Midgame strategy questions (number of cities, spacing and population) and early aggression

[Random Dude]

Hello! I'm a new player, just started playing a couple of weeks ago. I progressed through difficulty levels until I hit deity. With deity it seems like because of no prior experience with Civ games, I miss some important fundamentals and get crushed. Can you help me? I have looked through the forums of course.

QUESTION 1.

Up until immortal I did something like scout-slinger-settler, did "peaceful" fast expansion with settlers, grabbing a lot of land and defending as necessary, and it worked fine. This obviously doesn't work on deity, and besides, pretty much everyone says that "go kill your neighbor" aggressive expansion starting with 3 archers is better at high difficulties. My best last try was as Sumeria, I cleared some barbarians and took the nearest city-state with a warrior and 3 archers, then the next one with a couple more archers and war-carts to follow, then took Egypt out (they had two cities and were pumping chariot archers like crazy). That left me with 5 cities overextended across the whole continent, then barbarian horsemen hit everywhere at once, while I had no units at my capital. I think if I have been doing my regular peaceful expand, or at least stopped after the first CS, I might be able to hold it.

So why is the aggression considered superior, and how do people pull this off? Besides, I keep reading about that "10 cities before turn 100" stuff, but if I stopped at 3 cities to pump enough military for protection before going further, I highly doubt I could make it to 10@100 while also killing the neighbor. Do you always manage 10@100 on deity? Do you attack in several waves or non-stop? Do you rush for Agoge or Early empire first? (Any other tips?)

QUESTION 2.

Does district cost inflation apply to the same type of district, or districts in general? E.g. I have 20 commercial hubs and 0 industrials, does that mean that CH cost for me is inflated, but IZ cost is normal, or they both inflated, and district type doesn't matter? If it's the first, than is it a good idea to try to put all districts of the same type simultaneously? This is not about pre-placement, I already do it.

QUESTION 3.

What is the good amount of cities in the mid-game, spacing between them and population? Usually people say "wide rules" without any specifics.

In my emperor game I have been playing as Germany and had ~40 cities in Industrial era, put them as close as possible and didn't bother to grow them (basically, following some outdated pre-AOE nerf guide). Still, they each had at least CH and Hansa and ocassional entertainment, and I decided to abort the game because I was swimming in money and out-producing everyone, and micromanagement was hell.

In my immortal game I was playing as Rome and had ~15 cities, placed mostly 6 tiles apart (roads definitely help here), and focused on housing whenever needed, so they all were at least 7 population, had baths, etc. Accidentally overwrote quicksave after Australian patch. It was OK on production, but I think that I might have had problems with amenities with other civ if not for baths, and with defence if not for roads.

So I've been thinking. In order to have good amenities for bonus yields I have to limit the population somehow. So 13+ for the capital (and max space around it) with commercial, harbor, encampment and industrial for 5 production on trade routes, several 7-pop cities in hilly or forest-river areas to work the tiles, and the rest locked at 4 and put closely together because one ring is enough for them, but factories should make them useful. But to maintain 4 population, I should actively avoid settling on fresh water, avoid building farms, barracks/stables, universities, etc., and that seems more trouble than it's worth. So mostly 7 pop then? 4-5 tiles apart should be good, right? And how many? Multiples of 4 are obviously the best, so assuming peaceful games, is 16-20 enough for deity? After what number you usually say "okay, that's enough" if not going for domination?

Basically, it's possible to do any random stuff on lower difficulties and win, but I'd like to have a solid mid-game plan for deity, and I have absolutely no idea what to do.

 

[Bobfregman]

Generally I start off building 3 straight slingers and beeline for archery. I don't build any monuments, workers, or even settlers until I have my invasion force set up. Generally 2-3 warriors and 3-4 archers is enough to take out your closest neighbor at the very least.

The reason this strategy is generally superior is because the AI starts with 3 settlers and huge boosts to all their stats. Trying to out settle them and catch up is very difficult. By comparison that investment in 4 archers and 3 warriors could allow you to capture 6 cities for far cheaper than building the same number of settlers. Once you clear out an enemy or two you'll be able to settle the lands between relatively unopposed.

So generally for me it goes:
-complete army by then 30-35
-wage war till turn 60-75
-consolodate gains/settle until turn 100-110
-begin developing cities

District inflation comes from your civic/techs. Each new one researched adds a percentage cost to districts.

As for your final question it sort of depends what victory condition you're going for. You could land a science victory without ever going further than the number of cities you have at 100 turns. I usually go domination so I end up with 40+ cities on most games. You can use housing to control population if you're facing amenity issues but I rarely find it necessary in the long term. Even on diety the Colosseum is often obtainable and is basically the only contested wonder I shoot for. Late game the AI seems to ignore the Estadio so you can grab that as well, although the game is usually decided by this point. Pay attention to pop growth and build ED as necessary. There are a number of cards that give amenities as well.

 

[Random Dude]

So generally for me it goes:
-complete army by then 30-35
-wage war till turn 60-75
-consolodate gains/settle until turn 100-110
-begin developing cities

Thanks for the reply. After 20 tries or so I finally managed to kill a neighbor, getting 4 cities total by turn 70 with non-stop reinforcements, so I might be finally able to pull this off. Still, the fact that a neighbor (China) forward-settled me helped a ton, no idea how people can do it consistently.

The point about competing with 3 AI settlers makes sense.

District inflation comes from your civic/techs. Each new one researched adds a percentage cost to districts.

Isn't it both? Researched techs and finished districts?

As for your final question it sort of depends what victory condition you're going for. You could land a science victory without ever going further than the number of cities you have at 100 turns. I usually go domination so I end up with 40+ cities on most games. You can use housing to control population if you're facing amenity issues but I rarely find it necessary in the long term.

I usually go for science - looks like the same tech progression as domination until atomic, but with less micro required. Good to know that it's doable with less cities (I'll try it with 12), still I have a hard time understanding any possible pros of having less stuff (cities/districts) as opposed to having more stuff, other that not spending production on settlers.

 

[Bobfregman]

Doing it consistently has become more difficult with the latest update as the AI prioritizes walls quicker now. My strategy is to use your archers to clear out as many enemy troops as possible before putting them within range of a city. Once I've paved the way and let my guys heal up (using promos and waiting) i position as many archers as possible in firing position and use the warriors to siege the city. Make sure to use enemy farms as free heals for your units, usually you'll capture a builder that can instantly repair them after the conquest. Also worth noting is that the AI is still relatively dumb with their Settlers so it's not uncommon for me to steal a few from them.

I believe it's only civics and techs but admittedly I could be wrong. I'll need to look into it more.

A science victory on Diety is definitely possible with 9-12 cities. Of course this isn't always the case, it depends what civ you are and what the landscape of the cities looks like relative to your opponents. If you have the latest release theb Australia is great for science victories, if you have the right start you could probably pull it off with even fewer cities.

And yeah unlike Civ 5 this game really rewards going wide. The only real downside of having many cities, aside from time spent building settlers, is possible amenity issues but those are usually manageable.

 

[Evolena]

Isn't it both? Researched techs and finished districts?

It is only techs or civics (whichever tree you have most advanced in).
Your finished ditricts don't increase the cost of next districts, but you can have a 25% discount if you have less of a district than the average. This can lead to falsely think that the more of a district you build, the more it costs.

 

[ShinigamiKenji]

After the patch, the Battering Ram + Swordsmen/Horsemen combo seems better after the initial Archer rush. Use your Archers then to sweep enemy units while the Rams come close to the city. Horseback Riding is after Archery, so now it's quite a natural path to take for conquest. Swordsmen were buffed up, have a promotion to make them more resilient to ranged fire, and walk at the same pace as Rams, so they may be viable as well.

 

[Random Dude]

Thanks for the replies.

Make sure to use enemy farms as free heals for your units, usually you'll capture a builder that can instantly repair them after the conquest. Also worth noting is that the AI is still relatively dumb with their Settlers so it's not uncommon for me to steal a few from them.

This actually helps, previously I avoided pillaging farms because I forgot that repairing them costs no charges.

This can lead to falsely think that the more of a district you build, the more it costs.

Yep, for some reason I thought that's how it works. Still, then it appears than it's not a bad idea to put a bunch of districts of the same type at the same time if you have zero of that type and opponents have at least some. Not worth it to specifically prepare the timing, but if several cities are to hit 4 pop within several turns, then why not.

@ShinigamiKenji - Yes, already started going for rams if they start a wall and I really really need to take that city.

 

[Victoria]

Question 1
Yes 3 settlers makes it more cramped but you can and do get maps where the nearest civ is much further away. Barbs are normally a serious pain so getting military first is often (not not always) key. A lot depends on your map, a BIG amount depends on your map. Probably one of the largest things is your starting position. I tend to look only at food an prod values. I add them up and look at them individually. If you start in a position where all tiles only have a value of 3 (prod+food) then you will have a much slower start than a position where you have values of 4 or 5. Also prod far outweighs food in the long run. If you have lots of plains hills around you, each of these will be pumping out 5 production during mid-game, each with 1 food. A couple of good food squares and you have a powerful city at a size of 7. If I start on a coast and my first few barb moves identify mountains and city states I am significantly protected from both barbs and civs and starting scout builder settler may even be possible. If I start in the middle or I meet a civ in a few moves its now often slinger, slinger, slinger slinger, slinger, slinger clubman. clubman. Thats not a formula, just some idea that gives you the message.
If I am slinging away happily I will certainly go craftsmen route. If I am alone in the world I may sometimes choose to beeline early empire for faster settlers.
10 in 100 is just some vague measurement. I find 8 is fine often but more helps on deity in particular.
Good starting resources and either a flexible approach as to what to do or restart if it does not suit seems best way to play deity. I do like the former as it is more challenging and interesting to not know what victory you are going for at the start.

Question 2
From the formula thread at the top of this forum

District cost (production)
=[60*(1+9*Larger of [100*(Number of Techs/67 OR Number of Civics/50)]/100] * 75% IF you have less of this district than the "average" (either average of each type of district you have OR the average of this type of district over all players... conflicting reports)...
*50% if it is a UD
[ ]..round down

So it is about being below the average on the board for a 25% discount and 50% discount for a UD

Question 3
I say about 15 overall apart from tiny luxury collecting cities is about where I stop because as you say it gets a pain. I would rather a few more turns for a science than more management.
CV
8 is around optimal on Deity for culture speedily and it does seem like 1 per difficulty level apart from maybe 3 is the lower limit. Distance apart is to me a step too far. You build where it makes sense but over lapping has advantages like swapping tiles when needed to boost one city then the other. Just be flexible with this one. And getting Kumasi is the powerhouse of the early middle game.
SV
Tricky, I did well on Emp with just 10 last night but was playing Fred Barby Roasta (Germany). More is often merrier up to 15-20 depending on level. A highly skilled chap said 20 was great and fast on deity because the other civs had a lot of cities to steal. Sometimes its faster on deity because you can steal ready made and partially developed cities rather than found slow new ones. So 10-20 for SV.
RV
A very flexible number you do need a fair amount of faith unless playing a dom faith victory where you just cleanse each city you take with an inquisitor. You can try fast and sneaky if you get Yerevan so need few in this case but is a gamble.
Dom
2 approaches.
Genocide - Take everything, damn the negative modifiers and complain that so many cities made it hard to micromanage, warmongering and weariness penalties are too tough and the game is too easy. Its a very simple game, but different maps and some difference in army/navy it can be fun just to seriously own and is a great game. Sadly it can become a bit easy after a few times.
Capicide
Taking other peoples capitals is what it is all about. It is possible to do this and leave most of the civ intact. yes you get like -18 for having one of their cities and some penalties but it is about managing all of this with things like +5 for freeing a city state says forever (only one CS sadly). Just how many cities do you need for this approach?... Well all of the first civs cities you take and build science in quickly because this strat works well with a pushed science but maybe CD first, prod is king but gold is god. Maybe half the next civs cities... you may have to build the odd other to fill some annoying space or get a resource you need. For dom I tend not to build many farms, just prod. Maybe in one or two cities I know I can get great production from I will push population. This means you do not need many ED and can just concentrate on Commercial, science and ..maybe harbors (well I like harbors and we are all allowed to sin a bit)

A solid strat?.... well to me its about playing and learning the tricks and being flexible with them. the same formulaic game can get dull.
Early game expand, not necessarily with any strat but the longer you leave a decision the harder to change. Ideally by about 80 you need to know what you are doing. Then is about concentrating on that strat using all the tricks you can muster to better your chances.

As most good players will say... if you get to about turn 100 in good shape you have pretty much won as long as you keep an eye on other civ victories. Out expanding is important because they have HUGE benefits at deity against you and you need to catch up by using more, better placed districts and good card use and play. On deity you will still be lagging at turn 100, quite badly.... but from there you start snowballing.

Big Tip: Do not be afraid to restart... you are going to invest hours in a game, you may as well give yourself a better chance at the beginning.

 

[Victoria]

I have a hard time understanding any possible pros of having less stuff (cities/districts) as opposed to having more stuff,

The pro is just speed. And that does not mean you are fastest with 1 city. I have done a few tests using same map and civ and there does seem to be a sweet spot, more cities slow you down getting them all up to speed.
Try it some time on Prince where you know you should do OK. Save at turn 1 and go for a Cv or SV with different numbers of cities but the same build plan.

 

[Random Dude]

A lot depends on your map, a BIG amount depends on your map. Probably one of the largest things is your starting position.

I did a couple of dozen restarts just to see how the early game plays out, and noticed that the biggest difficulty drop comes from reducing the map size. Normally I play as random on 8-player continents with everything set to default, and on deity starting in the middle of a desert sandwiched between two aggressive neighbors is plain damn frustrating, it's actually quite challenging on immortal after the latest AI tweak. But on 6-player maps It's often on the side on a landmass where the plan is pretty straightforward: explore away from the ocean, but coast = safe and is a fallback area. It seems that playing on a smaller map can be a nice intro to the next difficulty if it's too hard initially. It also plays faster.

A very flexible number you do need a fair amount of faith unless playing a dom faith victory where you just cleanse each city you take with an inquisitor.

Ooooh, nice idea, I never have thought of that. I found "standard" religious victory quite boring, but purging entire religions by force sounds like much more fun.

Big Tip: Do not be afraid to restart... you are going to invest hours in a game, you may as well give yourself a better chance at the beginning.

More like "multiple evenings" rather than "hours", but while I agree with you in principle, I don't normally restart/reroll in games for personal reasons - it feels like borderline cheating and wins feel less satisfying, besides, losing can be satisfying too and adds to the learning experience, it it's not 90% losing, of course.

I have done a few tests using same map and civ and there does seem to be a sweet spot, more cities slow you down getting them all up to speed. Try it some time on Prince where you know you should do OK. Save at turn 1 and go for a Cv or SV with different numbers of cities but the same build plan.

Will definitely try. For now, getting to 8-10 cities ASAP with at least 8 units and then evaluating everything and making a choice between economy/tech/more expand/more military seems to be the most solid way to go.

@Victoria, your post was very helpful on both the specifics of Q3 and in general. I think I'm overthinking things way too much, and should go with the flow more rather than try to come up with the "perfect" strategy and make it work.


Just for the record, current game (random Gorgo, 6-player continents):

Spawned next to Sumeria, attacked them right away and took them out with 3 warriors and 5 archers. Found Egypt next being completely trashed by horse barbarians, had to clean up the whole mess. Met America, in a few turns he says "thanks for keeping peace on this continent", not sure if sarcasm, but decided to calm down with the conquest and go with the currency line. Somewhat cut Teddy on his part of the continent without settling too close to him. 8 cities at turn 88, 4 more settlers around 100, discovered the second continent and two other civs there. No idea on a victory condition, maybe should try quick "capicide" domination, it feels like it can be the quickest one here.


If anyone stumbles across this thread and finds it interesting, here's some random stuff I also learned recently:

1. On the early army: 3 warriors and 5 archers seem like the "sweet spot", able to defend and take down almost everything, and hit "have 8 units" eureka. Siege with 2 warriors sometimes possible, sometimes not (hexes are not controlled across the river).

2. There's a clear chioce between a peaceful (currency/commercial) and not so peaceful (iron working/horseback riding/encampment) tech paths. It's a good idea to look at available iron/horses before choosing.

3. Barbarians can be your friends, don't kill them! Seriuosly though, killing the first camp can be useful for eurekas (slinger kill, kill 3 barbs, clear an outpost), but the next ones can be left alone if they aren't right next to you. Scouts are afraid of military units and can be bounced back like a ball in Arkanoid. With two units, it's possible to divert a scout in a direction where you think the other civ is, and then they are busy with barbarians for a long time.

 

[Chibisuke]

I have been playing deity for a while now. Almost all of my games, except for I think my first game, have been on deity. I have attempted the peaceful settling game on several occasions and it is rarely successful in the way that you would want. The biggest problem is that the AI gets so many advantages at Deity that you simply can't build settlers fast enough to catch up even if you have a continent all to yourself.

As was mentioned, they are already starting with 3 settlers and I believe 6 troops. They also have I think 5 techs at the start. Plus they have large bonuses to production and gold so they can buy pretty much anything they need and crank out settlers much faster than you. What do they do with all this? Most AI seems to follow the path of (1) use troops to find and conquer a city state or closest civilization neighbor within the first 30 turns; (2) crank out settlers from one city and start expanding territory that way; (3) start bulding or purchasing strong early units such as swordsmen or horsemen once available and put up walls about the same time.

(1) and (2) will cause AI civs to grow at a massive rate that you cannot overcome simply through peaceful settling. The leading AIs will quickly have 10 - 15 cities before you can manage 7 or 8. And remember, they have HUGE natural bonuses to research, production and gold that will grow exponentially as they expand.

You could probably survive to the end of the game peacefully, and might even be able to grind out a win very very late after being behind for all of the game, but it will be more harrowing since at any moment a technologically superior AI might decide to send in a large army of advanced units and crush you.

It is much better to just go on the attack right away and try to take out at least one civ within the first 50 or 60 turns. That will allow you to at least catch up on city count if not surpass most AI civs before turn 100. Then you are in a decent position to not fall far behind in tech and might take an early lead depending on other factors.

Another suggestion I have is use your first warrior to find the closest civ and once you find them, assuming you did it before they founded their 3rd city, figure out where their last settler is (often it sits in the capital for a few turns until they decide where to settle.) Once you find it just camp with your warrior until it comes out. The AI does not usually give it an escort. Capturing that settler should be a high priority. So once it comes out, you should declare war (assuming it isn't surrounded by the AIs warriors) and capture it. That will help your growth hugely.

 

[Victoria]

@Random Dude I think you hit the nail on the head with the flow thing.
Some games the flow just feels better and works well. A good example being a builder first just does not flow but slinger slinger builder seems to get you craftsmanship nicely. While I could very rightly argue monument builder.... I suspect the third build in this case would be a tombstone.

Midgame is a little different, you need to break the 1-100 habits and get a new flow around your target but certainly if it does not feel right it often is not.