SGOTM 03 - The Real Ms. Beyond

[Iainuki]

Assuming we finish the Pyramids with a two-pop whip and that Compromise's notes about the penalty for poprushing wonders are right (we get 30 hammers rather than 44, before accounting for the stone bonus), we'll get 120 hammers from that. Clear-cutting Moscow's radius will get us another 400 hammers. Mining-chopping the hills and just chopping the flatlands will take 53 worker turns. As long we have red dot make a worker and send it back to help, we should be able to fit that in even with pasturing the cows with our first worker before 1000 BC.

I agree that we need that next settler out asap. I'm trying to determine if the fastest way to do that is to work on it directly at size 4 and whip it when we can, or whether it's quicker to work on warriors/pyramids until size 6, then whip it.

I'm pretty sure it's working on it directly at size 4. Without much overflow, it takes 5 turns to make it whippable at size 4, while it takes 14 more turns to grow to size 6.

 

[Compromise]

Comparing whether it's better to straight build the settler or to whip him upon growing to size 6. Here are my notes (rough!)

Direct settler growth:
Size 4(-8F) Working city (2F1H), fish(5F), clams(4F), quarry(1F4H), floodplains(3F): +7F, 5H
Settler in 150/12=13T

Grow till settler is whippable at pop 6:
Need 150-88=62H
62H/12=6T with overflow

Whip settler after pop6:
Grow to 6 in:
0: 2/42F @ +7F
1: 9/42F
2: 16/42F
3: 23/42F
4: 30/42F
5: 37/42F (stone connects?)
6: 2/45F (Size 5, 1 unhappy, so +5F) (might be off by a turn here)
7: 7/45F
8: 12/45F (whip anger abates, I think, so work grass hill (being mined), +6F)
9: 18/45F
10: 24/45F
11: 30/45F
12: 36/45F
13: 42/45F (Switch to settler)
14: 3/48F (Size 6, 1 unhappy, whip settler)

So, building immediately at size 4 gets the settler done a turn earlier. More importantly, it doesn't cost us any whip anger.

EDIT: Added the (duh!) case of growing the settler until it can be whipped at size 4. Takes 6 turns.

(Note: shave 2T if we apply chop to settler)

But, the cost is that growing the settler directly costs us 4Tx5H+1Tx(2x5H)+3T*(2x5H)+5T*(2x7H)=130/675H toward the pyramids.

I'll edit this in a little bit with a comment about working directly on the pyramid at size 4 (if we whip the settler and are happy-limited) vs. size 5 (built the settler directly).

Bottom line: As soon as we grow to size 4, start the settler and whip it 6 turns later when we can!

 

[Iainuki]

@Iainuki: good comments. Working the fish, clams and floodplain, agriculture is a 10T tech that will save us maybe 4T later. We probably can't afford to whip the granary if we're planning to whip the settler. If we whip the settler from Size 6 instead of building from size 4, then we don't need a granary so quickly (because we can work tiles rather than growing after the whip). If we can get agri(10T), pottery(12T-2*), then AH(15T-2*) before we need either a granary of a pasture, then agriculture makes sense. (2* is the estimated 2T savings for getting agri first).

Radical thought: if we don't need the granary started immediately, maybe we should get Animal Husbandry after the wheel? This means that after finishing the quarry roads, the worker could head north with the settler and start the cows ASAP, making red dot a lot more productive. If we're using the chop for the Pyramids, it doesn't matter much when it comes in; and the granary we can probably recoup the worker turns in getting the worker out at red dot noticeably faster.

Question: Do we want the hammers from the mining of the plains hill to go to the pyramids or a settler? Once stone is connected, the pyramids simply cost 338H and a hammer is a hammer. The connection of stone just determines when it's efficient to start the pyramids.

Still thinking past the discovery of the wheel....

That's one of the big questions, I think. Is there anything we can do to speed stuff up here? Because we're limited by whip unhappiness and don't have the stone hooked up yet, I don't think chopping the first settler helps, but it might be worth looking into chopping the second. (I.e., sending two chops into the settler so we don't have to prework it before whipping it.)

 

[Compromise]

I've whipped up a test game (it appears I started alone on an island there too) and am sitting there producing warriors.

Note: This test indicates that we only get 22.5H (fractional?!) per pop to whip the pyramids. With 23/675H it takes 30 pop to whip, with 24/675H it takes 29 pop to whip. (675-24)/29 ~= 22.448... or about 22.5H per pop. Weird. And not very efficient use of our pop.

Worldwide events not yet occurred in our game:

2020BC: Judaism founded
1120BC: Stonehenge
955BC: Oracle
760BC: Confucianism founded
745BC: Pyramids built! (Turn 117)

It takes about as long to set up 17 AI opponents as it does to hit "Enter" 117 times.

 

[Compromise]

Radical thought: if we don't need the granary started immediately, maybe we should get Animal Husbandry after the wheel?

I'm intrigued...! That might be very wise.

I'm going to whip up another sample game or two and see what kind of dates I get from it. I'd really like to see an earlier outlier date.

(But that estimate of 800BC for the pyramids was spot on here!)

 

[Compromise]

Sample game 2 dates (Reloaded 4000BC from last test game, hit "Regenerate Map", replayed): (This game is apparently not HoF-compatible )
(Alone on this island too)
1810BC: Judaism
1540BC: Stonehenge
985BC: Oracle
820BC: Great Lighthouse (I didn't wait for that last game)
760BC: Pyramids


Sample game 3 dates (alone; super-tiny island with 4 seafoods in capital's radius):
1900BC: Judaism
1150BC: Stonehenge
985BC: Oracle
640BC: Pyramids
580BC: Great Lighthouse (waited this time)


One more test: Dropped WB stone (often replacing marble) right next to every other civ's capital! (Alone again):
1830BC: Judaism (First 2 religions seemed delayed, eg Bud in 3100)
1630BC: Stonehenge
1300BC: Oracle/Confucianism (presumably CoL chosen)
925BC: Pyramids!
880BC: Empty Spanish Galley!
745BC: Moses (via stonehenge, no doubt)
640BC: Spanish Galley with archer and settler!
610BC: Great Lighthouse

 

[Iainuki]

I have some more data along similar lines. I generated an archipelago map under the same settings as this (AFAICT) with 17 civs; I picked the first, alphabetically, leader from each civilization, which gives a random sampling of Industrious civs but doesn't make everyone who could be Industrious. (We know China isn't, so I picked Mao.) Because I'm lazy, I then went into the worldbuilder and made a mountain wall around the settler/scout so I wouldn't have any annoying pop-ups coming up while I was hitting enter. I did this five times, making no other map edits. I noted cases where the Great Lighthouse fell earlier than the Pyramids, but I normally stopped right after the Pyramids were built. Here are the dates of significant events.

Buddhism FIADL 3640
Hinduism FIADL 3190
Judiasm FIADL 2110
Stonehenge BIAFAL 1390
Confucianism FIADL 910
The Oracle BIAFAL 910
Pyramids BIAFAL 280(!)

Buddhism FIADL 3640
Hinduism FIADL 3580
Judiasm FIADL 2500
Stonehenge BIAFAL 1360
The Oracle BIAFAL 880
Pyramids BIAFAL 865

Buddhism FIADL 3640
Hinduism FIADL 3320
Judiasm FIADL 2170
Stonehenge BIAFAL 1630
The Oracle BIAFAL 1420
Confucianism FIADL 1060
Pyramids BIAFAL 655

Buddhism FIADL 3670
Hinduism FIADL 3010
Judiasm FIADL 1690
Stonehenge BIAFAL 1660
The Oracle BIAFAL 985
Confucianism FIADL 865
Great Lighthouse BIAFAL 670
Pyramids BIAFAL 550

Buddhism FIADL 3400
Hinduism FIADL 2950
Judiasm FIADL 1960
Stonehenge BIAFAL 1660
The Oracle BIAFAL 1060
Confucianism FIADL 730
Great Lighthouse BIAFAL 655
Pyramids BIAFAL 580

I didn't notice any patterns with the fall dates of Hinduism and Buddhism and the Pyramids fall date. There might be a rough correlation with Judaism's fall date. (That hasn't happened yet in our game, right?)

Provided that Gyaathar didn't make edits specifically designed to encourage the AIs to go for the Pyramids, such as Compromise's "stone at every capital," here are my feelings at Pyramids completion times. If we build the Pyramids in the 1000-900 BC window we should be safe unless we get really unlucky. 900-800 BC has a probability of failure of less than 20%, but probably more than 10%. 800-600 BC has around 50-50 odds of getting them. Anything after that we still might get them but it depends on unusual luck happening in our game.

Opening up the saves and looking where the AIs had settled, it looks like as long as we get our fourth settler out not too long after 1000 BC, we should be safe from AI settlement plans. Busting the fog near our planned city sites is probably more important than settlers in the immediate future, since we don't want any barb cities spawning in bad spots.

I don't think the Great Lighthouse will be achievable, as good it would be.

 

[Compromise]

A few more notes before I really have to attend to real life:

Settler will be able to settle St Pete on 3rd turn after it finishes. Pasture will take 6T to complete (I think). Worker will get 5x4=20FH before pasture, then need 10T (10Tx7FH) to complete.

So, the setup of RedDot/StPete will require about (7 transit + 6 work) 13 worker turns before he can return to the north grass hill to chop/mine. (Note: AH may not be ready by the time StPete is founded, so the worker may not travel with the settler. ) In return, we get a worker in St Pete 20T after he starts his side job. I'd say that investment pays off quickly enough, especially since StPete can then pump out warriors (8T while growing), workers (13T) or settlers (21T).

Note: St Pete will essentially be protected from barbs by a single warrior located 1N of the cows. While a barb capture isn't irrecoverable--since the city won't be razed--it is, of course, bad. I think we need to start a warrior before the settler, then resume it after we whip the settler at size 4. After the whipping, the quarry will be hooked up and the warrior will finish quickly.

Upon getting that third sentry warrior (plus 1 MP in Moscow who can't get to St Pete in time in the event of trouble), he and the other sentries should take the positions on the sentry map I posted in post #227. (That one's not is that on the summary page right now, but should replace the other once red dot is founded.) The scout stays near StPete, I assume. Alternatively, northwestern warrior could fortify on the jungle hill for better defense at the expense of more fog. I guess the warrior could fortify in StPete, but that leaves the cows exposed to pillaging and allows more fog.

Bottom line(s):
I think AH is the best tech after the wheel.

I think the founding of StPete would be a good stop point for LK's turns.

Last minute correction: There might be a delay in the need for the worker at StPete while we wait for AH. We want it there right when we discover AH though.

EDIT: Crossposted with Iainuki. I agree with his risk assessment for the build date of the Pyramids. I'd like to see us get it finished before 900BC, preferably before 950BC. I haven't calculated, but I think founding St Pete and getting 2 more warrior sentries out there (as per my thoughts in recent posts) is compatible with that. Just the one settler from Moscow before finishing the pyramids.

And St. Pete could make a workboat for green dot since green dot will be a canal city.

 

[Iainuki]

Note: This test indicates that we only get 22.5H (fractional?!) per pop to whip the pyramids. With 23/675H it takes 30 pop to whip, with 24/675H it takes 29 pop to whip. (675-24)/29 ~= 22.448... or about 22.5H per pop. Weird. And not very efficient use of our pop.

Hmm. This has me thinking: do we want to use our third whip not on the Pyramids, but on another settler? This means it will be en route to green dot by 910 or so. We'd then have to build about 275 hammers into the Pyramids by hand, presuming clear-cutting, which would take about 15 turns total working the city center, the quarry, and the plains hill mine. (Of course, we'd spend some of this time growing, building a granary, and preparing for settler whips, so things will be tight; but I think we can probably squeeze out the hammers in between the settlers. More numbers on this would be good.)

 

[Iainuki]

I agree with his risk assessment for the build date of the Pyramids. I'd like to see us get it finished before 900BC, preferably before 950BC. I haven't calculated, but I think founding St Pete and getting 2 more warrior sentries out there (as per my thoughts in recent posts) is compatible with that. Just the one settler from Moscow before finishing the pyramids.

And St. Pete could make a workboat for green dot since green dot will be a canal city.

St. Pete can't really efficiently supply settlers until after the time-frame of this analysis: it simply has too few workable tiles pre-Iron Working. I just don't see that we can afford to delay the blue-dot settler until after the Pyramids: there's too much chance that an AI will grab that spot, since it will take us so long to even walk the settler there. As long as we're willing/able to cut enough forests, I think that's still compatible with a 900 BC or so finish date for the Pyramids. What we might consider doing is finishing the Pyramids before the third whip at Moscow for the green-dot settler, if that's workable.

We're going to want to supply the following things from St. Pete in pre-Iron Working: a worker, an obelisk (for the fish), two workboats (for the fish and green dot's clams), and as many workers/warriors/settlers as we can. 25 turns for a settler is a long time, but it would be more viable once we get the obelisk and workboat whipped so it will have the fish too.

My bottom line:

After reading Compromise's stuff, I think our best course is to start the settler so we can whip it right after Moscow's whip anger abates; research The Wheel then Animal Husbandry; finish the quarry and quarry roads then arrange to pasture the cows ASAP; and have Moscow build warriors until the stone is hooked up, then finish the last warrior and start the Pyramids. Lee, my feeling is that you should feel free to do as much of that as you want, but we should probably discuss again before deciding what to build in St. Pete, what our next tech should be, and what we should whip next in Moscow, if anything.

 

[LKendter]

I still oppose 2 settlers. IMHO is is a guaranteed loss of the Pyramids. I will try and look in detail once I get home in about 3 hours.

 

[Compromise]

Just a couple quick thoughts (not responding to the last few posts yet): Red dot settler: yes. Doesn't a settler (150H) pretty much cost the same as half the pyramids (338H w/stone)?

Also, another slight consideration is that earlier pyramids are an earlier great person. Pyramids before 2nd settler gets us a great engineer about 12 (?) turns faster.

We'll have to discuss how we want to handle our great person production and use at some point too. I assume we'll want to save an engineer for the Great Library and don't care about scientist versus engineer after that, but I haven't given it too much thought.

EDIT: Lee, I think making the settler for red dot (build warriors until size 4 (10T)?, switch to settler, whip asap (2T), finish whatever warrior is in the works (?T) while making max food is a good turnset. Wheel then AH. Move worker and sentries as appropriate. We're then at another decision point.

EDIT2: I just remembered that when I was looking at the save earlier, I realized I was wrong about when Moscow's borders would pop again. I'd been thinking it was at turn 51 (needing 100 @ 2cpt), but we're only at 88/150 culture for the next expansion. That may or may not affect my sentry coverage map near the capital.

 

[Compromise]

Not ever having been much of a mod user, I may find myself liking the HoF mod. On the options screen, I found the Civlerts options. That might help micromanagement by indicating when cities have grown or are going unhappy, etc. I'm looking forward to seeing how it works in my next turnset. I don't know if that's something anyone else might be interested in (though ignorant of) too.

 

[Kodii]

After skimming the notes, I don't think there was any intention to get two settlers out before Pyramids...

Either way, I'm against two settlers, as much as I want the blue dot to get settled ASAP.

AH after Wheel sounds good.

 

[Kodii]

I've updated the second summary post to the best of my ability. We need a definate build order, or at least what to begin with. I'll take another look at the posts I missed to see if I can dig that out.

 

[Iainuki]

I'm not sure I understand why everyone is against two settlers. The hammers are coming from two different sources: we want to whip the settlers, since we can't whip the Pyramids, which means that forests and built hammers will go into the Pyramids. Unless my numbers are quite off, we do have some slack here, and there's no real advantage in finishing the Pyramids early: if we intend to use the great engineer for the Great Library, the limiting factor in the next phase will be how fast we can develop Literature, not how fast the engineer comes out. If it turns out we don't have the production to do both, I agree, the Pyramids are more important, but if the delay isn't significant, I'd feel much safer with the settler heading to blue dot.

 

[Compromise]

We do need to discuss this 2nd settler issue.

But, is it agreed that during LK's turnset here, he'll be connecting stone, building warriors, whipping a settler (off to red dot) and starting animal husbandry after the wheel? (Heck, he's probably already done this!)

Also: I've been thinking about the GPP (great person points) that the pyramids supply. It's 4GPPx<nTurns that the pyramids are built before other option>. While that's a lot in turns, it's probably on the order of 50 GPP, and so maybe represents a few turns at the most later on when we have several wonders and some specialists.

 

[Kodii]

Well, lets see how this plays out. When we are in a position where we feel that we can whip another settler, lets see how confident we are in getting the settler and the Pyramids. If we want to do well (we're still actually "first", comparing others at the same date), then we may have to take that risk.

After all, fortune favors the bold

 

[Iainuki]

But, is it agreed that during LK's turnset here, he'll be connecting stone, building warriors, whipping a settler (off to red dot) and starting animal husbandry after the wheel? (Heck, he's probably already done this!)

Looks good to me.

 

[LKendter]

http://gotm.civfanatics.net/saves/civ4sgotm3/The_Real_Ms_Beyond_SG003_BC1930_01.Civ4SavedGame

2620 BC
I swap over to warrior in Moscow. I start fog busters playing musical chairs.


2500 BC
(IT) There are 17 civs on the map, and Mao rates the 5th largest.


2440 BC
(IT) A mean kitty barely dings our warrior.


2380 BC
(IT) Moscow has another forest by it.


2350 BC
We have absurd food power in Moscow. It could finish the settler in 13 turns naturally.


2200 BC
I whip settler #2.


2080 BC
St. Petersburg is formed at red dot. This team has me so paranoid I double checked the dot map before even moving there.

=======================

Summary:
I stopped at an odd turn number of 69, as the next key decision needs to be made. I set research to pottery, but NO beakers are invested.

The pasture is started by St. Petersburg. I realized I screwed up the timing by sending the worker to soon. That is why a road is complete for the pasture. I sent the worker there a bit prematurely, so you can yell me all you want.

Moscow is connected to stone now, so the Pyramids are a lot more viable.