Von Münchhausen IV

[Deau]

Guys, can we please try to use spoilers for SS.

Please! Its painful to try to read comments scrolling through un spoiler tagged SSs.


@Omni, Will do unless there is suddenly a huge debate over something.

 

[Deau]

I don't think I've ever fully understood food conversion during settler production. I've lost a lot of turns through hundreds of hours of games by not understanding that. Thanks for the explanation.

Just to be sure we're all talking about the same settling sites, I dotted avl8's map. Is this what everyone was thinking? Deau, are you suggesting that red site next to the cotton or the green site?

Spoiler :

I'm suggesting the green spot for the extra unique luxury happiness (marble) and because the tiles far east are all hills when this spot already lacks growth some so the deer/cows earlier in border expansion will be better.

 

[NotSure]

I should have thanked Deau for motivating me to finally get some clarity about settler production. I assumed that food was always converted at 4:1. Glieze went through the trouble of figuring out the exact conversions last month. Here is what he came up with.

Negative-0 food: +0 from food surplus
+1 food: +1 from food surplus
+2-3 food: +2 from food surplus
+4-7 food: +3 from food surplus
+8-11 food: +4 from food surplus
+12-15 food: +5 from food surplus
+16-19 food: +6 from food surplus
+20-23 food: +7 from food surplus
+24-27 food: +8 from food surplus
+28 etc (seems it continue with a 4:1 conversion rate)

Here’s the thread if anyone is interested:
http://forums.civfanatics.com/showthread.php?t=499930

If I understand this correctly, and settler cost is 106 hammers, then moving two citizens on hills purchased in Deau’s set will speed up settler production by 8 turns each (14 turns on four hills and 22 turns on two hills). If we farm just one tile, however, the number of saved turns is five. If we farm one tile and work the cattle, then we only save one turn on each settler after buying those hills (zero turns if the cattle are improved with a pasture). This makes me think we might be better served by building another worker while we improve those tiles. Then we can build two settlers and spend the gold buying the GBR tiles in our second city, or on a third worker.

Do I have this figured right? Is it more important to get a second city out ASAP? I’m still thinking we might be better off holding on to that gold for the moment.

 

[Deau]

You are missing the 1h from city tile(settled flatland) and 3h from building.

This is actually the thread that I've been linking left and right in S&T forums whenever a similar discussion about opener would occur. I've had a game case where someone said to build settlers before the granary because otherwise you are just wasting 1gpt and delaying the settler whereas on certain starts, it is possible for the granary to provide a few extra hammers.

E.g. size 4 pop and best tiles are 2 wheat and 2 sheep (hills). Assuming flatland settled capital, pre granary you would be running 10HPT during settler. Adding the granary would push that specific start to 13 HPT going from +0 food to +4 food.

However, as a general statement, hills come so late through natrual border expansion that you should have been done building settlers for a long time by the time they become availible. Assuming you are to build 3 settlers, purchasing those tiles will grant you an extra 12 turns of growth/production elsewhere, easily allowing to hard build the worker in that turn frame.

Another way of seeing it is that the 2 hills purchase will net you +4 production for 24 turns of building settlers or 96 hammers. This doesn't even compound the hammers/growth/science gained from settling each next city 4/8/12 turns earlier respectively.

You don't gain both previous paragraph's yields, only one. They are just different ways to analyze the return on investment.

Long story short, unless you ought to expand slowly to do NC pre expansion or because there are plenty more good settling spots than however many cities you ought to plant, purchasing hills is a very smart move.



*edit* Typically, if I look towards forward settling very far, I will often skip on hills purchase and actually work GPT tiles so as to save the worker-movement turns to get to the forwarded location. That is not quite the case here.

 

[NotSure]

I really screwed that up. I didn't account for the loss of food either. At the risk of digging myself into a deeper hole, I'll try again:

Copper hills only - 8 (0 from food surplus), 14 turns.
4 hills - 12 , 9 turns.
Copper hills w/ farm & pasture - 11 (2 food surplus, 1 pasture), 10 turns.

Looks like a good move I suppose. We do get the second city quite a bit faster and I suppose we can get the gold back meeting CS and selling resources.

 

[Optional]

I never knew about the food conversion maths, but looking at the save, producing a settler with the current tiles would take almost 12 turns (11.77). That's working the 2 copper-hills and the forest-silk.
With an extra hill bought for an extra citizen we'll get 2 more hammers per turn and shave off 2 turns (9,63).
If we buy 2 hills we can bring the citizen from the forest-silk onto a hill as well and shave an extra turn off (8.83 turns to produce the settler). But we lose 2 gold per turn from not working the silk on top of the money we spent on the tile.
If you want my view, I find buying the second hill for the 1 extra turn off a settler rather expensive. Then I'm basing my calculation on keeping the silk tile forested while building settlers, and improving the copper first, which is what I would do.

With building 2 workers soon I'd want some more protection from units as well, I wouldn't like to improve tiles only for barbarians to have pillaging parties. They come with horses and axe-throwers now. So a settler before a second worker is still my favoured option, otherwise a second town will come incredibly late.

I find the green position on NotSure's dotmap a good position. I see some other possibilities, like a combination of a coastal town and an inland town, but that area there I would like to claim rather soon.
In the north I suspect to be some iron deposits, I might want a Great Barrier Reef town further north for that reason, also taking in a desert hill and some plain-forest further north, but no position there looks really great. Without One with Nature for gaining some faith from the reef I wouldn't be eager to rush a settler in that direction, but with it, okay.

 

[Deau]

I really screwed that up. I didn't account for the loss of food either. At the risk of digging myself into a deeper hole, I'll try again:

Copper hills only - 8 (0 from food surplus), 14 turns.
4 hills - 12 , 9 turns.
Copper hills w/ farm & pasture - 11 (2 food surplus, 1 pasture), 10 turns.

Looks like a good move I suppose. We do get the second city quite a bit faster and I suppose we can get the gold back meeting CS and selling resources.

You could very well consider mining 2 hills instead of the 2 food in the hills purchased scenario to push 14 hpt instead...

You also actually need to build Granary to get 2 food surplus with 2 copper/cow/farm setup as otherwise your net empire base food is only 8 (cow 3 farm 3 cap 2) and that corresponds to 0 surplus for size 4 city.

Since we have to also hard build a worker before, adding granary to the pile will make settler production come in very, very late. We may have to delay regardless due to happiness block but still.


Realistically though, since our warrior is moving back up north/west and the scout is miles up north, we are still very, very far from 310 gold. We won't meet any additionnal CSs with scout for a while and they will all be worth 15g by that time making hoarding much less tempting in my book.

On the other hand, getting at least one settler out slightly earlier to forward settle alex and build a caravan for happiness/gpt might earn us a DoF and 240g lump from copper sale.

Anyway this is not set in stone, it is just my view on it. I'll gladly not purchase hills if such is the median opinion but I'm hoping I can teach a few tricks here and there along this topic of discussion ;P

 

[Deau]

I never knew about the food conversion maths, but looking at the save, producing a settler with the current tiles would take almost 12 turns (11.77). That's working the 2 copper-hills and the forest-silk.
With an extra hill bought for an extra citizen we'll get 2 more hammers per turn and shave off 2 turns (9,63).
If we buy 2 hills we can bring the citizen from the forest-silk onto a hill as well and shave an extra turn off (8.83 turns to produce the settler). But we lose 2 gold per turn from not working the silk on top of the money we spent on the tile.
If you want my view, I find buying the second hill for the 1 extra turn off a settler rather expensive. Then I'm basing my calculation on keeping the silk tile forested while building settlers, and improving the copper first, which is what I would do.

With building 2 workers soon I'd want some more protection from units as well, I wouldn't like to improve tiles only for barbarians to have pillaging parties. They come with horses and axe-throwers now. So a settler before a second worker is still my favoured option, otherwise a second town will come incredibly late.

I find the green position on NotSure's dotmap a good position. I see some other possibilities, like a combination of a coastal town and an inland town, but that area there I would like to claim rather soon.
In the north I suspect to be some iron deposits, I might want a Great Barrier Reef town further north for that reason, also taking in a desert hill and some plain-forest further north, but no position there looks really great. Without One with Nature for gaining some faith from the reef I wouldn't be eager to rush a settler in that direction, but with it, okay.

I agree at least partly. We will need to chop the forest shortly to get the happiness for 2nd/3rd/4th city. I was hoping to chop it on a 2nd worker to partly offset the worker purchase delay (I think I mentioned that in the first post discussing this). Partly also because workers can't benefit from the extra yield from starving during settler production (making it better to get 20 prod on the worker than the settler).

I'm fine with either/or. I think purchasing 1 hill and delaying forest chop might be alright if we opt for a build order like current worker/settler/worker/settler/settler or something similar. By the time we get to the last 2 settlers, cows should be improved and probably both coppers too so the net gain from a single 4th hill tile will be small saving us anywhere from 50-100g

 

[NotSure]

Anyway this is not set in stone, it is just my view on it. I'll gladly not purchase hills if such is the median opinion but I'm hoping I can teach a few tricks here and there along this topic of discussion ;P

I play SG games primarily to learn, so I appreciate the discussion. I can see that it takes a while to get good production from the combination of food surplus and hammer tiles. Considering the goal is to block off some land asap, I think your plan is the best way to go. I'll let you play now.

 

[avl8]

Buying hills for a faster settler production is my learning too.

Thanks, Deau!
Ive already used the trick in 3 GMR games I'm playing now

I've spend some time analyzing the map, here is my 50 cents:

City placement. I agree on green spots mentioned earlier, especially on the one with 2 TGBR. I also agree that a choke-point city to the south is great.
Though I think we should first scout the land west of the capital.
Atm we got the fog instead of clear data. There are plains and wheat in that area. I guess we can have a different luxury there, probably a sea based.
At the moment we should settle 2xTGBR city first to get our religion as soon as possible. Alex is going religion too, so its him or us to conquer the minds around. And these city-states around already crying "get some tithe from us". This one will also improve other stats, 2x2 beakers being the most important at the moment.

We might need a Plan B in case AI gets One with Nature first. In this situation we can have a city with 2 copper and banana on the coast first. I support the idea of cargo ship growth, 2xTGBR can still provide decent yields for some gold, being 3rd city in this case.

Hill purchase. I support the idea! Its really a good way to keep up with AI after a really weak start. And I think we need at least one hill near horses, so we put a get pasture. 4 gold per turn is quite good to get our gold back and get our next purchase as soon as possible.

Build order. I think we need to settle our cities as soon as possible. A worker is needed both to get happiness and to provide a cash flow/production from luxuries and horses. And we still need a second warrior or scout - to confront barbarians and escort the setter. We got 2 barb camps nearby - one to the east and second being somewhere in the west side.
One barbarian came to the capital from the west side on my watch. Eastern one could march to some CS, but next one can come to our worker/settler. More camps can pop to the north during next 10-20 turns.
I prefer a scout over warrior cause we're playing on pangaea-like map. And finding all AIs and CSs improves both our research and cash-flow. Plus we get all CS'quests sooner and probably complete some.

So I think worker-scout-settler is a nice way to go. As a crazy option - maybe even buy a scout now so he can move south as soon as possible, while first one will clear all fog to the north and escort the settler.

What do you think, guys?

 

[Optional]

I think we should first scout the land west of the capital.
Atm we got the fog instead of clear data. There are plains and wheat in that area. I guess we can have a different luxury there, probably a sea based.

The random map generator only places sea luxuries if a nearby civ has a coastal start position. If we find Alex south of us and like us he's not coastal, then probably we can forget about sea luxuries. But of course we want to see what tiles are there...

Hill purchase. I support the idea! Its really a good way to keep up with AI after a really weak start. And I think we need at least one hill near horses, so we put a get pasture.

Yes, that's a good point. If we improve that pasture it's as good as an unimproved hill for settler production as well.

I don't think I've ever seen an AI choose One with Nature, so I'm not afraid we can miss that. The jungle pantheon I've seen chosen.

 

[Deau]

Alright, as promised I hit my set tonight for Omni. My write-up will be somewhat boring. I uncovered like 12 tiles from having to travel warrior back north and scout south -_-.

T20 - Turn was done, hit next turn!

T21-T23 Marsh, jungle, hills. When will this end! (warrior still hasn't uncovered any fog tile). Scout also coming back slowly. Wasted a turn moving around the barb camp east of cap so that we don't lose our only scout later. Mining completed. I picked BW next with the similar train of thought as for AH. Even more so in hopes to see 2-3 iron within 3 tiles of the plains/GBR city position to offset the total lack of any other hammer source. If we go GBR/One with nature, we may want to consider sailing+optics detour. Its not like we'll get cities to build libraries anytime soon, we are luxury diverse starved.



T24 Pop4 - Time to work! Locked a copper tile. Now 6 turns to worker. I won't even get to play it -_-. 2 turns from a pantheon if we are lucky.

T25 We were not lucky. A civ got a pantheon, cost increased to 20. No harm though, they picked +30% ranged strength as usual. Trad opened, border to cow tile brought down to 3 turns. That will allow to swap 2 grassland for cow/silk.

T28 Cow unlocked, tiles reallocated and locked. Horray I found an ancient hut with warrior move. We are only 2 turns from pantheon however so instead of jumping on it immediately, I will leave it untouched for *hopefully* the next pantheon that we will get on T31. This way, if we randomly pop a faith pantheon, we will get +60 towards religion instead of +20. I am more and more leaning towards the GBR/one with nature setup.

T29 Moved warrior around the goody hut. I see fish/marble/cow lined up. Not an outstanding spot but there is more fog to unveil. Met Wu Zetian between turns. She already teched writing so we can sell embassy whenever we want. Since I don't see her borders anywhere yet, I've waited probably for at least another ~5 or so scout movement turns. If she is far enough not to forward settle us, let's do it.

T30 Worker complete. Moved him on a copper hill for mining. Purchased only one hill. The one that should unlock the horse tile shortly. I went with optional's proposal for 1 less hammers yet 2 gpt more working silk and 3 hills. I can't wait to see what's left near the warrior's position for a potential city. I intentionally put my mouse pointer on the tile I believe could be a potentially good 4th or 5th city given the current information. It gives us 3 far reach jungles across the ocean/mountain. I think it is somewhat lackluster though. I would love to see another lux or 2 more fish access NW.

Here are T30 SSs. There was really nothing else unveiled before from T20-T28 and the SSs clearly show that I skipped the goodie hut -_-.

Spoiler :

Hopefully pantheon next turn!

ROSTER
NotSure
avl8
Deau
OmniPotent42 - up next
mad-bax
Optional


I heard it would be great if I added the save file...

 

[Deau]

The random map generator only places sea luxuries if a nearby civ has a coastal start position. If we find Alex south of us and like us he's not coastal, then probably we can forget about sea luxuries. But of course we want to see what tiles are there...Yes, that's a good point. If we improve that pasture it's as good as an unimproved hill for settler production as well.

I don't think I've ever seen an AI choose One with Nature, so I'm not afraid we can miss that. The jungle pantheon I've seen chosen.

Improved cow is only 1h but even then, you are realistically not improving a hill in the same time so its difficult to really consider the pasture benefit in the equation. Unless you play as Attila.

 

[Deau]

On a side note, I meant to ask. I played Portugal yesterday for close to 200 turns to test the map (and it turned out to be a great/easy map) and I somehow managed to build HG in capital somewhere around T90. Is that "normal" for Immortal games? I mean I've seen similar odd cases on deity and the civ distribution on the map helped (Assyria/rome/Zulu/Suleiman on 2nd continent, Portugal/America/Mongolia/Morroco on ours leaving only really 2 possible AI civs for wonder spamming/trad). But I would've expected it gone somewhere between 60-80 based on deity timeliness.

 

[OmniPotent42]

On a side note, I meant to ask. I played Portugal yesterday for close to 200 turns to test the map (and it turned out to be a great/easy map) and I somehow managed to build HG in capital somewhere around T90. Is that "normal" for Immortal games? I mean I've seen similar odd cases on deity and the civ distribution on the map helped (Assyria/rome/Zulu/Suleiman on 2nd continent, Portugal/America/Mongolia/Morroco on ours leaving only really 2 possible AI civs for wonder spamming/trad). But I would've expected it gone somewhere between 60-80 based on deity timeliness.

It is a bit of a gamble now that most AI's go for Piety or Honor, with only a small percentage starting with liberty or tradition.

So we decided on the spot next to GBR? And the One with Nature pantheon, right?

 

[Deau]

It is a bit of a gamble now that most AI's go for Piety or Honor, with only a small percentage starting with liberty or tradition.

So we decided on the spot next to GBR? And the One with Nature pantheon, right?

That's what my intuition tells me as otherwise getting a religion is going to be an extreme stretch and since we are very isolated, if we don't have a religion, it will take years before we get spread to.

I'm not sure there was a concensus.

 

[Deau]

O I forgot to mention. Until we have all our luxuries improved, from the moment we have at least one copy of a luxury we own, please check ALL available luxury duplicates from neighboring civs every turn. With so many fewer DoWs between AIs in BnW, you basically have exactly 1 turn to trade or they will trade it to someone else. If we can trade our first 2-3 copper immediately to some AI and keep the trade rolling, we will be fine on happiness. However, if someone misses an opportunity, odds are we'll have to sell for 6gpt and we will be screwed on growth.

 

[OmniPotent42]

That's what my intuition tells me as otherwise getting a religion is going to be an extreme stretch and since we are very isolated, if we don't have a religion, it will take years before we get spread to.

I'm not sure there was a concensus.

I guess it would make more sense to focus on infrastructure and culture first. Religion is optional, I guess. I guess if we just went through tradition normally we'd be okay, as the #1 things we need are food and science.

 

[avl8]

Nice set, Deau! That patience with an ancient ruin is really good move!
We got already lvl 4 capital and got all necessary techs, so the only one outcome that we need is bonus faith. So I think its a good idea to check if we get a pantheon, and if we have one on t31, get the ruin and see if we get our prophet faster.
Really a very insightful game we're playing, guys!

I think One with nature is better than Sacred path. OWN got potential to get fast religion benefits and convert it into everything we need for a CV later in the game.

I agree on the "ruin" city spot Deau pointed for a 3rd-4th city, it has a nice potential with fishes, jungle, marble, cows and a harbor all-in-one. It will also have a nice hill farm near the lake, which is really good for early growth.

Legalism seem to be an obvious choice since we still lack monument. Next social policy is still a very interesting choice. With a settler spam during next 10-15 turns Monarchy or piety opener both seem to be a good idea for me. I still don't understand is it worth going Landed elite for a faster settler production, or Monarchy is better.

I still think that one scout is not enough for a pangaea-like map. Early exploration have a huge impact on the late game, especially when we started so slow. Every AI we met lower tech costs, every CS we met will get us some quests, every natural wonder will help us grow. And we probably need another religious CS to improve our faith income. Vatican got a reealy hostile/bad Pope this time
So settler-scout build order looks very promising to me.

 

[Optional]

Improved cow is only 1h but even then, you are realistically not improving a hill in the same time so its difficult to really consider the pasture benefit in the equation. Unless you play as Attila.

I wasn't speaking about the cows, but the horses, as I was responding to avl8's remark that buying a hill would get us the horses soon...

I meant to ask. I played Portugal yesterday for close to 200 turns to test the map (and it turned out to be a great/easy map) and I somehow managed to build HG in capital somewhere around T90. Is that "normal" for Immortal games?

I've never seen it go that late, I consider Hanging Gardens impossible to get, so I'm surprised. But I've seen the Pyramids not being built in the first 100 turns since BNW.

I wonder how soon we are planning the settler for our third town. To build a unit inbetween seems a wise choice. Once Bronze Working is researched (I didn't open the save again, but that's probably researched when the settler is built) we can make a final decision on where we want our presently being built settler to go.

A thing I want to ask is; for a cultural game, is it a good idea to smash barb camps asap, so that more come up and you get left with more cultural artifacts later on in the game?

Omnipotent42 is up. It is a good habit that he person who is up confirms that he's got the save and posts some plans for his set.