Broad brush early game strategy talk

[Adreno]

Since the previous thread derailed into a flame war and got closed, I thought I would open a new thread about early game strategy in 6-player FFAs with ancient start. I'm still learning the game and I intend to use this thread to talk about multiple ideas and ask questions without starting 1000 different threads. Hopefully this will remain about early game strategy and not derail into anything else.

Hokay! First post! Today was the first day I ever settled on snow! Had some really good terrain, unfortunately England was next to me and I managed to pretty much suicide to longbows. In the future if I have China/England near me I should really consider a warrior/swords rush before they get to their special units.

Anyhow, interesting city placement decision:



I wanted to build 4 cities without sacrificing any yield from bonus resources (if you settle on a bonus resource, the yield is the same you would get without the resource. I don't know if it's improved by granary/stables/etc but it's still not the complete yield).

So I wanted to settle Ankara on the coast near the Pearls. I couldn't settle on the deer because I would lose the yield. If I would have settled on the Tundra above the pearls, I would have had to settle the 4th city on the upper cattle, losing the yield from the cattle (because you can't settle cities within 4 tiles of each other).

It kind of dawned on me that it doesn't matter if everything south of Ankara is unusable land. I still get the 2 food 1 production yield for my city base, and I can still grow the city to 6 pop on great tiles, possibly loan tiles from Istanbul, and beyond 8 pop I'm gonna have citizens working the university anyway. My 4th city had great tiles as well: 2 cattle to start, expanded to the wheat on pop 3, and later on I was loaning some river tiles from Istanbul.

In civ3 I used to build cities the "optimal distance" from each other so that each city would get the max amount of tiles - without building cities too far so that some tiles between cities were left unused. I've found that in civ5 you can micro really well if cities share some tiles. This way you can utilize all the best tiles and you have more room to maneuver when you hit unhappiness and want to maximize gold/production.

***

Unrelated, I also realized you don't want to produce great engineers in early/mid game. Since each GP costs more than the one before, producing a non scientist will limit the total number of scientists you will be able to produce in the game. You can still take a GE from the liberty finisher, and you can still produce GE/GA points from cities as long as you don't complete the units. In an optimal scenario, you would have produced only scientists up to modern era, and you would have a lot of half way built engineers and artists. When all the crucial techs are discovered, you would switch from producing scientists to producing engineers/artists. While you're mostly burning GP for golden ages, the production/culture yield at that point benefits you more than the science yield.

 

[KM]

Some thoughts. I think the screen dump is good for city placement discussion. I would have placed my cities different. If you look towards england and the mountains there is a river. yoou have not scouted fully there but it looks like it is possible to place a city there. the place 2 tiles from silver between two mountains. With a gran and mill it would be a good producction city. And it would be able to chare some of the tiles with capital hen chifting growth and producction between cities.

I fnd that you notrmaly dont want to take any echs towards sailing early so the pearls will steal to much producction to get anyway and Ankara is in a really spot. If any I d put a city between the wheat and cheep on the thundra and palce other cities forward towards england.

Regarding GE, I sually fnd that a free GE early is okey even very good but late on you only want GS. So one GE from GE points is not to bad sinc u can use it to snatch some ealry wonder that an give u a good boost when u most need it.

 

[Adreno]

Some thoughts. I think the screen dump is good for city placement discussion. I would have placed my cities different. If you look towards england and the mountains there is a river. yoou have not scouted fully there but it looks like it is possible to place a city there. the place 2 tiles from silver between two mountains. With a gran and mill it would be a good producction city. And it would be able to chare some of the tiles with capital hen chifting growth and producction between cities.

I fnd that you notrmaly dont want to take any echs towards sailing early so the pearls will steal to much producction to get anyway and Ankara is in a really spot. If any I d put a city between the wheat and cheep on the thundra and palce other cities forward towards england.

Regarding GE, I usually fnd that a free GE early is okey even very good but late on you only want GS. So one GE from GE points is not to bad sinc u can use it to snatch some ealry wonder that an give u a good boost when u most need it.

Thanks for the input! You're right I should've scouted the mountains better and expanded forward instead of back. I actually planted my 5th city later in the exact spot you described (turns out it had marble on the other side) - and I built granary and mill real early to grow it.

The GE question is interesting. The opportunity cost of taking a GE from liberty finisher is exactly 1 scientist, but it's harder to estimate the opportunity cost of building a GE from a city.

If you're employing a specialist to work GE points in a city, and you complete the specialist in early/mid game, here's a list of things you gain/lose:
- Small amount of unhappiness from a specialist (not sure how much?)
- Instead of working a tile that would produce 3-5 of something, you're producing 2 production (early/mid game doesn't typicly have specialist bonuses)
- When you complete the GE, it increases the points required for the next GP, and it means you will eventually get 1 scientist less. However, because this loss will occur in the future, it's not as detrimental as the opportunity cost with the liberty finisher (where you literally choose between a scientist and an engineer right now). Even if you plan to save all your scientists for late game, simply the option of bulbing one early is an asset, and getting 1 extra scientist late in the game is an inferior choice to getting one early.

Now that I wrote this out I'm drawing a different conclusion than earlier: whether it makes sense to produce a GE depends mainly whether you have to unwork a bad tile or a good tile to employ the specialist. Also depends a bit on whether you have any crucial wonders in the pipeline (a city will not produce a GE in time for HG/ND/PT). In general I would say it's a good idea only when a GE is worth significantly more to you than a GS.

For example, if you are not working a 2food2production tile because you employ the workshop specialist, your opportunity cost for 3 GE points is 2.7 GS points + 2 food + 0.25 happiness (I pulled the 0.25 and 2.7 out of my ass... the 2.7 represents "a bit under 3" because the loss of a scientist occurs in the future instead of present)

 

[KM]

well the only specialist slot u can work early for GP is workshop and u ll probably get that quite late. Soe early wonders thought provde GP points.

Seems the early techs routes for ancient seems to be after initial lux techs IW etc either go for Oracle, HS combo (GP) used usually for for GSto get x-bows or head quite directly towards x-bows getting collos up. Workshops is ust not time for early on.

so mainly first time u get a good choise between what GP u ll get and asign specialists is usually around medival, and by then it s pretty obvious it should be GS poits and not GE.

Only exception would be if u got a strategy where u get workshops early on then it might be worth it to asign specialist in capital for an early GE before only getting GS.

It would be nice to get some input on workshops. Seems players take them later now then before since everyone rush for x-bows. but might be possible to work something out to get them up early and still not get x-bows to late....


regards

 

[PrimEval]

Hi Adreno,
Nice screenshot, I think I will try to get into the habit of doing the same. KM suggested that you expand towards you neighbor, I agree with this. Your choice and rational for the placement of ANKARA is fine, but it should have been a later city (city 5-6). You should prioritize scouting around your cap, then around your neighbors cap before far wandering on the map. You would have seen that the land to the left of your cap was uncontested and free to be settled later. The land between you and England however is contested and needs to be settled ASAP.
For the specific positioning of EDIRNE, it’s great to have a city in a choke that acts as a nice boarder between you and the Indians to the north. The problem is that it appears to be on a hill, with no hills below it. This means you will not be able support it with ranged fire from the rear. Also, prioritizing city placement in that general area should have been later (city 4-5), since India is unlikely to settle a city past a city-state and a mountain range.
At the start of the game I would also have been tempted to move my settler to the gems, river, mountain tile. As I'm sure you know, observatories are amazing, plus it's a good defensive location.
-CIV on NQ

 

[Adreno]

Hi Adreno,
Nice screenshot, I think I will try to get into the habit of doing the same. KM suggested that you expand towards you neighbor, I agree with this. Your choice and rational for the placement of ANKARA is fine, but it should have been a later city (city 5-6). You should prioritize scouting around your cap, then around your neighbors cap before far wandering on the map. You would have seen that the land to the left of your cap was uncontested and free to be settled later. The land between you and England however is contested and needs to be settled ASAP.
For the specific positioning of EDIRNE, it’s great to have a city in a choke that acts as a nice boarder between you and the Indians to the north. The problem is that it appears to be on a hill, with no hills below it. This means you will not be able support it with ranged fire from the rear. Also, prioritizing city placement in that general area should have been later (city 4-5), since India is unlikely to settle a city past a city-state and a mountain range.
At the start of the game I would also have been tempted to move my settler to the gems, river, mountain tile. As I'm sure you know, observatories are amazing, plus it's a good defensive location.
-CIV on NQ

Thanks! I agree 100% about expanding forward. My settler originally spawned on the grasslands, so I already moved it once to get on the hill between the rivers. I guess I could have moved it again and settled next to the mountain to get the observatory (it would be sitting on lux in both cases) The reason I hate moving the settler is I will get all the crucial techs 1 turn later for every turn I don't settle my cap.

 

[Thresian]

if you settle on a bonus resource, the yield is the same you would get without the resource. I don't know if it's improved by granary/stables/etc but it's still not the complete yield

I don't think that's right. If you'd settled on the deer, I think you'd still have got the bonus +1 from it, so your city would have 3 and 1. And I'm pretty sure you also get the +1 from the granary. What you lose is the ability to build a camp on it.

 

[Adreno]

I don't think that's right. If you'd settled on the deer, I think you'd still have got the bonus +1 from it, so your city would have 3 and 1. And I'm pretty sure you also get the +1 from the granary. What you lose is the ability to build a camp on it.

I once tried settling on sheep and the yield was the same as it was without the sheep. You probably still get the bonus from granary/stables but that's it. I assume it's the same with every other bonus resource. Luxuries and strategic resources are still good to settle on.

edit: I just went into SP and tried what yields I get
grasslands, hills, river: 2food 2production 1gold
grasslands, hills, river, sheep: 2food 2production 1gold
grasslands, river, stone: 2food 1production 1gold

 

[Tabarnak]

I don't think that's right. If you'd settled on the deer, I think you'd still have got the bonus +1 from it, so your city would have 3 and 1. And I'm pretty sure you also get the +1 from the granary. What you lose is the ability to build a camp on it.

I once tried settling on sheep and the yield was the same as it was without the sheep. You probably still get the bonus from granary/stables but that's it. I assume it's the same with every other bonus resource. Luxuries and strategic resources are still good to settle on.

edit: I just went into SP and tried what yields I get
grasslands, hills, river: 2food 2production 1gold
grasslands, hills, river, sheep: 2food 2production 1gold
grasslands, river, stone: 2food 1production 1gold

You are both right.

Tip about deer : Since production is better than food at some point in multiplayer, cutting a forested deer on a plain is beneficial because you don't lose anything. I will let this forest alone on grassland because you lose an hammer per turn(after 13 turns ). I will cut the forest from a hill if sufficient food around. But in singleplayer getting extra happiness and food is easier so i will cut a forest everywhere excepted on a toundra.

 

[Adreno]

You are both right.

Ok I went into singleplayer again and settled on deer in 3 different locations, tundra, plains/river and plains/hill. Everything gave the normal yield (2 food, 1 production +1 from hill, 1 gold if river). I did find something surprising: settling on grasslands cattle yields 3 food 1 production. It's like every other bonus resource yields nothing, but cattle yields +1 food. Weird.

What I didn't try is if it makes a difference once you research animal husbandry/trapping.

 

[KiffeLesBiffles]

Ok I went into singleplayer again and settled on deer in 3 different locations, tundra, plains/river and plains/hill. Everything gave the normal yield (2 food, 1 production +1 from hill, 1 gold if river). I did find something surprising: settling on grasslands cattle yields 3 food 1 production. It's like every other bonus resource yields nothing, but cattle yields +1 food. Weird.

What I didn't try is if it makes a difference once you research animal husbandry/trapping.

No nothing weird here. Simple rules

Food output of city tile = Max(2 ; Food output of the tile including ressources including building that improve ressources)
Prod output of city tile = Max(1 ; Prod output of the tile including ressources including building that improve ressources)

The only exception is Flood plain. You dont get the extra food settling on Wheat flood plains.

Exemple :

1) Grassland, deer => Food = Max(2, 3)=3 / Prod = Max (1;0) = 1
2) Plains, deer => Food = Max(2, 2) =2 / Prod = Max (1;1) = 1
3) Plains, deer + GRANARY => Food = Max(2, 3) = 3 / Prod = Max (1;1) = 1

 

[Thresian]

Thanks, Kiffe, that clears things up.

 

[Adreno]

No nothing weird here. Simple rules

Food output of city tile = Max(2, Food output of the tile, including ressources, including building that improve ressources)
Prod output of city tile = Max(1, Prod output of the tile, including ressources, including building that improve ressources)

The only exception is Flood plain. You dont get the extra food settling on Wheat flood plains.

Exemple :

1) Grassland, deer => Food = Max(2, 3)=3 / Prod = Max (1;0) = 1
2) Plains, deer => Food = Max(2, 2) =2 / Prod = Max (1;1) = 1
3) Plains, deer + GRANARY => Food = Max(2, 3) = 3 / Prod = Max (1;1) = 1

Took me a while to get it but now I understand. So the food output of the city base is always 2, unless the underlying tile produces more food. And production of the city base is always 1, unless the underlying tile has greater production.

And how to apply this information into practice: Try to settle on hills to get +1 production for city base. Avoid settling on hills with a bonus resource on it, because you would lose the bonus yield. If you have to settle on level ground, avoid settling on a 2food1production tile (since u get the 2food and 1production on ANY tile u would be wasting the good tile but settling on it). However, a 3food0production tile would be good to settle on. Following the same logic on avoiding waste, it's better to settle on desert/snow/tundra than on grasslands/plains.

 

[Adreno]

I just played a 4-player FFA as Russia where I tried infinite city sprawl and synchronized growth for the first time. I didn't intend to to play this strat but there was a lot of unclaimed terrain. I did pretty well despite failing to get Notre Dame. One thing baffled me tho: I settled a city on flood plains, wheat, river (3 food, 0 production, 1 gold). Instantly after I was settled, the city base was yielding 2 food, 2 production, 1 gold. What the heck? I noticed the flood plains was turned into desert when I settled, so maybe that's why I lost the extra food from the wheat. But where did the extra hammer come from? I think all my cities had an additional hammer in the city base. Is this a Russian thing? Civilopedia doesn't say anything about it.

 

[KM]

+1 hammer in each city from a liberty policy?

 

[KiffeLesBiffles]

I just played a 4-player FFA as Russia where I tried infinite city sprawl and synchronized growth for the first time. I didn't intend to to play this strat but there was a lot of unclaimed terrain. I did pretty well despite failing to get Notre Dame. One thing baffled me tho: I settled a city on flood plains, wheat, river (3 food, 0 production, 1 gold). Instantly after I was settled, the city base was yielding 2 food, 2 production, 1 gold. What the heck? I noticed the flood plains was turned into desert when I settled, so maybe that's why I lost the extra food from the wheat. But where did the extra hammer come from? I think all my cities had an additional hammer in the city base. Is this a Russian thing? Civilopedia doesn't say anything about it.

Yes, as I said, flood plains is the only exception.

Extra hammer must come from liberty policy.

 

[Adreno]

Ah of course, forgot about that. Here's another question: when does it make sense to build aqueducts? I typicly build the aqueduct as a priority if I've completed HG or I'm the Aztecs/India. But in most games I actually end up NOT building them, because a lot of my cities stagnate in post university era when I'm working specialists. Even when my cities are growing, happiness is an issue and I always have something more urgent to build than aqueducts. I guess this is the reason I'm almost always behind in population.

Related to growth, what's the most efficient way to improve tiles? This is how I do it:

- Improve lux first, improve the best tiles after lux (prioritize horses/stone for circus/stone works)
- Time chops to early wonder building or early settlers
- Save jungles for universities, unless they have bananas
- Chop forests everywhere except tundra, or hill cities that need extra food, or strategic positions to block ranged units
- Farm river hills
- Trading posts everywhere else, or if I'm already getting a lot of gold, lumber mills and mines instead. I imagine most players don't do trading posts, but they yield +2g in GA while the lumber mill/mine only yields +1.2 hammers. Although if you research cannons before Economics, the lumber mill/mine yield improves earlier than the trading post yield. Rationalism gives a huge bonus to trading posts too.
- Roads after all worked tiles are improved, not before size 5 in any case, and depending a bit on distances.

 

[Tabarnak]

I build an aqueduct in my capital only. Rarely elsewhere unless i have another amazing site with lot of wheat and stuff. The thing is most games are finished around the industrial era or before.

 

[Adreno]

Fractal, 8-man NQ FFA:



I move the warrior south, desert beach, I move the warrior north, tundra beach.. so it's an island, right? Guess I'll do worker-monument-granary and take the wonder bonus from tradition first. Maybe I'll build a huge defensive navy with my 1/3 upkeep costs and try to go for the cultural victory for the first time in MP.

Then my borders expand. What do I find?



And that's not all. I scout a few more tiles to find El Dorado... no gold. What was that advice about scouting around your capital again? I built the 2nd city on the spot where my warrior is to connect the waters in case we have a closed ocean. I also figured since I'm ottomans I can build naval defence for both cities. But that turned out to be real boring game so nevermind.

I had a really interesting game earlier, 6-man fractal FFA. Everyone was on the same continent, but I was protected behind a mountain range, connected to the rest only by a 1-tile pathway. Rome found El Dorado at the start and started conquering everyone 1 by 1. I thought I would be able to out tech him, but I only had 4 cities and only one of them was by a mountain. Once we got near Industrial Era it was obvious I needed a new plan. We were around the same % in literacy, but he was producing more scientists, he held most of the map (more modern resources), and he was crushing me in manufacturing. Nevertheless I decided to pull off "a Coachi".

My only chance was ninjaing his capital when he doesn't expect it.

I planned my bulbs to get Panzers and Mech, and I had a ton of units ready for upgrades. Then I set off to sea with about 5 Panzers, 5 Mechs, 4 Artillery, a GG, Destroyer, Battleship, and 2 Caravels to scout front & back. It was a long trip across the map, and I was building more mechs while the units were on the way. I needed a distraction because Rome had Stealth Bombers and a ton of land units. I managed to pump out 4 more mechs while the convoy was at sea, and I intended to use those units to draw his firepower to our border (his capital was really far from our border).

Then I landed my distraction party near the bordertown, and watched the aircraft count on his capital go from 5 to 2. I made sure I moved those units first when the new turn started, and sure enough, he took the bait. After he spent the movement points on his bombers, I landed my ninja assassin convoy on his capital. Because there was limited shoreline in the vicinity of his city I actually had to deploy my units over a period of 2 turns. He had massive amounts of land units near the capital and the site quickly became a graveyard. But he couldn't kill everything.



After taking this screenshot he bombed my mechs so that 1 died and 2 remained in red health as the turn was coming to an end. He had enough firepower to take all of my mechs the next turn so it was gonna be a clickfest. I figured the city would regain some health and he would bomb the red health units first, so I planned the following order for actions:
1. Kamikaze attack with the red mech
(Let the red panzer die)
2. Bombard with 1 artillery
3. Attack with the full health mech that's standing on the hill.

BLAM! You have conquered Rome!

Not only that... but conquered with the last remaining melee unit. How lucky is that for execution?

You can see the relative strength of our civs from the minimap. This really was my only hope to win the game and I was surprised to succeed.

 

[KM]

Regarding scouting.

I dont know if it s optimal but for me this is what s working.
In duel. Move worrior towards enemy. allways good to svout out enemy capital. Also move around to try to find the natural wonders for early happines but besure to head back towards the enemy capital when the policy for liberty settlr is about to finish (Mabye u get lycky) but at least u got a chanse to block him from his first pick of second ctiy spot or deley him.

I allways build a scout after monument in my own cap even if it means deley grannery when doing gran. start. I want to be able to scout around in good time for the best second city location and If my oponent do the same thing i do. I need something to protect my settler with...


In 2vs2, 3vs3 same thing as duel... trying to find enemy cap. the exception is if i m in the back I still move worrior towards the front, but If i emeadetly spot a good second city location and want a gran. start I might skip scout until after gran.

Cton. Since there are usually two enemies (on on each side) I dont really walk up to enemy capital and park there. rather just scout with worrior to see whos closer and who they are and concentrate on trying to find out where to put a defencive borderline and in what diretion to put a more offencive stance aswell as good city locations for myself. then with the scout I just try to as early as possible take a walk around the map before I get cut of to find natural wonders and get some good idea about my enemies....