[BTS] SGOTM 25 - Home Slices

[LowtherCastle]

Confirmed: Deity AI wonder builds benefit from the 5% AIPerEraModifier

The GLH, for example, costs Stalin 150 base-h, assuming I'm correct that he's in the Modern Age. With his industrious trait and OR, he needs only ~86 base-h to complete it. Here's my list of current AI eras:

Further data on AI tech levels:

Renaissance Era (AI modifier: -15%):
Churchill
de Gaulle
Mehmed
Julius Caesar

Industrial Era (AI modifier: -20%):
Home Slices
Tokugawa
Bismarck

Modern Era (AI modifier: -25%):
Stalin
Roosevelt

This is easily determined by looking at the city buildings. First notice that the above groupings have the same buildings types in the Real Save. Go into Worldbuilder in bbp's test save, home in on any primitive AI's city, add to him techs by era. When you finally add Renaissance, the buildings flip to that grouping. Next Industrial Era and the buildings flip again. Then Modern and once again. I haven't looked at how this affects the AI score values.


Two Caveats: bbp noticed some WB inconsistencies and I noticed that giving a tech that flips the buildings and then removing that tech does not flip the buildings back.

 

[babybluepants]

With his industrious trait and OR, he needs only ~86 base-h to complete it.

Good thing our archer is going to steal his worker and pillage his cow pasture.

Edit:
For GLH, I think we will beeline it when we capture Shanghai, but can’t do much more about it.

Any chance there is another precondition for an AI to build GLH - like having more than one coastal city? Based on experience, I suspect not...

 

[Kaitzilla]

Confirmed: Deity AI wonder builds benefit from the 5% AIPerEraModifier

The GLH, for example, costs Stalin 150 base-h, assuming I'm correct that he's in the Modern Age. With his industrious trait and OR, he needs only ~86 base-h to complete it. Here's my list of current AI eras:

Wow, had no idea era modifier affected everything.

 

[LowtherCastle]

Wow, had no idea era modifier affected everything.

My testing showed varying results. Settlers are 92h in all eras. Builds seem to have a 3% modifier, food 4%, and wonder 5%, although the food was slightly inconsistent in my spreadsheet. The 5% wonder is certain.

 

[LowtherCastle]

Tentative Bomber Plan 1
T21 Bombard Mongolia
T22 Strike
T23 Strike
T24 Move to Gua
T25 Bombard Hong Kong
T26 Strike
T27 Strike
T28 Move to Beijing
T29 Bombard Manchou
T30 Strike
T31 Strike
T32 Bombard Shanghai
T33 Strike
T34 Strike

Manchou and Shanghai are interchangeable. Changquing is also bombable from Beijing.

Tentative Bomber Plan 2
T20 Bombard Mongolia
T21 Strike
T22 Strike
T23 Move to Gua
T24 Bombard Hong Kong
T25 Strike
T26 Strike
T27 DoW dG, teleport to HK, Bombard Shanghai
T28 Strike
T29 Strike
T30 Move to Beijing
T31 Bombard Manchou
T32 Strike
T33 Strike

 

[LowtherCastle]

I'm also noticing how AIs road right through our culture to connect their cities to the capital. How do we want to manage this? We want some worker to road to Europe, right? AI workers will do that almost twice as fast as us, so we ought to let one do it, although we might want to guard it from barbs.

 

[LowtherCastle]

Big Picture
I think this scenario could go different ways, depending on our choices.
1. If we don't capture AI cities, they'll tech and develop slowly, perhaps not helping us much with anything.
2. If we capture AI cities, they'll tech faster and focus more on their homeland, maybe building a wonder for us.
3. Overall, I think they're going to be way too slow for our needs, no matter what. I also think Domination will take far too long. Too many cities to settle. My gut hunch is that our fastest victory is Conquest via Astro, thusly:

100 military
100 civil
58 diplo
28 tech
100 construction
34 culture (construction and culture can be balanced as needed)​

We beeline Kremlin, capture all nearby AI cities asap, settle some more prime sites while continuing to kill all Ais off, except Mehmed and Roosy (for example). We tech, trade or steal our Medieval techs as we go. We chop the Pyramids early on, run Rep and specialists to spam any GPs we need, in particular two GSes for Astro. This fuels our Medieval techs effortlessly. We keep building wonders while also running failgold as needed. Basically, everthing falls into place simply by focusing on:

A. Fast warring
B. Kremlin+Mids(+GLH)
C. Some REX + lots of worker stealing

I think this path can be done really fast. Really fast. However long you imagine this game going, divide that in half.

The reason I bring this up now is for general thinking plus considering 1) tricks for how we're going to steal lots of workers and 2) whether we capture all the nearby cities, including Roosy and Bismarck, or leave some as worker pumps, or what?

 

[LowtherCastle]

One more idea would be to build Globe in Gua or similar asap and whip wonders there.

 

[LowtherCastle]

The rationale for 100 construction rather than 100 culture is that wonders will be easy to whip in big cities at game end. This frees us to focus our GPs on two GSes for Astro and golden ages.

No need for lots of spies. (hammers)
No need for lots of cities. (hammers)
No need for lots of beakers.
No need for lots of commerce. (eps)
No need for great attack units.

Just Astro + Conquest + food-rich cities in which to whip wonders.

 

[WastinTime]

I think mids and glh are not necessary and will severely slow down the war efforts.
Both seem capturable. Is that a word?

 

[LowtherCastle]

I think mids and glh are not necessary and will severely slow down the war efforts.
Both seem capturable. Is that a word?

Okay, we'll see. Not important right now. All depends on how fast the AIs are compared to us.

 

[babybluepants]

The rationale for 100 construction rather than 100 culture is that wonders will be easy to whip in big cities at game end. This frees us to focus our GPs on two GSes for Astro and golden ages.

No need for lots of spies. (hammers)
No need for lots of cities. (hammers)
No need for lots of beakers.
No need for lots of commerce. (eps)
No need for great attack units.

Just Astro + Conquest + food-rich cities in which to whip wonders.

This is what I have been thinking, too. The wonders / culture equation, I mean.

I would still like to think that we can win by dom without Astro. Mainly because I would like to tech as little as possible. We do still need enough techs to unlock the wonders, but the less the better.

One more idea would be to build Globe in Gua or similar asap and whip wonders there.

I like that.

The other thing is, I don't think the AIs are very likely to get GP's before we kill them. We will capture at least 4 other holy cities (Paris, Berlin, Moscow and Istanbul). Should we actually try to get some GPs? Especially if we do decide with more map knowledge that Astro isn't necessary. Yan'an could get a few out.

Tentative Bomber Plan

I don't think we need to strike twice each time, but the big question is how do we get the ground units in place. The bomber is a lot faster than our production and troop movement. With this timeline, are we:
- revolting to Tao/OR?
- building a second worker in Yan?
- building a forge, followed by Kremlin in Beijing?

 

[babybluepants]

I've been thinking more about units. Chokos are awesome for the early conquests. They are nice 2pop whips, they have the best odds against the LBs and cause more collateral to the back-up archers; the first captures are of cities very close to us. However, they are not as good as 2-move units in picking off workers, and the survival rate for chokos and swords is quite high. Once we capture the nearby cities, they become almost obsolete, since we can't walk them all the way to Europe.

So, how many choko + sword do we need to capture the cities around us (including or not including Chong)?

I'm still kinda like the idea of teching HBR. We can get it in onl 15t.
Here is some testing of HA vs. chariot:

                            Chariot    Horse Archer
3.7 LB (1 strike)            23.38%    73.69%  
2.4 Archer (1 strike)        44.69%    90.34%  
3.8 PRO LB (1 strike)        17.72%    67.01%
2.4 PRO Archer (1 strike)    36.41%    87.46%
3.0 LB (2 strikes)            40.88%    89.17%
3.0 PRO LB (2 strikes)        32.45%    86.03%

The advantages are:
- we would need less units for an initial attack.
- much higher survival rate means that they are much more hammer efficient, and that they can promo, heal for 2-3t and just keep rolling.
- they are 2pop whippable with or without Kremlin.

LC suggested that we would use choko or sword built in Europe for conquest there. That may be fine, with Kremlin and hopefully decent size captured cities. But it also has potential to be slower and more awkward. We can move HAs all the way to Paris, via Moscow and Berlin, in under 20t.

 

[WastinTime]

I would normally skip forges for non-IND civs. I would consider one if you say it will not delay Kremlin. If all it does is save 1 forest, but delays Kremlin then skip it. But, maybe whipping it with OF speeds up Kremlin? Then of course we do one.
The extra happy is nice, but can't we work around that?
Only other reason for forge is probably for Colossus.

I suppose I should come up with a Kremlin plan and you guys focus on the wars. How much military are you counting on coming from Beijing?

 

[babybluepants]

If we whip it for max OF into Kremlin, it doesn't slow the Kremlin down compared to whipping something else. Edit: I don't think it does anyway, based on my calcs and testing to date, but there are lots of variables. And then we have +1 happy and +25% production in Beijing.

It depends on whether we want OR, and how many units we want from Beijing before Kremlin.

Edit: we could also just aggressively 2pop chokos and OF into Kremlin, and gift / recap the city after. We should think about what tiles to improve, as well. I prefer only iron / stone / FP watermill, if we are going with aggressive whipping. If we are spending time at pop 5 hammering stuff out, ivory watermills are good.

Edit2:
Without OR, forge saves 40 base-h from Kremlin. It costs 78 base-h to set up for a max OF whip. So, the net cost is 38 base-h. Comparatively, a choko costs 24 base-h to set up for the same whip, so the forge has an extra cost of 14h early.

With OR, forge saves 33 base-h from Kremlin. It costs 59 base-h to set up for a max OF whip. So, the net cost is 26 base-h. A choko still costs 24 base-h to set up for the same whip, so the forge has an extra cost of 2h early, but we also lose 2t to anarchy (and gain +1 happy and the OR bonus in Yan / Gua).

It’s all pretty close. I don’t think we can speed up Kremlin too much.

 

[LowtherCastle]

I would still like to think that we can win by dom without Astro. Mainly because I would like to tech as little as possible. We do still need enough techs to unlock the wonders, but the less the better.

I'm not doubting we can win by domination. I just think Conquest will be much faster.

The other thing is, I don't think the AIs are very likely to get GP's before we kill them. We will capture at least 4 other holy cities (Paris, Berlin, Moscow and Istanbul). Should we actually try to get some GPs? Especially if we do decide with more map knowledge that Astro isn't necessary. Yan'an could get a few out.

Different thoughts. First, I'll keep an eye out for specialist builds (temple or library) in the holy cities so we'll know what they're doing. Second, if we have time, we could capture the city, build the temple, gift back to them. (Doubt we'll do that, but could.)

I don't think we need to strike twice each time, but the big question is how do we get the ground units in place. The bomber is a lot faster than our production and troop movement. With this timeline, are we:
- revolting to Tao/OR?
- building a second worker in Yan?
- building a forge, followed by Kremlin in Beijing?

Funny, I was thinking the bomber was the lagger... I haven't started looking at when Tao/OR because it's not clear we want or need it yet. Wasn't planning the second worker in Yan. Whip a choko for Mongolia, then chariots for Manchou. Open for debate.

 

[LowtherCastle]

I've been thinking more about units. Chokos are awesome for the early conquests. They are nice 2pop whips, they have the best odds against the LBs and cause more collateral to the back-up archers; the first captures are of cities very close to us. However, they are not as good as 2-move units in picking off workers, and the survival rate for chokos and swords is quite high. Once we capture the nearby cities, they become almost obsolete, since we can't walk them all the way to Europe.

So, how many choko + sword do we need to capture the cities around us (including or not including Chong)?

I'm still kinda like the idea of teching HBR. We can get it in onl 15t.
Here is some testing of HA vs. chariot:

                            Chariot    Horse Archer
3.7 LB (1 strike)            23.38%    73.69%
2.4 Archer (1 strike)        44.69%    90.34%
3.8 PRO LB (1 strike)        17.72%    67.01%
2.4 PRO Archer (1 strike)    36.41%    87.46%
3.0 LB (2 strikes)            40.88%    89.17%
3.0 PRO LB (2 strikes)        32.45%    86.03%

The advantages are:
- we would need less units for an initial attack.
- much higher survival rate means that they are much more hammer efficient, and that they can promo, heal for 2-3t and just keep rolling.
- they are 2pop whippable with or without Kremlin.

LC suggested that we would use choko or sword built in Europe for conquest there. That may be fine, with Kremlin and hopefully decent size captured cities. But it also has potential to be slower and more awkward. We can move HAs all the way to Paris, via Moscow and Berlin, in under 20t.

HAs are killer, especially with the bomber. My thinking is to send our spare chokos west asap, fight it out when they get there. Chariots catch up as needed. One of the problems is having a road. It also depends on how much we want them to settle cities. I'm thinking pretty much stone-age them after they've teched the Medieval techs for us. Not sure how realistic all this is.

My biggest problem with HBR is that teching it might delay the Astro path, depending on when we would get the 2 GSes. (Yes, I know that's still hypothetical, but that's why we're at 0% slider.)

Bear in mind that Astro is more than just access to RoosyLand. It's also galleons, which rock against the barb galleys and in terms of moving our units around.

 

[LowtherCastle]

I suppose I should come up with a Kremlin plan and you guys focus on the wars. How much military are you counting on coming from Beijing?

I've slotted 2 chariots and a choko, the choko for Manchou. After that, no plans or needs necessarily. The choko could be whipped into the Kremlin. I would be delighted if you planned the Kremlin out and I adjust things to that plan. I'm whipping the first chariot to get the two chariots to Mongolia in time.

 

[LowtherCastle]

bbp, note I edited my post to you above...

 

[Kaitzilla]

Gift our excess resources to Roosevelt now?

Accounting for our war plans, we are gonna be at war with everyone at least once.(-3 diplo, more for lost open borders bonus and lost +1 long peace)

If there really is an American continent, Roosevelt will use that Medern era bonus to settle it and get big.
So he should be our main trading partner.


Standard rule of thumb says we should lose 2 turns for tao/Org as soon as we get 5 workers out.
That loses the least amount of food/city development.

War plan devolopment sounds good.
I can't imagine AI stealing Circumnavigation Bonus before us, so it can wait for after bomber plan.