Optimum Early Growth Strategy

[Grimz101]

if your playing raging barbs try ganhdi for fast workers, on fast games not much difference, but normal+ you will really notice the difference, gotta love gandhi for chopping cities really fast, on small maps, on standard+ though id rather have financel, since costs will start killing after a while, only recommend gandhi on standard+ if your playing a team game.

 

[ohioastronomy]

Good article! Would you compare the the early growth for the Marathon game speed too? I usually do worker/worker/settler at normal game speed and warrior/warrior/worker/worker/settler at the marathon game speed with very little chopping and some queue swapping. The reason for this is because there are usually too many barbarians in my game and I need to save the forests for swordman rush.

As for building the early fishing boat, I do not build a fishing boat until after I have conquered around 30% of the world. The reason for this is simple: it takes too much time to research sailing and building ships. On top of this, seafood is the AI favorite target. Even when they are dying, they would still try to rush a ship to go after my seafoods. Therefore, I think acquiring seafoods is a waste of time and because I can't really protect my seafoods in time of war. With enough swords and axes, I could easily claim another food source on land. Plus I need a strong military early to deal with the barbarians too. Of course, I could be wrong about this; there is still much about Civ4 that I don't know (probably because I'm still playing Civ3 and very little of Civ4 so far).

The build orders that you describe are optimal at high difficulty levels because of the slow pace of technology research at marathon, with a couple of possible exceptions. If you start next to a commerce resource (gold or grassland/gems are ideal) then your worker can build a mine and the early tech game will go much faster. This translates directly into increased production. Because of the long build times, you can also benefit from making it a priority to build on top of a plains/hill (4 base production), since this shaves a lot of turns off the early units. It is less important to have animals/quarries in the fat cross because they will not be developed for many turns. Financial civs can also swap worked tiles (alternating a forest and lake, for instance) to accelerate tech development. In general, you'll come out ahead if you can move forward the first usable worker turn.

 

[ohioastronomy]

How can you ignore commerce when commerce is the only thing thats going to allow you to settle more cities? This becomes more apparent in the higher difficulties but ultimately its determines your expansion rate in all of them.

I should have worded that a bit more carefully. My real point was that the choices you make at the very start of the game have limited impact on commerce. You have to have workers and cities before you can build cottages, the important buildings, gold mines, etc. Yes, it does matter tremendously outside the time frame that I'm describing here. You'll have more of everything, including commerce, if you focus on workers early and take advantage of the stored hammers in your trees.

 

[civsim]

You are obviously a great player: but as for communication and clarity . Am a good player of civ3 and starting on civ4. Was looking for a good sound strategy to follow for my early days, on civ 4. Obviously lots of thoughts and research on yours , but for the ordinary player who enjoys the game , your essay is heavy to say the least. would it be possible for you to make your strategy less technical , and just simply say what is required at the start without the techno bable. As i understand , Chop wood (forest), and build up workers ( for resources; pig, cattles, horsesetc) for them to develop before expanding (with settler) would be a simpler advice and the numbers are less important when you just want to enjoy the game. And what about when you face constant harrasement from barbarians, who destroy your improvements??? That should be included as well...
If your advices where met for advanced players who wouldn't need these starting pointers, ok but if its for the beginners at civ4, you should be more to the point.

 

[ohioastronomy]

You are obviously a great player: but as for communication and clarity . Am a good player of civ3 and starting on civ4. Was looking for a good sound strategy to follow for my early days, on civ 4. Obviously lots of thoughts and research on yours , but for the ordinary player who enjoys the game , your essay is heavy to say the least. would it be possible for you to make your strategy less technical , and just simply say what is required at the start without the techno bable. As i understand , Chop wood (forest), and build up workers ( for resources; pig, cattles, horsesetc) for them to develop before expanding (with settler) would be a simpler advice and the numbers are less important when you just want to enjoy the game. And what about when you face constant harrasement from barbarians, who destroy your improvements??? That should be included as well...
If your advices where met for advanced players who wouldn't need these starting pointers, ok but if its for the beginners at civ4, you should be more to the point.

The main point is that workers are extremely valuable for the start of your game, followed by settlers, and that city growth is a distant third. For a newer player I'd recommend starting with the following:

1) make your first city build a worker and your first tech bronze-working. Choose a civilization that starts with mining. Qin is a good choice (mining and agriculture, plus financial=faster technology and productive=faster wonders and forges).

2) After your first worker comes out, have them improve something while you build a warrior. Pick other technologies according to the type of game that you want; agriculture and the wheel are pretty important first off; writing is also important for a lot of other reasons.

3) Send the warrior out to your second city site. This should be close, ideally along a river that your first city shares, and a spot with 2 or more resources within two tiles.

4) Have your first worker chop a tree to build a second worker. Have both of them chop a tree to build a settler. You can cut trees up to three tiles away from your capital without penalty; just be careful of animals if you go outside the cultural boundaries.

5) Send the settler to the new city and send one of the workers with them. Have the other worker stick around the main city and develop it. At the second city, have the worker chop a tree for another worker. You'll want to found later cities with the same pattern - send a military unit, settler, and worker together and have the first thing that the worker does be to make another worker as long as there is a usable tree there.

6) Leave one worker around each city. Extra workers can be put to work connecting cities with roads. Once the core cities are fully developed switch their workers to roadbuilding and assisting the new cities.

7) Don't expand indefinitely. When your empire gets too large the maintenance costs get high - the point where this happens depends on the difficulty level. At Prince, it starts to kick in around 4 cities. Make getting Code of Laws and courthouses a priority. If your tech is dropping below 60-70% it is time to stop building new cities and make your existing ones better.

Hope that helps!

 

[A+ombomb]

I think the optimal build depends on if you are playing on what the game speed is. On quick, I would go right into worker/settler with a warrior building inbetween chops on the settler so the warrior could defend a second city from barbs. On slower speeds, the worker seems to pop out way before bronze working is done, so its probably better to do something else.

 

[Zombie69]

The extra production power of the first workers is so large that even a substantial benefit, such as fish, turns out to be only roughly equal to just making a worker.

If it's roughly equivalent in terms of food and production, then i'd say it's certainly worth doing, if only for the extra commerce from the tile with the fish.

 

[ohioastronomy]

I think the optimal build depends on if you are playing on what the game speed is. On quick, I would go right into worker/settler with a warrior building inbetween chops on the settler so the warrior could defend a second city from barbs. On slower speeds, the worker seems to pop out way before bronze working is done, so its probably better to do something else.

Your starting position also matters. At longer speeds and high difficulties it may be most efficient to focus on commerce improvements, so that when you make the workers they have something to do. I've been messing around with archepelago (preparing for a run up to deity). There aren't nearly as many forests as a normal start, and hammers are precious. Improving a gold, grassland/gem, or ivory first speeds tech up so much that you end up with faster early production. A similar comment applies to fish; clams/crabs are less critical. If you're financial you can juggle the worked tiles to get bronze-working and the first worker timed to appear at the same time (switch off working a 2P/3C tile and a 3P tile). Similarly, although you lose theoretical production, at higher difficulties you can do well with a modified build sequence:

Worker/BW
Worker does an improvement, city puts out a warrior
chop worker 2
double chop settler

 

[Briquette]

Many of us play with random leaders (prefer the challenge of adapting to a situation rather than playing game with known and rehearsed strategy). Thus won't be able to pick and choose leader traits to match this tactic.

Many play with roaming barbarians (already discussed the damages they can do with unprotected improvements and workers).

On mid-range difficulty levels (where many people still playing), the goodie huts can be very valauable and are generally missed out on if you don't get a couple units looking around right away.

Not actively exploring may also prevent locating the best location for cities to stake your claim to an important resource (copper, iron, etc.)

And of course finances are a concern if you continue this strategy to expand and build too many cities early on.

Then there's the thought of saving trees for "critical" production of a key wonder when their value chopped may be 60 or more due to civics, etc.

Don't get me wrong, I think the idea has great merit and is obviously well researched and documented. I just think there are sometimes more elements to consider that may or may not outweigh the 2 worker chop rush strategy.

 

[ohioastronomy]

Many of us play with random leaders (prefer the challenge of adapting to a situation rather than playing game with known and rehearsed strategy). Thus won't be able to pick and choose leader traits to match this tactic.

Many play with roaming barbarians (already discussed the damages they can do with unprotected improvements and workers).

On mid-range difficulty levels (where many people still playing), the goodie huts can be very valauable and are generally missed out on if you don't get a couple units looking around right away.

Not actively exploring may also prevent locating the best location for cities to stake your claim to an important resource (copper, iron, etc.)

And of course finances are a concern if you continue this strategy to expand and build too many cities early on.

Then there's the thought of saving trees for "critical" production of a key wonder when their value chopped may be 60 or more due to civics, etc.

Don't get me wrong, I think the idea has great merit and is obviously well researched and documented. I just think there are sometimes more elements to consider that may or may not outweigh the 2 worker chop rush strategy.

There isn't really a contradiction - of course, you're correct. What an exercise like this does is to quantify the tradeoff. If you can get away with it, the sequence of productive things to do is well-defined (workers, settler, then grow). If you have a nifty special resource, improving it first also yields good production. There has been a lot of discussion about tree-chopping; this exercise confirms that the benefits are real. Ditto for queue-swapping, although I think this is really a short vs long term payoff question.

More generally, it suggests that on average you are better off if you can send a worker/settler pair to found a new city. Chopping trees to accelerate the process is economically very efficient - enough so that you might gain enough production later to more than make up for not having the trees later. I stopped after 2 cities because, again, you're correct - you want to expand to the appropriate number of cities for the difficulty level. Beyond that point, you need Code of Laws and courthouses - or enough gold mines, etc. that you simply don't care much.

If you are running with raging barbarians, or you're running at Emperor+, you can't as easily follow the strategy that gives you the most production. In effect, this represents the handicap that you're taking with those choices.
I hope this helps!

 

[NobodyImportant]

I think I understand this, but I wonder whether your conclusions still hold when we complicate the analysis to include commerce.

Consider my current game as a case study. I'm playing as the Americans on Noble level at Normal speed, with Washington as the leader. Washington (the city) is in a lovely spot, on a plains hill by a river bend. Playable resources include Wheat, Gold, Ivory and three flood plains squares. It's 3480BC, and I've just completed my first worker.

As I understand it, your strategy would have me start a Settler next, while my worker irrigates the Wheat and then chops forest to speed production. That seems like it would be effective. But my intuition tells me that I should push out a Warrior and wait until my city grows to size 2, in order to get a bloke working the gold mine before I start Settlers. It'll take 7 turns to grow the city, but the gold square will be worth a whopping 9 commerce and 3 production once the mine is built.

Once again, I think I understand your point about the importance of making Settlers early, but I wonder if it's worth the opportunity cost of missing that gold square in this case. I'd be interested to hear your thoughts on the matter.

 

[ohioastronomy]

I think I understand this, but I wonder whether your conclusions still hold when we complicate the analysis to include commerce.

Consider my current game as a case study. I'm playing as the Americans on Noble level at Normal speed, with Washington as the leader. Washington (the city) is in a lovely spot, on a plains hill by a river bend. Playable resources include Wheat, Gold, Ivory and three flood plains squares. It's 3480BC, and I've just completed my first worker.

As I understand it, your strategy would have me start a Settler next, while my worker irrigates the Wheat and then chops forest to speed production. That seems like it would be effective. But my intuition tells me that I should push out a Warrior and wait until my city grows to size 2, in order to get a bloke working the gold mine before I start Settlers. It'll take 7 turns to grow the city, but the gold square will be worth a whopping 9 commerce and 3 production once the mine is built.

Once again, I think I understand your point about the importance of making Settlers early, but I wonder if it's worth the opportunity cost of missing that gold square in this case. I'd be interested to hear your thoughts on the matter.

In this case, chop-rushing actually has the nice benefit of accelerating early commerce as well as production, with a modest change to the cycle. Here is how it would work:

If you wait until the city grows to size 2, you usually have to work some tile other than the gold mine for the population to have excess food. Before the worker improves a tile, this growth will be slow - 3 food/turn on an unimproved wheat. You therefore have to wait 8 turns on normal before you have a spare person to work the gold mine, plus the lag time to build a worker who can build the mine.

In the worker-worker-settler setup, your first action is to build a worker and improve the gold mine. Once it is improved, have the city work the gold mine and pump out early commerce. You don't get growth anyway when you're building workers and settlers, so you might as well pick a tile with nice commerce. Then you chop a second worker. Have the second worker chop for a settler and then improve the wheat. (Improving the wheat first doesn't help, because your city can only use one tile at size 1, and you want that tile to be the gold mine). Now you have an improved gold and wheat and you set the city to grow, working the wheat tile. Since you're now at +6 food, growth will be quick (4 turns). More details:

Work the wheat turns 1-15 (make worker) and turns 16-20 (building gold mine). Worker1 chops on turns 24, 28. Worker2 appears on turn 24 and chops turn 28. Settler appears turn 31, wheat is improved on turn 33.
Work the gold turns 21-33, then work the wheat turns 34-37, followed by wheat+gold. You get 13 turns of high commerce out of the first 37. The mathematic maximum if you never grew and only worked a developed gold mine is 17. Everything is a bit faster with extra production from a starting hills/plain, but the principle is the same.

If you improve the wheat first, followed by the gold, grow to 2:
Work wheat turns 1-25 (1-15 for worker, 16-20 to farm wheat, 21-25 to mine gold). Growth to size 2 will occur on Turn 24. Turns 25+ you work gold and wheat. Worker1 chops turns 29, 33. Worker2 appears turn 29, chops turn 33. Settler appears turn 34. By turn 37 you've had 12 turns of high commerce, one fewer than the other approach. You've also delayed your new city by three turns (but gained one turn of growth on your home city), for another small net negative.

In either case you have the ingredients for a very, very solid start.

 

[KerThud]

Very good analysis and discussion. I can't wait to see the followup article: what the tradeoffs are for various courses of action at turns 40-100. Also, what do you think of the idea of looking around for a plains/hill to settle on? I thought that was an interesting point. Again, it neglects commerce because you can't research until you found that city, but you'll certainly get some production, and will have some extra scouting (from the hilltop) for a future city location.

 

[hollebeek]

The reason you can ignore commerce is because the commerce from your
palace (8) will overwhelm any additional commerce due to city size (e.g. +1).

And remember that building a second city, especially along a river, is a
bigger commerce boost, since you get +2 for the two squares it works
and +2 for the two trade routes.

The one case where I think it DOES make a difference, is with gold. With
a gold start, I will often mine the gold and switch the city to "max commerce"
while I am chopping. This can provide a quite significant commerce boost,
allowing for fast early research.

 

[junior7]

Great article! On prince now, on standard, and this is useful knowledge - might enable me to progress to monarch!

I could follow most of the numbers/scores in first post but cannot reconcile Case A (Worker,Worker,Settler,Improve). You quote:
Case A: Worker2 T23, Settler T27, Imp T32, 21 worker turns, 16 overflow, Score=+153.5, chopped=3 trees
The numbers I have trouble with are overflow (and that element of score) plus number of chopped trees. If Worker1 chops T19+T23+T27 and Worker2 chops T27 then this is 4 trees and production is 27x4 + 30x4 = 228 less 220 (settler+worker+worker) which leaves overflow=8.
Sorry if I missed something but I cannot see it.

Also does your score take into account that early improvements in Case C & D provide additional production (after workers and settler) over Case A which does not have improvement available until T32?

 

[ohioastronomy]

Great article! On prince now, on standard, and this is useful knowledge - might enable me to progress to monarch!

I could follow most of the numbers/scores in first post but cannot reconcile Case A (Worker,Worker,Settler,Improve). You quote:
Case A: Worker2 T23, Settler T27, Imp T32, 21 worker turns, 16 overflow, Score=+153.5, chopped=3 trees
The numbers I have trouble with are overflow (and that element of score) plus number of chopped trees. If Worker1 chops T19+T23+T27 and Worker2 chops T27 then this is 4 trees and production is 27x4 + 30x4 = 228 less 220 (settler+worker+worker) which leaves overflow=8.
Sorry if I missed something but I cannot see it.

Also does your score take into account that early improvements in Case C & D provide additional production (after workers and settler) over Case A which does not have improvement available until T32?

Good catch; you're correct that the overflow is only 8 for the base case and it uses 4 trees. Blame a complex Excel spreadsheet :0 I also did not count production differences after the workers and settler are made. For the best case (wheat/corn, or the lucky copper mine) you'd get
24 excess production (case A, worker/worker/settler/good special)
8 excess production (case A, worker/worker/settler/mine)
12 excess production (case B, worker/settler/worker/good special)
4 excess production (case B, worker/settler/worker/mine)
36 excess production (case C, worker, improve special, worker. settler)
10 excess production (case D, worker, mine, worker, settler)

These changes make improving a special resource the best choice, also saving a tree. Note that this can only really be done with one nation (China), which starts with agriculture and mining. However, it does open up some interesting production choices for nations that start with agriculture but no mining (US, France, Persia, Inca, Egypt). If they have the right starting position they can research mining and bronze-working and get essentially the same production as the nations that start with mining.
Interesting side note...

 

[ohioastronomy]

Very good analysis and discussion. I can't wait to see the followup article: what the tradeoffs are for various courses of action at turns 40-100. Also, what do you think of the idea of looking around for a plains/hill to settle on? I thought that was an interesting point. Again, it neglects commerce because you can't research until you found that city, but you'll certainly get some production, and will have some extra scouting (from the hilltop) for a future city location.

This depends a lot on game speed. At normal, you're trading off one turn of technology research for enhanced early production. For a worker/worker/settler starting approach, the stats for base +5 production are improved quite a bit. Here is what they look like:

Worker after 13 turns (one turn delay in start)
Worker2 on turn 19 (chop on turn 17)
Chop on 21, 23, 25, settler on turn 25, improve on turn 28, 20 overflow
for four trees. This accelerates everything relative
to the usual timetable. The advantage is even bigger at marathon.
You also have a game-long production advantage in the home city. The cost
is more subtle: there may be a sweet set of resources that you could
reach at the first city site that are out of reach if you move. Commerce development will also be better longer term (the odds that you'll be able to use a commerce resource at one of your first two cities are pretty good).

 

[MyOtherName]

Forgive me if you've mentioned this, but I didn't see it addressed when skimming the thread.


Your analysis seems to completely ignore anything but the shortest-term considerations.

Forests are (roughly) a fixed resource -- when you chop them, they are gone forever. Each chop should have a penalty attached to reflect this fact.

However, you do exactly the opposite of what you should: your score significantly rewards burning through your forests in lieu of building sustainable improvements.


I'm aware that you said you only wanted to analyze the first 40 turns, but because of its short-sightedness, I think it's only relevance is for the person who has already decided they are going to clear cut, and is only interested in how fast he can do it.

It seems so easy to account for your "long"-term potential: simply eliminate the 7.5 hammer per turn bonus for worker-turns, and introduce a -30 hammer penalty for each chop. The only question is how to properly account for a city-turn. (I guess it's not that hard on emporer -- I imagine each city-turn is worth exactly what the city could produce at its happiness limit... say, 8 hammers per city-turn for a good second city?)

 

[ohioastronomy]

Forgive me if you've mentioned this, but I didn't see it addressed when skimming the thread.


Your analysis seems to completely ignore anything but the shortest-term considerations.

Forests are (roughly) a fixed resource -- when you chop them, they are gone forever. Each chop should have a penalty attached to reflect this fact.

However, you do exactly the opposite of what you should: your score significantly rewards burning through your forests in lieu of building sustainable improvements.


I'm aware that you said you only wanted to analyze the first 40 turns, but because of its short-sightedness, I think it's only relevance is for the person who has already decided they are going to clear cut, and is only interested in how fast he can do it.

It seems so easy to account for your "long"-term potential: simply eliminate the 7.5 hammer per turn bonus for worker-turns, and introduce a -30 hammer penalty for each chop. The only question is how to properly account for a city-turn. (I guess it's not that hard on emporer -- I imagine each city-turn is worth exactly what the city could produce at its happiness limit... say, 8 hammers per city-turn for a good second city?)

That is a good point, especially applicable to queue-switching (trading extra trees for units and growth while building workers and settlers).
It has been a very good discussion on this thread, and I did touch on this in posts 2 and 19. Even after docking the tree-cutting strategies for the opportunity cost of the trees, they still come out well ahead of cutting no trees because you get an earlier head start on the improvements in your capital and the start of your second city.

Worker turns are still very valuable because you can use them to develop improvements earlier or to start a city earlier, so that aspect of the calculation isn't affected by the early (as opposed to late) usage of trees. The trick is to compare two scenarios where you do everything the same, differing only in what the worker can do. Here is an example that might be useful. Imagine that we both found our second city on the same turn, but I have a worker there the turn it is founded and yours comes five turns later.
There is an irrigated corn there. I improve it five turns later when your worker arrives. If you look at my city when you've improved the corn,
I've gained 15 + 30 = 45 food directly from the corn. I grew to size 2 after seven turns and with one extra food I could be size 3, for a total of 49. I just improved a second resource. You gained 32 production (growing to size 2 on turn 8, but only getting one extra production because it is unimproved.) But now my city is working the second tile for the next five turns until you catch up, and this pattern will repeat for every population point that I add. It's therefore fair to count the benefit of one turn of 4-5 improved tile for a worker turn, and similar comments apply to getting the third city out.

This also has intereting implications for what you do with your workers - the yield from being efficient is pretty high. Not overimproving a city that won't work extra tiles for many turns and not obsessing about roads will have a surprisingly large yield, as will having improvements in place when a city grows into them...

 

[Orca]

Your analysis seems to completely ignore anything but the shortest-term considerations.

Long-term considerations are not very important ! I see this fault in thinking so often here on the boards, what you gain in short terms will multiply over the years and overgrow the longterm penalties by far.