Prepare to cringe - bad maps (deity)

AcaMetis · Jul 25, 2019

Spoiler :

Fippy · Jul 25, 2019

Did you expand beyond the first hill city in your new try, Krikav?
I wonder if Monty is more likely to plot if he sees better targets than Prague (with an Archer sitting in there).
Will consider sending settlers towards some spots (double fish), but not found them until i reach friendly (unless Monty settlers show up).

AcaMetis · Jul 25, 2019

I know that if you share a large enough cultural border (8 or 9 land tiles, was it?) you're considered a Land Target, which makes an AI more likely to plot on you. I'm not sure if it makes an AI more likely to plot, though, which matters in this case since

Spoiler :

krikav · Jul 25, 2019

@Fippy

Spoiler :

Anysense · Jul 28, 2019

AcaMetis said:
I think it's based on the relative distances between the AI's capitol and yours (and possibly the capitols of other AIs), rather than a set distance. If it is a set distance it's more than 9 tiles, I just tested it and I got a liberation...9 tiles from the AIs capitol. I swear this heat...anyway, another quick test, and I could still liberate a city even when it's 10 tiles from the AI's capitol.

It is definitely not just distance, there are other factors.
For example, I can liberate this city

Spoiler :

Four turn later it can't be liberated! What's that about?

Spoiler :

AcaMetis · Jul 28, 2019

I take it you used that Missionary to generate culture in that city for those four turns? If so I have to assume it's because of that, I'm not sure what else could have changed in those four turns.

Kaitzilla · Jul 29, 2019

Anysense said:
It is definitely not just distance, there are other factors.
For example, I can liberate this city

Spoiler :

View attachment 530794
View attachment 530795

Four turn later it can't be liberated! What's that about?

Spoiler :

View attachment 530797

(Liberate) is pretty simple.
If a new city is closer to an AI capital than the player capital, the player can (Liberate) it to that AI for +1.5 diplo points.
Diagonal is 1.5 tiles, and in the case of a tie I think it is considered closest to the player. :hmm:

Anyway, at the start of the game a city that can be liberated to an AI will naturally have a popup after 1 turn to let you know an AI capital is nearby.

Spoiler :

Strategy
2 liberated cities for +3 diplomacy can be a crucial move when an AI has a large tech lead, and the player can afford to give away 200 :hammers:

if there is enough room to cram in 2 cities close to the AI capital.

Some have thought to use 3 city gifts and multiple DOW's to force the AI up a net +1.5 diplo each Cease Fire, but there are many flaws to doing it.
3 cities and 3 military units are an expensive investment.
Each DOW is -1 diplo with every AI pleased with the target AI.
Capturing the same 3 cities over and over causes a lot of war weariness to build up.
The AI freaks out and starts building military and stops teching so fast to deal with the situation.
All 3 cities have to be size 2 once to not automatically raze.
Open Borders and the 60 turns of peace bonus both vanish, and trade routes lose the +150% bonus.
Some AI take a few turns to talk peace even with 30 war success.

It is cheaper to use the AI's own cities to force them to Friendly, but the buildings have a 33% chance to be destroyed each time, the turns to come out of revolt resets each time, and the city loses 1 population each time.

Not something I do too often.
(Liberating) a city to an AI while at war with them is an exploit of course.

Culture
Once a city starts adding city culture, then the (Liberate) mechanics start changing in very weird ways!
@WastinTime is one of the only ones to try and puzzle it out.

WastinTime said:
When I say 'distance' I mean sid's distance, which means if it's off continent, then it's doubled. So with that baked-in to ourDistance...

OurDistance > OurCulture/HisCulture*HisDistance
(you had the culture flipped)

Not sure about ties. Seems like the first player checked wins, so we would NOT be able to liberate to anyone if we have tied scores.
If two AI tie then it's the first one processed. *not tested.

This formula has problems during testing, but it feels close?
0 :culture:

is supposed to act like 1 :culture:

during the calculations, but they give different results.
I'll try to list out some distance values from a world builder game to hopefully get a correct formula.

Spoiler :

But wait, it gets even weirder!

If a city is across some water on an island or another landmass, the math changes again. :eek:

WastinTime said:
I loved playing with Pollina because she would always pop in with an exact answer--seemingly glad to research it. With me, it have to go kicking and screaming into the code. I often come up with a partial answer. It can give me a headache, but this time it didn't take too long.

The answer is: if the capital is not in the "area", which I take to mean Toku is off-continent, then you double the distance.
Therefore, our capital is 'closer' to Shanghai than Tokyo is.

See, that's a partial answer. I doesn't explain the culture thing. Want me to go back in? <= See, this is me complaining

WastinTime said:

This shows the doubling I was talking about:

Code:

int iCapitalDistance = ::plotDistance(getX_INLINE(), getY_INLINE(), pCapital->getX_INLINE(), pCapital->getY_INLINE());
                    if (area() != pCapital->area())
                    {
                        iCapitalDistance *= 2;
                    }

And this shows how culture and distance are both used to pick the winner: (Pretty short and sweet, but it gives me a headache)

Code:

                    int iCultureTimes100 = getCultureTimes100((PlayerTypes)iPlayer);

                    if (bConquest)
                    {
                        if (iPlayer == getOriginalOwner())
                        {
                            iCultureTimes100 *= 3;
                            iCultureTimes100 /= 2;
                        }
                    }

                    if (GET_PLAYER((PlayerTypes)iPlayer).getTeam() == getTeam()
                        || GET_TEAM(GET_PLAYER((PlayerTypes)iPlayer).getTeam()).isVassal(getTeam())
                        || GET_TEAM(getTeam()).isVassal(GET_PLAYER((PlayerTypes)iPlayer).getTeam()))
                    {
                        iCultureTimes100 *= 2;
                        iCultureTimes100 = (iCultureTimes100 + iTotalCultureTimes100) / 2;
                    }

                    int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance);

                    if (iValue > iBestValue)
                    {
                        iBestValue = iValue;
                        eBestPlayer = (PlayerTypes)iPlayer;
                    }

Yes, the distance to the gift city is considered Doubled if it is on another landmass from a capital.

But wait, it gets EVEN weirder!!!
Normally vassals can't be gifted cities, but it turns out you can still (Liberate) a city to a vassal.

In testing, I even (Liberated) a city that had 50,000 :culture:

that was 5 tiles from my vassal. :eek:

But as soon as I un-vassaled them, the city could no longer be (Liberated)

I'll try some more tests to figure things out later.
I give up, my brains are mush.

Anysense · Jul 29, 2019

Kaitzilla said:
In testing, I even (Liberated) a city that had 50,000 that was 5 tiles from my vassal.
But as soon as I un-vassaled them, the city could no longer be (Liberated)

This is weird, I was under impression that vassals don't take cities with less than 50% of their culture. But I rarely gave my cities away for purposes other than avoiding domination, so I wasn't paying much attention. It's religious games that brought me to it.

AcaMetis · Jul 29, 2019

Kaitzilla said:
Diagonal is 1.5 tiles, and in the case of a tie I think it is considered closest to the player.

Just tested it, ties are indeed considered closer to the player. At least, a city equal distance between my capitol and an AI capitol wouldn't Liberate.

Whisker · Jul 30, 2019

@Kaitzilla
Staring at the numbers and code snippet you provided, here's how it works (I think).
For each player we have int iCultureTimes100 = getCultureTimes100((PlayerTypes)iPlayer); which simply returns players culture multiplied by 100
Then whoever founded the city gets his/her culture multiplied by 1.5
and if one player is vassas to another the culture gets further modified by:
iCultureTimes100 *= 2;
iCultureTimes100 = (iCultureTimes100 + iTotalCultureTimes100) / 2;
I have no idea what that iTotalCulture above is, maybe combined culture of both players involved in current calculation?
Finally comparison value is get by:
int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance);
Where first part is bigger between 100 & iCultureTimes100 and second part is distance to capital or 1 (I guess this is for cases when there is no capital because total kill required option is checked).

Looking at numbers provided: at culture 0 the values for each player are simply 100/distance so with player distance 21 and AI 20 values are AI 5, player 4,76 which gets trunctuated to 4, AI wins and city can be liberated. If distances were player 22 and AI 21, both values should get trunctuated to 4 and city could not be liberated (can somebody check?)

If player has 1 culture, value is culture times100 times 1,5 (because he is the original owner), so 150/21=7,14 trunctuated to 7
and AI need to lower his distance to 12 to get 100/12=8,333 so that is trunctuated to 8.

For player distances between 21 to 19 and culture of 1 the result always get trunctuated to 7 so results do not change. For distance 18 result get trunctuated to 8 and AI needs further reduce distance to get result that gets trunctuated to next full integer so 100/11=9,09=>9

Somebody with more time than me and access to worldbuilder should test these.

shakabrade · Jul 30, 2019

Thanks guys for the great inputs on liberation mechanics. I have used it very often (most of the games) and it is one of the most valuable tools in consistent deity non deaths (and then wins). It is also useful on Pangaea maps, you can often put a cushion crappy city between you and a warmonger (gift it to peacemonger). The best way is to consult the peaceweights or just a current love hate situation. When a warmonger has the guy it hates as a land target as well, it is surely going to result in a war. When you are not the target of the first war, you can piggyback other AIs and take easy captures or simply neglect the wars and tech. Arguably, you could do it by bribing, but I hate bribing as it accelerates the tech pace of your future victims plus you can only do something well after the alphabet when it could be late. Since liberating is such a frequent weapon in my arsenal, I value IMP quite a lot in masochistic deity setting. Its diplo benefit is greater than SPI and you save so many early hammers on settlers. It is especially helpful on tough maps where you can't just capture everything, while totally irrelevant on lower difficulties or rich maps.

Kaitzilla · Aug 1, 2019

Whisker said:
@Kaitzilla
Staring at the numbers and code snippet you provided, here's how it works (I think).
For each player we have int iCultureTimes100 = getCultureTimes100((PlayerTypes)iPlayer); which simply returns players culture multiplied by 100
Then whoever founded the city gets his/her culture multiplied by 1.5
and if one player is vassas to another the culture gets further modified by:
iCultureTimes100 *= 2;
iCultureTimes100 = (iCultureTimes100 + iTotalCultureTimes100) / 2;
I have no idea what that iTotalCulture above is, maybe combined culture of both players involved in current calculation?
Finally comparison value is get by:
int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance);
Where first part is bigger between 100 & iCultureTimes100 and second part is distance to capital or 1 (I guess this is for cases when there is no capital because total kill required option is checked).

Looking at numbers provided: at culture 0 the values for each player are simply 100/distance so with player distance 21 and AI 20 values are AI 5, player 4,76 which gets trunctuated to 4, AI wins and city can be liberated. If distances were player 22 and AI 21, both values should get trunctuated to 4 and city could not be liberated (can somebody check?)

If player has 1 culture, value is culture times100 times 1,5 (because he is the original owner), so 150/21=7,14 trunctuated to 7
and AI need to lower his distance to 12 to get 100/12=8,333 so that is trunctuated to 8.

For player distances between 21 to 19 and culture of 1 the result always get trunctuated to 7 so results do not change. For distance 18 result get trunctuated to 8 and AI needs further reduce distance to get result that gets trunctuated to next full integer so 100/11=9,09=>9

Somebody with more time than me and access to worldbuilder should test these.

Whoa, looks like you solved it again Whisker! :love:

====================================
Culture = 0 :culture:

for player, 0 :culture:

for AI.

Player founded the city, so 0 gets multiplied by 100, then multiplied by 1.5.
The number for iCultureTimes100 becomes 0.
Then int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance); happens.
The iValue for player = std::max(100, 0) / std::max(1, iCapitalDistance) = Floor (100 / iCapitalDistance)
The iValue for the AI = std::max(100, 0) / std::max(1, iCapitalDistance) = Floor (100 / iCapitalDistance)
If iValue for AI > iValue for player, the city can be (Liberated)
If iValue for AI < or = iValue for the player, the city can't be (Liberated)

====================================
Culture = 1 :culture:

for player, 0 :culture:

for AI.

Player founded the city, so 1 gets multiplied by 100, then multiplied by 1.5.
The number for iCultureTimes100 becomes 150.
Then int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance); happens.
The iValue for player = std::max(100, 150) / std::max(1, iCapitalDistance) = Floor (150 / iCapitalDistance)
The iValue for the AI = std::max(100, 0) / std::max(1, iCapitalDistance) = Floor (100 / iCapitalDistance)
If iValue for AI > iValue for player, the city can be (Liberated)
If iValue for AI < or = iValue for the player, the city can't be (Liberated)

====================================
Culture = 2 :culture:

for player, 0 :culture:

for AI.

Player founded the city, so 2 gets multiplied by 100, then multiplied by 1.5.
The number for iCultureTimes100 becomes 300.
Then int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance); happens.
The iValue for player = std::max(100, 300) / std::max(1, iCapitalDistance) = Floor (300 / iCapitalDistance)
The iValue for the AI = std::max(100, 0) / std::max(1, iCapitalDistance) = Floor (100 / iCapitalDistance)
If iValue for AI > iValue for player, the city can be (Liberated)
If iValue for AI < or = iValue for the player, the city can't be (Liberated)

====================================
Culture = 3 :culture:

for player, 0 :culture:

for AI.

Player founded the city, so 3 gets multiplied by 100, then multiplied by 1.5.
The number for iCultureTimes100 becomes 450.
Then int iValue = std::max(100, iCultureTimes100) / std::max(1, iCapitalDistance); happens.
The iValue for player = std::max(100, 450) / std::max(1, iCapitalDistance) = Floor (450 / iCapitalDistance)
The iValue for the AI = std::max(100, 0) / std::max(1, iCapitalDistance) = Floor (100 / iCapitalDistance)
If iValue for AI > iValue for player, the city can be (Liberated)
If iValue for AI < or = iValue for the player, the city can't be (Liberated)

If iCapitalDistance value ends in 0.5, round down.
If (Liberate) city is on a different continent/island than the capital, double the iCapitalDistance value.

====================================
You have successfully cracked the math formula to perfectly describe (Liberate), nice work!

I really like that 1.5 factor based on who founded the city, it was my main source of confusion.
I also like the multiplied by 100 culture operation, really fixes cities with 0.1 :culture:

for integer operations.

I'm tempted to abuse (Liberate) now that I see how much city culture affects it.
Found 3 cities near my capital, give them to a Cultural Civ, let him build up culture 10 turns, then seize and liberate them over and over to get the AI to Friendly. :devil:

The math for the vassal (Liberate) still is not quite working right, but I barely collected any data, and the code is solid:

and if one player is vassas to another the culture gets further modified by:
iCultureTimes100 *= 2;
iCultureTimes100 = (iCultureTimes100 + iTotalCultureTimes100) / 2;
I have no idea what that iTotalCulture above is, maybe combined culture of both players involved in current calculation?

shakabrade said:
Thanks guys for the great inputs on liberation mechanics. I have used it very often (most of the games) and it is one of the most valuable tools in consistent deity non deaths (and then wins). It is also useful on Pangaea maps, you can often put a cushion crappy city between you and a warmonger (gift it to peacemonger). The best way is to consult the peaceweights or just a current love hate situation. When a warmonger has the guy it hates as a land target as well, it is surely going to result in a war. When you are not the target of the first war, you can piggyback other AIs and take easy captures or simply neglect the wars and tech. Arguably, you could do it by bribing, but I hate bribing as it accelerates the tech pace of your future victims plus you can only do something well after the alphabet when it could be late. Since liberating is such a frequent weapon in my arsenal, I value IMP quite a lot in masochistic deity setting. Its diplo benefit is greater than SPI and you save so many early hammers on settlers. It is especially helpful on tough maps where you can't just capture everything, while totally irrelevant on lower difficulties or rich maps.

Yup, (Liberate) is a wonderful and expensive way to improve diplomacy.

The later the game gets, the less expensive it gets relatively speaking, but the less room there is to place (Liberate) cities.
Then again, that is what Razing is for!

*Razing a city to make room for (Liberate) cities costs -2 diplo*
**Autorazing a city costs -0 diplo weirdly enough** :crazyeye:

Fippy · Aug 1, 2019

Hiya guys, been on short holidays and will try to continue the very challenging Charly map later today ~~
Meanwhile i created a new map, this time without Buffy or any other mods (we can mix that up a bit). Just Bug.

Mao on Isolated (i thougt about that classic Iso map with him

):

Spoiler :

It's overall not that terrible, okay start as you can see. Some decent areas to work with.
However also some annoying ones, and commerce poor. Could enjoy or curse this one, who knows

AIs were left on random, no Mansa etc.
..and i forgot changing the name again lol, still Toku.

Fippy · Aug 1, 2019

Game over for Charly, greed killed

Spoiler :

I could have waited with settling, but in such a bad shape either way.
Urgently wanted his forest chopped before culture expands, but he stole fish some turns later

Barb city which stole a good spot maybe made me feel so hopeless in the end.
I guess i see no real chance but rng luck with settling towards Monty, but maybe something could be done north.

AcaMetis · Aug 1, 2019

Goodness, what a monster that Charly map turned out to be :crazyeye:

. This is looking less winnable than that cooked Prepare to DIE! Wang Kong map at this point.

Spoiler :

Fippy · Aug 1, 2019

Yup, if i replayed i would go like this:

Spoiler :

shakabrade · Aug 4, 2019

@Fippy

Spoiler :

AcaMetis · Aug 4, 2019

Spoiler :

drewisfat · Aug 7, 2019

I really felt like my third peacemonger attempt was going to be the charm but.....

Spoiler :

drewisfat · Sep 5, 2019

Found this monster in a random Shuffle game.
Temperate, normal sea level, semi-isolated, WvO

Spoiler The start :

Spoiler What makes it cringeworthy? :

Apologies, the save is BAT mod.

Prepare to cringe - bad maps (deity)

Deity

Mycro Junkie

Deity

Theorycrafter

Emperor

Deity

Lord Croissant

Attachments

Emperor

Deity

Warlord

Praise Vivec!

Lord Croissant

Mycro Junkie

Attachments

Mycro Junkie

Attachments

Deity

Mycro Junkie

Praise Vivec!

Deity

King

King

Attachments

Similar threads