granary or settler first?

[Rustwork]

i have been over this in my head numerous times and i feel that a quick settler, allowing another city to develop while the granary builds is more advantageous overall than getting the granary up first. i think this works best when food is high. does anyone have any examples of what is better. seems the best players go granary first. looking for some input before i play through a 50-60 turns from the same start to see for myself.

 

[Hellfiredoom]

I build a settler first. As you mention, I like having that second city growing and to spit out units so my capital can focus on settlers (usually being my factory). This also gives my workers time to hook up a luxury or two to support my capital growing larger and avoid running into happiness issues.

 

[CommandoBob]

I'm not an expert on this, since I don't build a lot of granaries.

Overall, I think the answer to your question is:

It depends.

Difficulty level, map size, land mass type, number of opponents and barbarian settings, in addition to the initial tiles visible, all factor into the decision.

Are the settler or granary the only two first builds to consider? What about happiness concerns?

I don't pretend to know the answer to the general question, though I favor the granary since it can help build more settlers faster (less food needed to grow back after the settler is birthed). Note that I favor the granary, but that is not how I usually play, so don't use that as a firm guide!

As I write this I sense that my tone could be taken as heckling and belittling your question. I assure you it is not and I am not. You've asked a serious question, but to answer it fully, or at least, more intelligently, requires some more information.

Could you post a screen shot of the starting location you are looking at?

 

[Opies]

I always do Warrior -> warrior -> settler -> granary -> warrior/worker -> settlers...

after the first few warriors the city is already level 3 so getting out that settler is quick and easy, and you get a second city going while the granary builds. I don't see switching the order to be advantageous.

 

[Regina]

I won't say this is the most sound advice ever, I don't know. I pretty much taught myself how to play Civ II years and years ago and I developed a system for branching out quickly which has always worked for me.

Most of the time I've tried playing for culture, although part of the reason I've been so lousy at Civ III was because I didn't truly understand all the win conditions and I'm still working on the trade thing and what to research and what not to research, etc.

Sometimes I build a granary, sometimes not, depends on how fertile the land is. I always try to make sure my capital has plenty of food sources for quicker growth. If not I keep trying for a better map. (I load a map, save, then run my units around for around twenty turns to see if I want to play on that map or not--keep in mind this is mainly because as a fairly novice player I want as many advantages as possible.) I won't play a map if there isn't enough good conditions to build at least ten cities before I have to start clearing jungle or swamp land, and even in those cases I usually build cities all the way around the jungle or swamp to get it into my territory, then clear it later after I get to the middle ages, and then build cities to fill in the gaps.

I don't take all my settlers out of one city, either. I take up to four out of the capital and get them started. Then each of those cities builds two settlers and each city after that builds at least one. This way my empire expands exponentially and very quickly.

If there is one wonder I like to build in ancient times, it's the Pyramids for the granary in every city. I usually start on that in my capital as soon as it becomes available which means it's even more important that I have a few other cities started for getting out settlers.

Like I said, this may not be the most sound advice. It's just a method I came up with eons ago that has served me very well in fast expansion.

Of course one caveat is if you start out with a civ with the agricultural trait. If you do that you won't need granaries or the Pyramid.

 

[Lanzelot]

I was never sure of the "correct" way to do this either. Usually I let my stomach decide on a game-by-game basis...

One factor to consider in this equation is: fast early research. On large maps and starting with Emperor level, it is essential to finish Writing quickly, otherwise you run the risk of missing the slingshot. If you first build a couple of warriors (for scouting and MP) and then the granary, then the capital has already grown to size 6-7 by the time the granary is done. And all that time you have collected a lot of beakers (if there are a couple of good commerce tiles to work on). Usually this get's me the first couple techs faster than building a settler first and having two small towns do the research.
Coincidentally, in the cases where this matters (large maps) the AI usually takes longer to find you, so it is not that important to have an early town providing military protection.

However, if I expect the AI to be close by, I usually prefer an early settler over the granary.

Lanzelot

 

[ThinkTank]

I had a fairly detailed discussion with MAS and others on this subject 2 1/2 years ago, in this thread.

To summarize my view: purely focussing on growth, granary first is strongest. However, there can be other priorities, like having an early town on the coast to produce curraghs for contacts, or having an early town for producing veteran ground units. So for these reasons I normally go for a settler first and then for the granary.

EDIT 1: I'll see if I can re-attach the pictures and Excel sheets to the old thread, I must have thses around somewhere.

EDIT 2: done.

 

[Spoonwood]

Could you post a save for the people here Rustwork? Who's your tribe? What sort of land do you have available?

 

[Fiddlin Nero]

I was never sure of the "correct" way to do this either. Usually I let my stomach decide on a game-by-game basis...

I don't see how "correct/incorrect" apply here. Most efficient in getting REX going I suppose but that doesn't really allow for enjoyment. When I first get to 2 or 3 population I like to build 1 or 2 workers to get that road building efficiency so I can stay ahead of my settlers. IMHO granaries first.

 

[Opies]

I was never sure of the "correct" way to do this either. Usually I let my stomach decide on a game-by-game basis...

One factor to consider in this equation is: fast early research. On large maps and starting with Emperor level, it is essential to finish Writing quickly, otherwise you run the risk of missing the slingshot. If you first build a couple of warriors (for scouting and MP) and then the granary, then the capital has already grown to size 6-7 by the time the granary is done. And all that time you have collected a lot of beakers (if there are a couple of good commerce tiles to work on). Usually this get's me the first couple techs faster than building a settler first and having two small towns do the research.
Coincidentally, in the cases where this matters (large maps) the AI usually takes longer to find you, so it is not that important to have an early town providing military protection.

However, if I expect the AI to be close by, I usually prefer an early settler over the granary.

Lanzelot

The next game I play I'm going to move on to emperor so this has really made me think twice about settler first. I never considered the tech advantage, and at level 6/7 2 settlers will come fast anyway. At monarch I've always been able to get the slingshot doing a settler first but that might change, especially if I don't start with alphabet.

 

[Rustwork]

i don't have any screenshots. it's something i consider each game, and was just looking for logical reason for one or the other. or criteria to follow for deciding which to build first. replies have been helpful. i might try a couple maps, same start, and see how it effects the game.

 

[Raliuven]

I am a regent-monarch player so it is hard to argue with those that play at a higher level. For me, it depends on the lay of the land. If the city has the potential to become a settler factory, I usually go for the granary first and plan on the factory to help out expand the AI or the settler first strategy. I hadn't thought of the tech bonus - that is interesting.

If it doesn't have the potential, I scout for a location while I build a settler to place a settler factory. On low levels, a factory isn't necessary to survive. On higher levels it can be the deciding factor.

 

[GamezRule]

I usually build a settler first, but I have been reading some old stories lately and the high level games usually go warrior > warrior > granary > settler...

 

[Raliuven]

I thought I would report on a little test that I did over the last 24 hours. My goal was to try both approaches on a map and see what effect it had on the game. I kept track of each move and each city - food and shields produced. I was actually surprised by the outcome.

CIVIII Gold (PTW) - Standard Map, Archipelago, greatest land setting, normal, warm, 4 billion years, Chieftain (I didn't want the AI to pester me), all defaults but not culturally linked, 7 rivals. Roaming barbs – though I didn’t encounter anything but GH.

In hindsight, I should have picked a Civ that started with pottery, but I chose the Greek. As it turns out, it didn't matter because when I founded my city it popped a GH and I got pottery! No need to restart. Saved base game for both plays.

Two models - Model #1 - create granaries first, then settlers. Model #2 - create settlers first, then produce a granary. Produce more settlers if possible. I played the first 51 turns. I'm not saying a better player couldn't have created better results.

The map generated offered no bonus food anywhere near me, so that doesn't factor into this. My start position did offer 3 BG, which was nice. I settled in place. I did not produce any additional workers. I duplicated the worker moves and I believe that they were generally the best moves available. I am more doubtful in Model #1 - I think I should have mined & roaded a BG near city #3 rather than pursuing an upgrade of city #2.

Outcome of Model #1 - 4 cities created. 1 settler just produced and waiting. City #1&2 had granaries, total of 7 citizens, and 1 settler and 1 granary due in 6 turns. 13 gpt economy. I only produced warriors after a granary was built and while waiting for my population to recover. My army was 4 warriors. My exploration was pitiful and I didn't pop any more GH. I had only 4 techs (I set research to 40t at the beginning). America, my nearest neighbor built towards me and our cities were already bordering each other. City #2 founded Turn 27, City #3 T33, City #4 T45.

Outcome Model #2 - 4 cities created. 1 settler just produced and waiting. Only City #1 has granary. Total of 4 citizens and growing slowly, Granary in 11 and Settler in 24. 9 gpt economy. I immediately produced warriors while waiting for my population to grow to the point that I could start a settler build and while waiting for my population to grow between builds. My total army was 7 warriors. My exploration was much, much better and I popped 3 more GH, 1 for a worthless map, 2 for techs (CB and IW). The really interesting thing is that America did not build toward me this time. I had suspected that the AI would evaluate the situation the same – there were plenty of bonus food and luxuries between us – which we were both growing towards. At the end of turn 52, America still hadn’t moved towards our common ground. Also, because my scouting was so much better, I placed city #4 in a much better place – possessing 1 cow and set to another cows and a game with a culture expansion. I thought it would be dishonest to settle city #4 in the same place as model #1 since I had much better information - which is a strength of this model. City #2 founded in turn 23, City #3 T45, City #4 T46 (yes, only 1 turn later).

I found the results very interesting. Model #1 might have had a stronger core to work from, but Model #2 was in a better tech place, our knowledge of the world was broader and is it possible the fast building military convinced the AI to build away from the great Greek civilization? Both produced 5 settlers by turn #51, but I think it is clear that Model #1 was set to pull away from Model #2 – settler due in 6 turns as opposed to 24 turns. Also, Model #2 did produce the first city faster (turn 23 vs. 27) but then lost pitifully on City #3 (T33 vs. T45) and by 1 turn on City #4. City #5 would be produced at the same time but City #6 would be in Model #1 favor.

Of course the model is only so good as the test (and player). Most of us would not produce the same queue in every city before evaluating overall needs. If you are in a tight space or need military to intimidate, model #2 may be the solution. If you know you have a lot of room to expand, maybe model #1 is better.

That is what I found, for what it is worth. I might try it with an industrial Civ and see if that changes things. For a fair portion of the time cities #3 & #4 were working unimproved tiles.

 

[NickyH]

My experience is that granaries become less important on higher difficulty levels, combined with small maps. It's possible that you'll have to compete with your neighbours for space already after your first 5-7 cities, and in that case, you'll be able to grab more land if you build settlers early, not only from your capital city, but from any city that has enough population to build a settler. That means that it will take a lot longer for your cities to become powerful, but on a small map on Demi god or Deity level, you might not get space enough for more than 5 cities if you take the time to build a lot of city improvements and let your cities grow, before you build your settlers.

This depends very much on how crowded your part of the world is from the beginning, of course. You'll notice how far away your neighbours are, as you scout around, and make your decisions based on that.

(It was a very long time since I played a game like that, so I might be mistaken when I say 5-7 cities, I don't remember exactly how many how many you normally can build with those settings. But I'm sure you understand what I mean anyway.)

I think standard size maps on Emperor difficulty is the kind of game that I've played most, and with those settings, I usually build at least one settler, maybe even two, before starting on the granary. I'm more likely to build only one if I have a nice food bonus in my capital, and I'm more likely to build two settlers before any improvement, in case I find a couple of nice spots with cows for my second and third cities.

And then there are those games that I like to play nowadays, of course. 20k culture on Warlord level, on a small map. In those games I never build a granary, nor a settler in my capital. If I don't find a free settler from a goody hut early, and get a scientific leader (to build the Pyramids) within the five first techs that I research, I just feel that it's a lousy start, and I start a new game.

 

[wear146]

I will go for a granary if I start with pottery, other wise I usually go with a settler

 

[del62]

Interestingly some have said it depends on how many city sites are available close by with food bonuses, (cows, wheat) etc

 

[templar_x]

@ del - that´s logical in my eyes (um, well, i am one of those who supported that strategy ), if you look at settler BEFORE the granary as more "expensive" than those you build AFTER the granary. you´re using twice the food=growth for the first ones, compared to the latter.

thus you´d only lavish that money if you can get a good deal for it, i.e. plant that settler down on a good or very good site. this site should offer more than just a few additional 2f-tiles to work. otherwise it would be "cheaper" to build the granary first and enjoy the faster pop growth in your capital/SF-town!
on the other hand, if you can use an early settler to found another SF or WF-like town, you´ll gain ANOTHER town which adds population to your civ fast.

therefore, in my eyes, the early settler only makes sense if i have food rich-places to make use of. otherwise i´d always wait for the granary to finish.

templar_x

 

[del62]

templar x- it was you I was thinking of when I posted.

 

[Spoonwood]

On higher levels, even with food bonuses around, I've gone granary first and then settlers (not 20k, I don't put a granary in the capital in that sort of game). If you want to progress to higher levels faster, the easiest way to step up comes as to make game conditions as easy as possible and start playing at those levels as soon as you can. Barbarians on higher levels really only help the AI, so get rid of them. Lower the aggression level to decrease the probability of unwanted demands and attacks. Try to look through some starts to find a food rich start (grassland wheat or cow). If you don't want to insist on a 5 food start, at least look through 3 or 5 or some number of starts before you settle on a game which has bad starting territory and you end playing a game that you'll lose, or have more difficulty of winning.

With respect to the granary, once one masters managing a settler factory, with no barbarians around and AI's with low aggression, even though the AIs may initially have 3 or so cities before your granary finishes, it becomes possible to have about the same number of cities as the AI by the end of the expansion phase, up to and including Deity level.