’t have a scout). Unfortunately, if you don’t have a scout then you won’t have pottery as a starting tech either (they are both expansionist civ bonuses) – so you won’t able to build a granary until you’ve researched it.

My guess is that people who can’t keep up with the AI might benefit from a bit more reading on shield/food/commerce relationships and testing out different civs and terrain types rather than just settling for what works for the way someone else plays. The better you understand the mechanics the more creative you can be with your build order.

The beauty of Civ3 is that it throws up so many different scenerarios at you, so it isn’t really a “formula” game. It rewards the player who can respond flexibly to different challenges.
 
OK, I have just tested two situations where I started as Russia on terrain that had very little shields but good food (including) a 3 food square. For the sake of the test I parked the worker and free scout and let the computer choose which squares to work and began building settlers immediately (no warrior). The numbers worked out beautifully and I got my first settler exactly on cue. From then on it pumped settlers without wasting either shields or food, exactly as the pop hit 3. I stopped when I had 5 cities (who were pumping out warriors - just to make the test easier to judge). I also had one city make a granary just to keep the test a fair comparison.

The next run through I tried your build order (again having the new cities make only warriors for the sake of the test).

The outcome was that the “wrong” method did better than yours. By the same year, they had more settled cities and more overall population. Somewhat surprisingly, the warriors came out equal.

The second attempt (in a different site – mostly grassland, but also with a 3 food square) produced similar results up to 5 cities, although much closer. The only difference being that the “wrong” way ended up with 1 more population and a better science research result (the slider had to be dragged back to zero for a while when the treasury went bankrupt near the end of the granary build).

In both cases the difference was pretty modest, but it did seem to prove the weakness in “best build order” theories. Particularly as the strategy used was your “never do this” one. (And I would have avoided it too).

Not trying to pull you down here (in most cases I think your order would work fine for people who want a "default" option that will work well in most cases, until they feel more confident about using more unusual orders). I'm trying to use this as an exercise to improve my own knowledge (I actually started out testing on the assumption that your method would beat the other one, and it was sheer luck that dished up a starting position where it didn't seem so appropriate). :)
 
Good point Polonius.

Pottery is one of the cheaper techs..but I do not go for it earlier.

This is still good info on growth. I usually build warriors until my pop hits 2 close to moing to three..then go for the first settler. I usually just bide my time by exploring. I will have to look into this though using expansionist civs. This could be the huge bonus of pottery (except being closer to getting mapmaking). It is worth a try. It is easy to ignore granaries..thanks.
 
I agree with what Polonius has said. I don't believe that there is an absolute optimum build stratety as factors such as civ, game size/difficulty level, starting position, etc can't be ignored. My starting strategy is to build specialist cities based almost entirely on nearby bonus resources. Because one spends so long in Despotism in Civ III, mining and irrigating really only work on bonus resource tiles (and flood plains). I always play expansionist and my first build is always a scout. The two scouts first find nearby prime city locations then as many goody huts for technology as possible. My second and third cities are always settler farms, one, preferably two irrigatable bonus food squares and pop goes from 1 to 3 in the same time that a settler is produced. Cattle are best for this, with wheat, growth is faster than building.

I look for a site with two game tiles for my unit producer (only tile with two food/shields under despotism). Another of the first six cities is my worker farm, one irrigatable bonus food resource and pop 1 to 2 is matched by worker production. The next two cities are my Wonder/Science cities - all rounders with long term food and shield (i.e. hills) potential. My capital fills whatever of the above roles that it's starting position best suits it for. In the game I am currently playing (Monarch, huge, 9 civs), my capital produced a scout, two settlers for my settler farms then 22 consecutive workers until I became so ashamed of my empire's capital not even having a temple that I made another city my worker farm and started building some infrastructure in London.

This system will keep pace with the AI early, you will need to add other settler farms later on bigger worlds to continue to keep pace. One also has to add unit producers steadily to maintain an adequate defence force - or offence force. Obviously, this system is also dependent on starting position, since one needs to find some good city sites. About all that I can say about that is that on huge maps where expansion by conquest early is really difficult because of the distances between civs, if you don't get a good starting position, you better be good at playing catch up through the mid and end games because you will be out expanded
 
Thanks for a great post Anglophile. Apart from the fact that I've just gotta love a post that starts "I agree with what Polonius has said." - there's a whole bunch of well fleshed out tips in there that I will try out next time around. :)

I just love the start of the game, when the board is clear and everything rushes along without all that tedious waiting while 10,000 little men run around the map. So I can sometimes get a bit carried away and lose concentration on optimising all my placements, city goals etc. It certainly does pay to nut it out early though, and tidy things up a bit before the other civs pop out of the fog and distract me. :crazyeyes
 
Well, polinus has brought up quite a few points, but the goal of my build order isn't higher pop points, it's faster expansion. As you can see, it's geared towards exponential expansion. The settlers are made, then warriors to defend, then spearmen to defend the city, a setter, then the city's fortified warrior is sent out. The scouts/warriors find the best sites and goody huts, and your capital is BIG and can get 1 or 2 wonders quite easily, or produce units every 3-4 turns.

You end up with much better defense, and because there a granary there it grows MUCH faster. Since you've already got a fairly larish army you can coerce your way to huge amounts of money, and pillage everything with your warriors. Add to the fact that you have much better city positions and you have a killer start-up.

Another thing is that, if you don't fancy scouts, just leave them out. You'll get everything else done faster. The most important thing to build is the granary, a granary-settler-settler... is the fastest city roduction scheme you can muster.

At no point did I say my build order was the best, nor that there was one. I posted one that did a pretty good job, and has shown to be a great deal better, infact then the settler-settler-settler start.

One thing to bring up is that your captial city is by large worth more then half of the rest of your civ, because everything else is corrupted to there. Hence: Bigger captial cities are a better start-up.

It also make sense to note that you should spend maybe 3-4 turns just finding the very best start position, because you'll probably never move your captial, and besides, you only get 2.
 
Hi Excilus,

I can understand that you’re a little disappointed that we didn’t all queue up to praise your method but I think it’s worth being realistic about how far it can take you. When you post remarks like “The sad fact is that most of you have been doing it the wrong way” you’ve got to expect a few of us to pop up and say “hang on now!” (And I’ve done exactly the same thing many times – posted my great “discovery” to a forum and had it fall just a little flat!)

There are tens of thousands of us playing Civ3 now. I would imagine that a big percentage settle for a favourite build order that is some variation of your suggestions. Certainly, the posts I’ve seen here and on other forums seem to support that. It shouldn’t take too many hours of play before you realise that just banging down a build queue in any old order isn’t working very efficiently.

But “your” way isn’t the only way, and it isn’t even always the fastest. (My test showed that in some circumstances, the setter, settler, settler way can actually build MORE cities than yours, rare though this may be). I did also test Granary, settler, settler, settler and found that also outdid your original preferred order (Interestingly, you actually added this in your second post).

Your post will probably largely be preaching to newbie’s who may take it a bit too literally. Apart from all the terrain details and general game settings, a lot depends on what style of game you are going to try and play – Civ3 has such a big range of options. Sure, a good fast start is probably always the best foundation, but the decisions you take in the first 50 moves or so can have a profound effect – for better or worse – depending on what you are trying to achieve. You have a focus on troops that many of us wouldn’t copy for instance.

As you no doubt know, all aspects of the game are still hotly debated because there are so many variables with play style, civ attributes etc. Some players even deliberately avoid Pyramids because they feel the granaries unbalance their cities later in the game, (I tend to be in more in your camp on this one) and there have been a number of threads debating that issue. To build immediately or “waste” a few precious turns exploring is also a hotly debated issue. (For newbies it might be worth saving 4000BC, having a quick look round and then reloading, until you get a feel for whether you’re a “build now” or a “look round” type! Cheesy but instructive!)

Thanks for raising a great topic. It’s made me spend some extra hours practising starts, trying out methods, and generally thinking about it all. And writing all this waffle out has made me try and crystallise my thinking a bit more. My feeling is still that working out default build orders like yours is a very handy “mid point” strategy until you get a strong handle on how to really manage building creatively, efficiently, and appropriately to the circumstances.

Cheers. Polonius.
 
I've seen post where people have taken a starting setup and tried it both ways, just to see which gives faster expansion. I tried the same thing myself, and found two results.

First, it depends a lot on your starting position. Especially extra food squares. The fewer extra food squares, the more of an impact building a granary has. Also, the size of your city may be limited by difficulty level. That's also going to depend on how many warriors you can build (not scouts--they don't help happiness) between settlers. And your starting position for luxury availability.

Second, it took quite a while before one or the other system pulled ahead enough to show which way was better. I had one starting position that took me until almost 1300 BC before I could see much of a difference.

I'm thinking of trying a hybrid. Maybe build granaries in the cities that don't have extra food squares, or wait until after producing a couple of settlers to build granaries in the high food production cities.

There are some other tactics that help. When rush building a granary, if you can get 21 shields before you rush it, it only costs one citizen. If it looks like you are going to get a population increase a couple of turns before rushing it, change production squares to produce less food. If you build the granary BEFORE you get the pop increase, your food box starts out half full. Could save a few turns.

When pop rushing settlers, if you pop rush when your city is size 3, it will only give you the free settler on lower difficulty settings. It bumps you down to size 2, which is too small to produce a settler. I understand on chieftain and maybe warlord diff levels, it will give you a settler anyway. But when I try it on monarch and up, I have to wait until the population gets back up to 3 before the settler pops out and bumps my city back to size 1.

When pop rushing stuff, choosing when to actually rush build can give you extra commerce if you do it right. Sometime extra food, too. The optimal time to do the pop rush seems to be just after the pop increase that will put your city into civil disorder due to unhappiness. You may as well build wealth until one turn before that, if you don't mind the micromanagement. The problem is if you mess up and don't notice the pop increase, you go into civil disorder and lose two turns of production rectifying the situation.

I usually play with enough opponents that I can trade for pottery pretty early on. I don't even bother researching any of the startup techs any more.
 
OK..I admit that popping out settler and expanding as fast as possible is important. At some point though you HAVE to concentrate on military. Any good tips out there for when to stop. I have found that is isn't necessarily when you run out of room. I have been attacked by the AI who only had like 4 cities..but concentrated on military. I was a little weak..because of my rapid expansion. So I have found that sometimes it is best to stop expanding and just build our forces up and prepare to launch an attack.
 
The start location is a key factor. Usually I just struggle to get as many and as fast as I can. A 5 pop town produces a settler faster, but takes longer to replenish. I had one start with 2 cows, and I could not make settlers fast enough to keep happiness under control...

But I also dicovered another issue. It is hard to keep up with AI when they start with two towns. I saw this twice at Warlord level where I discovered the second -- not capitol-- city on turn 5.
I find it very difficult to get a settler to go 5 squares, build a town, grow to pop 2 and have a defender in 5 turns --and they did this from go from go. To say nothing of the time it takes to make the socond settler.
:confused:
 
"i tend to neglect granary's"

Zoolook,

How could you "neglect" granaries? They Halve the time it takes you to grow your city!?!? I don't understand this...

-J.:confused:
 
Granaries are a surprisingly complex topic as it is not as clear cut as just halving the time to increase a pop point. The first downside is the 'opportunity cost', i.e. the time taken to build the granery is time not spent building a settler. Therefore while the city will turn out settlers faster after the granery, you will have two cities turning out settlers slower if you don't build the granery. The second point is that building a settler requires both sheilds and pop points, so whichever is slower will be the rate limiting step. Halving the time for population growth will not necessarily double the rate of settler production. As a number of other replies have noted, the two approaches provide very similar outcomes. Presumably one could build a spreadsheet to model this and provide a more rigorous analysis; however, I would rather play the game.

The third comment I have has to do with level. The lower the level, the better graneries will be as you will be able to have larger cities, producing more shields and then you do have a shot at actually doubling the rate of settler production. At higher levels, you will also get bigger cities faster but unless you are lucky with luxury resources (and I never am), or spend valuble settler building time on temples and garrisons, half the population will be entertainers so the larger population will not be a benefit.

My own view pretty well matches Maven's - it depends on the starting position. The better the starting position with respect to food generation, the less likely that building a granery will prove beneficial. I build graneries early in cities with limited food resources, i.e. hills, forests, or plains with few bonus tiles. I have to admit that I just restart if I get jungle/mountains/tundra, the game is tough enough without that level of handicap.

Finally, I have not intended to be critical of Excilus when agreeing with other postings. The thread he has started covers as critical an aspect of the game as any and his opening post not only has a number of good ideas but has sparked a lot of very interesting and useful discussion. One of the things I really like about Civ 3 is the aggressive expansion of the AI. In the previous versions, one could win routinely on deity by just surviving through the early game because although the AI would have 10 cities by the time you got 5 or 6, you would have 50 by the time the AI reached 20, if it even expanded that far. Civ 3 provides a much more competitive AI for expansion and corruption (which I hate but in my heart know is an excellent addition) makes that shear force of numbers more difficult to acheive.
 
Top Bottom