[BTS] SGOTM 26 - Home Slices

[LowtherCastle]

Peter probably is not teching Monotheism since with any overflow at all, he'd have it at the start of our T104 I think. ​

I think AIs adjust last turn to finish a tech when close. Observable in their graphs, the GNP graph, with little spikes the turn they finish a tech.

 

[WastinTime]

I think AIs adjust last turn to finish a tech when close. Observable in their graphs, the GNP graph, with little spikes the turn they finish a tech.

That's a fun little fact. So they don't do much overflow from 1 tech to another. I like that.
More predictable.

Kait, did your 8 test Peter's have 80-100 bpt? Cus I'm sure it matters how many turns the tech takes. That's probably why Education blocks the AI pretty well in the Lib race.
Your test wasn't picking MC or Feudal and that would be cus they're not doing enough bpt I'm guessing.

 

[Kaitzilla]

Kait, what makes you think Peter is running 80% slider? Base on maintenance costs?

I don't remember ever seeing an AI run along at 100% slider before.

Let me check the 8 Peter's test.

80
50
40
70
60
50
50
70

Hmm, almost ever one is running 20% espionage slider
Did one cause all the rest to respond?

Yes, I suppose 100% is possible.
That top one is 80% and 20 with 0%

 

[WastinTime]

The biggest weight factor seems to be 'choose cheap tech'
Seems like they almost always pick a tech they can research in 1 turn (on normal speed) 2 turns epic and 5 turns Marathon.
Mono is still 2t for him.

 

[Kaitzilla]

That's a fun little fact. So they don't do much overflow from 1 tech to another. I like that.
More predictable.

Kait, did your 8 test Peter's have 80-100 bpt? Cus I'm sure it matters how many turns the tech takes. That's probably why Education blocks the AI pretty well in the Lib race.
Your test wasn't picking MC or Feudal and that would be cus they're not doing enough bpt I'm guessing.

Ah, no they did not have 80-100 bpt

We should probably just hold on to Paper to be safe unless Sitting Bull or Peter demand it.
Sitting Bull is a loser who won't trade it to anyone else, and techs too slow to capitalize on Liberalism.

 

[Kaitzilla]

I don't remember ever seeing an AI run along at 100% slider before.

Let me check the 8 Peter's test.

80
50
40
70
60
50
50
70

Hmm, almost ever one is running 20% espionage slider
Did one cause all the rest to respond?

Yes, I suppose 100% is possible.
That top one is 80% and 20 with 0%

Well, I feel silly.
All the AI run 20% slider because I infiltrated them with a Great Spy!

They never run 100% slider?
I'm an idiot!

Of course they must run 100% slider

 

[Kaitzilla]

Gah! I missed that.

Assuming 80% slider
T101 - +60
T102 - Anarchy
T103 - +80
T104 - Anarchy
+140 in something right now.
Or +153 in Currency since he knows Math+Alphabet
Maybe add on another 4% since we know Monotheism/Currency/Paper and Peter has met us.

Emperor AI techs cost about 80% of what we pay I think:
Currency - 496
Monotheism - 149
Paper - 744
Nationalism - 2232
Feudalism - 868
Aesthetics - 372
Calendar - 434
Horseback Riding - 310
Metal Casting - 558
Compass - 496
Construction - 434

Peter probably is not teching Monotheism since with any overflow at all, he'd have it at the start of our T104 I think. ​

Continuing the foolishness, emperor/immortal/deity AI pay Noble costs for techs.

Only the human pays inflated tech costs on higher difficulty.
For emperor, our techs all cost +20%

Here is the actual cost for techs for Peter.
Currency - 520
Monotheism - 156
Paper - 780
Nationalism - 2340
Feudalism - 910
Aesthetics - 390
Calendar - 455
Horseback Riding - 325
Metal Casting - 585
Compass - 520
Construction - 455

Emperor AI Peter has Civil Service and has advanced to the Medeival Age (iAIPerEraModifier =2), so he gets an era modifier of -6% cost to everything I think, including tech costs?
So this list of tech costs can be reduced 6% to get the true number for Peter I think.
That would make Monotheism = 146​

 

[Kaitzilla]

Techs
T105 - Education 100% (548 -82) +178 Education@ 715/2808
T106 - Education 100%, (466 -87) +193 Education@ 2576/2808 Great Scientist bulbs for 1647
T107 - Metal Casting 0% (379 +51)
T108 - Metal Casting 0% (430 +47)
T109 - 100% slider on Peter (477 -63) +126

This is actually kind of interesting.
According to the test game, we get different amounts of espionage at 100% depending on the turn and our state of empire.

T107 - +142
T108 - +127
T109 - +126

To get a look at what Peter is teching, we need a one turn burst of about 135 I think.
So 100% slider on T108 or T109 by itself would not do it.

We must also consider if Peter will retaliate with us using his slider.
It would slow down his tech rate and cause us to lose vision on what Peter is teching.
Peter's slider is more effective than normal because his Castle gives +25% and his capital is running bureau for +50% commerce.

We really don't want to trigger Peter's slider.

 

[WastinTime]

We really don't want to trigger Peter's slider.

Excellent. So can we put the last nail in this coffin?
No slider for now.

 

[Kaitzilla]

Testing seems to indicate Peter will not crank up his espionage slider as long as we don't go much above his threshold to see what he is teching.

Excellent. So can we put the last nail in this coffin?
No slider for now.

Ok.

It is really only useful for when we have actual Friendly AI​

 

[Kaitzilla]

Continuing the foolishness, emperor/immortal/deity AI pay Noble costs for techs.

Only the human pays inflated tech costs on higher difficulty.
For emperor, our techs all cost +20%

Here is the actual cost for techs for Peter.
Currency - 520
Monotheism - 156
Paper - 780
Nationalism - 2340
Feudalism - 910
Aesthetics - 390
Calendar - 455
Horseback Riding - 325
Metal Casting - 585
Compass - 520
Construction - 455

Emperor AI Peter has Civil Service and has advanced to the Medeival Age (iAIPerEraModifier =2), so he gets an era modifier of -6% cost to everything I think, including tech costs?
So this list of tech costs can be reduced 6% to get the true number for Peter I think.
That would make Monotheism = 146​

Sigh, again this is wrong!

Testing indicates iAIperEraModifier does not affect tech speed at all!
As the AI advances through the eras, it takes less food to grow cities and less hammers to build units and buildings, but the amount of beakers to finish a tech remains the same.

Here is the actual cost for techs for Peter.

Currency - 520
Monotheism - 156
Paper - 780
Nationalism - 2340
Feudalism - 910
Aesthetics - 390
Calendar - 455
Horseback Riding - 325
Metal Casting - 585
Compass - 520
Construction - 455

Final Answer Regis!​

 

[Kaitzilla]

Oh, surprise, surprise. How simple.

tl;dr
If he only has 3 Tao cities or 5 total cities, we can bribe him with 2 Confucian spreads.
Otherwise it takes 3+ confu cities. I'm pretty sure we would need 3 confu or more.
I'm surprised it has nothing to do with population.
Anyone want to do a test game to confirm?

Formula:
N = nPeterCities / 2 (round down). Let's say it'd be half of 6. N = 3

N2 = nPetersTaoCities - 1. Let's assume he will have 4 Tao cities, so N2 = 3 also

AI will refuse due to DENIAL_MINORITY_RELIGION if:
#ConfuCities < min(N, N2)

        if (getHasReligionCount(eReligion) < std::min((getHasReligionCount(getStateReligion()) - 1), (getNumCities() / 2)))
        {
            return DENIAL_MINORITY_RELIGION;
        }

Yes, testing reveals it works exactly like this.

As long as Peter has either 5 cities or only 3 Tao cities, we can bribe him into converting with only 2 spreads.

Attitude Thresholds
Will open borders: Cautious

Will trade techs: Annoyed
Will trade extra happy resources: Annoyed
Will trade extra health resources: Cautious
Will trade extra strategic resources: Pleased
Will trade map: Cautious
Can plot war at pleased: Yes
Can be bribed to war against AI: Cautious (If he is pleased or lower with target AI)
Will sign trade embargoes: Pleased (If he is cautious or below with target civ)
Will adopt religion: Pleased
Will adopt civic: Pleased
Will possibly peace vassal to human player: Friendly
Will form defense pact: Friendly
Will form permanent alliance: Pleased

Thankfully Peter is not one of the nuts that needs to be Friendly before they can be bribed to switch religions.​

 

[Kaitzilla]

Kait, what makes you think Peter is running 80% slider? Base on maintenance costs?

I wonder if anyone has ever tried to tease out precisely how fast an AI is teching using the GNP number?

Careful monitoring of all cities and buildings after Open Borders and before caste/music should let someone know exactly how much an AI produces each turn.
is going to be just +4 per turn.
Pull that out of GNP and the GNP number is just Science(with multipliers) + Commerce which represents the 2 main sliders we care about.

Then calculate how much maintenance the AI is probably paying.
Then calculate how much the AI has each turn from the espionage screen and city numbers.
Brute force out what the commerce slider must be set at.

I think with an absurd amount of effort, per turn might be able to be pulled from GNP data?
..
……..
Ok, probably not.

 

[LowtherCastle]

To add to the ep confusion, once Peter meets SB, what will he do with his ep allocation and slider?

All the more reason to not mess with it pre-SB.

 

[Mitchum]

Peter's slider is more effective than normal because his Castle gives +25%

This reminded my why SB was getting 5 per turn against us early on instead of the normal 4.

I still want to see Peter's research but let's cross that bridge a bit later. Him spending more than 1 turn teching a duplicate tech to us could potentially add 1 turn for 1 turn to our finish date...

 

[LowtherCastle]

This reminded my why SB was getting 5 per turn against us early on instead of the normal 4.

I still want to see Peter's research but let's cross that bridge a bit later. Him spending more than 1 turn teching a duplicate tech to us could potentially add 1 turn for 1 turn to our finish date...

right and getting techs from Peter is our best opportunity to overcome any disadvantage from settling later than PR.

 

[Kaitzilla]

I am comfortable now with all the decisions and will post the PPP for the next 3 turns.

The only way Peter is teching Monotheism (156) now is if he had no overflow from Alphabet and is 80% slider or lower.
Much more likely, he had overflow from Alphabet.
Or is running 90% or higher on tech slider considering how fast he did indeed get Alphabet.

In the tiny chance he gets Monotheism next turn, it is not the end of the world.
The earliest we can get Friendly Peter is T116.
We can just grab a monopoly tech from T117 Friendly Sitting Bull and make more trades with Peter.
Or we can wait for another +1 diplo on T122 (shared civic) reaches +3 for Peter and do the Philosophy then if we have not traded 5 techs to Peter yet.
We get an additional +1 diplo on T128 from Open Borders, and have until around T130 to get that Philosophy.

 

[Kaitzilla]

PPP
T105 to T107​

Continuing from Mirrorguard's turnset here and here

Screenshots

Spoiler Empire :

Spoiler AI and World :

Spoiler Extra :

​

Using micro from earlier post.

Techs
T105 - Education 100% (548 -82) +178 Education@ 715/2808
T106 - Education 100%, (466 -87) +193 Education@ 2576/2808 Great Scientist bulbs for 1647
T107 - Metal Casting 0% (379 +51)

Workers
Cowboy
T105 - Chop forest 2E of Whale City to complete missionary
T106 - Move to forest 2S2W of Silver City
T107 - T110 Chop forest

Heretic
T106 - Finish Mine 2S1E of Canal City
T107 - T109 Move to forest 2N2W of Canal City

Phillip
T105 - Road 2S1E of deer tile to complete New Deer Road
T106 - Chop forest 2S1E of deer tile
T107 - Chop forest and STOP!

Joan Dark
T105 - Move to forest plain hill 3S2W of Science City
T106 - T108 Chop forest into University

Shepherd
T106 - T107 Road 1W of Science City

Rex
T105 - T108 Chop forest 4S1W of Science City into University

Chariot
T106 - T108 Explore west and try to meet the 4AI as fast as possible

Cities
Science City
T105 - Temple; Sheep, Rice, Horse, Wine, x3 Plainsriverfarm, Plainsforest
T106 - Temple; Sheep, Rice, Horse, Wine, x3 Plainsriverfarm, Plainsfarm, Plainsforest
T107 - University; Sheep, Rice, Horse, Wine, x3 Plainsriverfarm, Plainsfarm, Plainsforest

Cow City
T105 - Sankore (goes to 18/550); Crab, Cow, Deer, x2 Plainsriverfarm, Tundrafur
T106 - Monastery; Crab, Cow, Deer, x2 Plainsriverfarm, Tundrafur, Plainshillfur
T107 - Monastery; Crab, Cow, Deer, x2 Plainsriverfarm, x2 Coast

Canal City
T105 - Wealth; x2 Clam, Crab, x2 Coast, x2 Scientist
T106 - Wealth; x2 Clam, Crab, x3 Coast, x2 Scientist
T107 - Maoi (goes to 160/250); x2 Clam, Crab, x5 Coast

Clam City
T105 - Moai (goes to 92/250); x2 Clam, x3 Coast
T106 - Library; x2 Clam, x3 Coast
T107 - Library; x2 Clam, x3 Coast, Mine

Profit City
T105 - Monastery; Crab, Whale, Stone, Coast
T106 - Monastery (goes to 22/60); Crab, Whale, Stone, x2 Coast
T107 - 2-pop Whip Monastery; Crab, Whale, Stone

Silver City
T105 - T106 Confu Missionary; Deer, x2 Crab, Clam, Silvermine, x4 Coast
T107 - 2-pop Whip Missionary; Deer, x2 Crab, Clam, x4 Coast (Boards the Galley T110 right as it starts to sail west again)

Whale City
T105 - Missionary; x2 Crab, Clam, Horse, Silvermine (1 Chop lands)
T106 - Moai (goes to 159/250); x2 Crab, Clam, Horse, Silvermine, Fur
T107 - Courthouse; x2 Crab, Clam, Horse, Silvermine, Coast

Units
Galley in Peter's land

Spoiler :

T106 - Move 2E to clam
T107 - Move 2N2E (Regretable we don't have time to unfog the corner, but we can at least see if it is coastal or not)

Galley 2W of Whale City

Spoiler :

T105 - Wait
T106 - Collect 2 Confu Missionaries and sail 2W
T107 - Sail to 1S of copper and unload 2 missionaries 1S1W of copper

Galley south of Sitting Bull

Spoiler :

T105 - T115 Sail to Whale City

Confucian Missionaries 1 and 2
T106 - From Whale City, board Galley 2W of Whale City
T107 - Exit Galley 1S1W of copper

Warrior (2S2E of Deer tile)
T105 - T107 Move to garrison Clam City

Diplomacy
on Peter
T105 - Trade Meditation to Sitting Bull for 20 + switch to Organized Religion (The map will show if SB had any extra map knowledge, tiny chance of a few squares) (We know from buying Peter's map that Peter does not)
T105 - Once Sitting Bull is in Anarchy, sell Math to him for + his map so we can get +4 diplo
Keep our map until Peter and SB meet
----------------------------
Give in to any Peter or Sitting Bull demand except:
1) Stone is a NO
2) Gold demand will be discussed with team
Everything else is a Yes including Paper or our map.

Every turn
Check if SB and Peter are close to meeting (This involves 2 SB scouts near Silver City, Peter WB near Profit, and SB workboat north of Canal City)
Check if Confu spread somewhere
Zoom in on Sitting Bull's + Peter Cities and screenshots
Record the espionage values on all AI cities
Save and the start and end of each turn

Stopping point
SB and Peter accidentally meet early or are about to meet next turn
Peter or Sitting Bull demands , pause and discuss with team the amount
Meet a new AI
A world wonder is completed
Something unexpected happens
Reach start of T108

**Edit**
100% Done, please comment ​

 

[WastinTime]

Shepherd
T106 - T107 Road 1W of Science City

I'd rather not have this road. Are you using it to get to the forest for Oxford?
Not a big deal if we have the road I suppose (thinking about forest growth)

T107 - Maoi (goes to 160/250); x2 Clam, Crab, x5 Coast

work the new mine? for +6-7 in Moai this turn?

 

[Kaitzilla]

I'd rather not have this road. Are you using it to get to the forest for Oxford?
Not a big deal if we have the road I suppose (thinking about forest growth)

work the new mine? for +6-7 in Moai this turn?

Yes, the road 1W of our capital is for getting to the Oxford forest chop.

T107 mine in Canal City would later lead to Maoi being over 250 and being unable to fit a good chunk in.
I had to carefully lower the amount of overflows going into Maoi on this city to make them all fit.

Canal City
T105 - Wealth; x2 Clam, Crab, x2 Coast, x2 Scientist
T106 - Wealth; x2 Clam, Crab, x3 Coast, x2 Scientist
T107 - Maoi (goes to 160/250); x2 Clam, Crab, x5 Coast
T108 - Courthouse; x2 Clam, Crab, x5 Coast, Mine
T109 - 4-pop Whip Courthouse; x2 Clam, Crab, Coast, Mine
T110 - University (goes to 45/200); x2 Clam, Crab, x2 Coast, Mine
T111 - Monastery; x2 Clam, Crab, x4 Coast
T112 - 2-pop Whip Monastery; x2 Clam, Crab, x2 Coast
T113 - Moai (goes to 245/250); x2 Clam, Crab, x3 Coast (1 Chop lands)
T114 - University; x2 Clam, Crab, x3 Coast, Mine
T115 - University; x2 Clam, Crab, x4 Coast
T116 - 4-pop Whip University; x2 Clam, Crab, Coast

Maybe if we are ditching the Courthouse, but I really like the Courthouse
Getting University down to a 4-pop whip is heavenly compared to a 5-pop whip.

The earliest we could have Forges is the start of T113.​