[BNW] The "Getting Tributes from Deity AI" Strategy PART 1

Thank you for the kind words. I remember that you said one time you only play England. I like them very, very much too... one of the most versatile civs, I think. Amazing UUs, amazing UA (the spy especially)...

About the monument...I like it in all new cities as first building, but in the capital I don't usually prioritize it.

1) When Tradition: I don't build it, I wait for the Legalism policy. I build scouts and rush the shrine instead if I have good dirt for a pantheon.

2) When Liberty: I don't prioritize it, because:

- either my warrior or scouts reveal some good place to settle (a natural wonder, perhaps), and I will produce a settler quickly without waiting for Liberty policy that gives one. The monument gets delayed until after shrine, granary, water mill (I like food buildings a lot). It's still good for border expansion, after all.

- there is no good enough place to hurry to settle, so I can wait for Collective Rule. And while the monument helps you get faster to it, I still prefer to build other things like shrine, granary, or a wonder.

3) When Honor or Piety, I build it ASAP, but after 2 scouts and the shrine. (I still take Tradition Opener when I want to go deep into Honor or Piety, 3 culture and fast borders are just too good to skip).
 
Thanks @Tiberiu — this guide is some next level stuff! I will give it a try the next time I notice my GPT is looking good! HT to @vadalaz for pointing it out here here.

One caution: GPr from Liberty finisher is not an option if you do not have a pantheon. I have gotten burned by that a couple of times!
 
Thanks @Tiberiu — this guide is some next level stuff! I will give it a try the next time I notice my GPT is looking good! HT to @vadalaz for pointing it out here here.

One caution: GPr from Liberty finisher is not an option if you do not have a pantheon. I have gotten burned by that a couple of times!

Thank you kindly. Yes, I know of the Liberty finisher pantheon requirement. Thankfully it isn't usually very difficult to get a pantheon with a quickly built shrine. I am more annoyed by the fact that the Liberty Prophet also increases cost for the Scientist/Merchant/Engineers too... but a religion is worth this cost, regardless.

edit: I am following that game, I never played a succession game, I think everybody would hate me after my 10 turns :D I delay NC, and do weird things.
There seems to be a certain appeal to it, though. Trying to combine different styles of play in order to win. But I think the players should have more authonomy over their decisions during their 10 turns of play. If I want to go Honor, and the other players don't want to, I would have to Veto the decision.

(Unrelated to beetle's comment, I'm using this opportunity to make some general comments and write down my thoughts about this strategy, and what a solid strategy could/should be based on, in Civ 5.)

There are some small updates that I never got around to doing because I think a general strategy is too hard to compile, so instead writing about all aspects of the game in a more "modular" way is what I hope to be able to do at some point, but I came to civ a bit late and there's little demand for such resources. I really like Zigzagzigal's civ-specific presentations of civ 5, but I feel like they never get into depth so as to be able to get an Immortal Player to win on Deity. This was my initial intention with this guide, to provide something that can be used to win Deity with consistency.

However I always felt that my presentation has a certain flaw - since everything could be potentially dismissed as based on an exploit (I know vadalaz at least, considers it an exploit, which actually got me thinking - what value can a strategy have if it's based on an exploit?). To some extent, I would agree, however, all Deity wins are based on exploiting the weak AI, so there are strong arguments in abusing the AI. The player only plays with the rules that he is given - and as long as a feature is not a bug that was not intended by the developers (like, for example, the obscure "eternal war" bug ), then it is legitimate to use it in order to win. Even shooting and retreating with a promoted Chariot or a Camel Archer can be considered an abuse/exploit since deity AI will never do this in a game. Stealing workers, continuously liberating same worker to barbs, are things AI will never do, so I would/could argue what is an exploit is ultimately a matter of opinion an preference, instead of an objective fact (as long as there are no obvious codes introduced, or files tampered with, obviously)
Selling a lux for 240 early game by DOW and peace in 6 turns is in a way no better "morally" as the GPT exploit, or stealing gold and breaking DOFs, selling/capturing the same city state multiple times is also one of the most powerful abuses that is borderline legitimate. Declaring war on a CS and making peace same turn, multiple times, is, however, borderline cheaty.
My point is Diety AI is unable to do most of the things that a human player does in order to win. That is maybe exactly the point - they were given a huuuge amount of bonuses in order to try to mask this and create an entertaining environment for the casual players, and in this mission, I think they have done their job somewhat well).

For all these reasons, sometimes I think that the best general strategy would be one that works in a self enforced "permanent war + no CS abuse/exploit" setting. This should be the ultimate test. However given the rules by which a Deity game is governed, I think that the only way would be to simply spam army and conquer everything. And, of course, this would also be in a way, exploiting the inability of Civ 5 AI to use their units, combat fortifications, or great generals. Which defeats the purpose that was initially set, i.e winning without exploits.

So, I sometimes think that, in the end, all ways of winning Deity AI (maybe excluding Duels or Tiny maps, and also "cooked" settings), are based on exploits/cheats, etc. This discussion wouldn't have even existed if the company had provided a good AI that can play the game.

/end thoughts
 
Last edited:
I know vadalaz at least, considers it an exploit
Exploit or not is an endless argument and I would agree that it ultimately comes down to personal preference. I don't mind the use of this game mechanic in non-competitive play at all, and if the rules of a competition allow it I'm fine with that as well.

What I don't like is that this game mechanic isn't documented in Civ 5 itself, so it is essentially hidden from players. People can put thousands of hours into the game and miss it entirely. This makes your guide all the more valuable.

I found out that gpt affected peace deals by accident while playing low difficulty games. I noticed that AIs would only offer 1-3 gpt to me for peace in early wars when I had a Gems start, or something similar. First couple of times I must've thought it was my military strength that influenced this, but then it became clear that it was related to my gpt. Having too strong of a military in early wars on low difficulties is actually detrimental, because the AIs want to offer you a city, except they don't have any expos so they offer you nothing at all.

I guess this is another gripe I have with this game mechanic. I think that having an army so strong that the AI is willing to give up a city should never result in worse peace deals than having an army comparable to (or weaker than) theirs. But as you point out in the guide, this happens even on Deity. Being offered 260 gpt is a lot better than being offered the AI's worst expo, some 3-pop city in the middle of nowhere that you can only sell for maybe 40 gpt.
 
I would certainly welcome a final patch made by Firaxis that, at the very least, prevents the AI from ever making such absurd trades.

That multiplayer game, short as it was, helped me see how very much different a game is when you can't sell your useless horses/iron to somebody. This is how the game should be. The fact that Civ 5 Vanilla had the AI buy open borders for 2 gpt led to feedback from the community that is this is outright wrong - and then, what they did was to release 2 expantions where they still left the AI to buy useless open borders for 2 gpt... This shows they either didn't care or failed to ever understand how serious this apparently small thing is. Surely the fact that a horse/iron sells for 2 instead of just 1 is a simple rounding error that they never bothered to correct, since this allows the bizzare situation where in a turn you can get 1 free gold by buying a horse from somebody and selling it back for 2. Your horse quantity remains unchanged but you made gold out of bogus deals. I used to play Civ:Beyond Earth and found that the same applies there: you can buy favours for 1 science if I remember correctly, and then sell that favour to get 1 science and 5 gold. If you are really desperate to get 100 gold, you only need 20 such deals and you're all set.

What frustrates me the most about all of these issues is that the fixes are so easy to make... it's not as if there is some complicated thing that requires a lot of financial resources, or work from programmers. They just don't care.
 
After so much time playing this game I only found now that the Deity AI only needs 100 faith for his prophet, and almost surely for their missionaries too. This is not documented anywhere as far as I can tell. This explains how Deity AI can sometimes found a religion very early in the game. This also explains why the AI can spam missionaries apparently out of thin air (a missionary with Mandate of Heaven would only cost 80 faith to buy for the AI)
 
especially considering the ridiculous amounts of cheats that the AI uses against you and the fact that it has bonus production when building Wonders, too, despite that most online guides wrongly say that they don't. The bonus is such that Cristo Redentor, for example, would require something like 937 or 967 (forgot exact value) hammers instead of the normal 1250 hammers.

After so much time playing this game I only found now that the Deity AI only needs 100 faith for his prophet, and almost surely for their missionaries too. This is not documented anywhere as far as I can tell. This explains how Deity AI can sometimes found a religion very early in the game. This also explains why the AI can spam missionaries apparently out of thin air (a missionary with Mandate of Heaven would only cost 80 faith to buy for the AI)

The AI "Per Era Modifier" applies to World Wonder Production costs as well as building Production costs. This is not documented in most guides but it can easily be confirmed in the game's code:
Code:
    if(!isHuman() && !IsAITeammateOfHuman() && !isBarbarian())
   {
       if(isWorldWonderClass(pkBuildingInfo->GetBuildingClassInfo()))
       {
           iProductionNeeded *= GC.getGame().getHandicapInfo().getAIWorldConstructPercent();
           iProductionNeeded /= 100;
       }
       else
       {
           iProductionNeeded *= GC.getGame().getHandicapInfo().getAIConstructPercent();
           iProductionNeeded /= 100;
       }

       iProductionNeeded *= std::max(0, ((GC.getGame().getHandicapInfo().getAIPerEraModifier() * GetCurrentEra()) + 100));
       iProductionNeeded /= 100;
   }

In addition, their discounts to unit Production costs also apply to Faith costs, which would explain your second observation.

I would certainly welcome a final patch made by Firaxis that, at the very least, prevents the AI from ever making such absurd trades.

That multiplayer game, short as it was, helped me see how very much different a game is when you can't sell your useless horses/iron to somebody. This is how the game should be. The fact that Civ 5 Vanilla had the AI buy open borders for 2 gpt led to feedback from the community that is this is outright wrong - and then, what they did was to release 2 expantions where they still left the AI to buy useless open borders for 2 gpt... This shows they either didn't care or failed to ever understand how serious this apparently small thing is. Surely the fact that a horse/iron sells for 2 instead of just 1 is a simple rounding error that they never bothered to correct, since this allows the bizzare situation where in a turn you can get 1 free gold by buying a horse from somebody and selling it back for 2. Your horse quantity remains unchanged but you made gold out of bogus deals. I used to play Civ:Beyond Earth and found that the same applies there: you can buy favours for 1 science if I remember correctly, and then sell that favour to get 1 science and 5 gold. If you are really desperate to get 100 gold, you only need 20 such deals and you're all set.

What frustrates me the most about all of these issues is that the fixes are so easy to make... it's not as if there is some complicated thing that requires a lot of financial resources, or work from programmers. They just don't care.

The Vox Populi / Community Patch Project mod fixes almost all of the blatant AI stupidity. Unfortunately because Firaxis hasn't released the multiplayer code, its compatibility in multiplayer is limited.

You're far from the only one who is frustrated with their AI design and lack of patching the game. :wallbash:

Q1: How will I survive the early game wars if I declare wars so quickly?

Because the tactical AI is AWFUL and can't even move and shoot with their ranged units on the same turn.
 
Last edited:
I am probably going off topic here, but I am with you on the relative value of the Shrine over Monument.
Yes, I know of the Liberty finisher pantheon requirement. Thankfully it isn't usually very difficult to get a pantheon with a quickly built shrine.
The two times I got burned by this, (1) I had the Shrine early, and (2) decent faith dirt. And both times some far away civ enhanced before I got a pantheon. So I was really counting on GPr for the Liberty finisher! At that point, there are religions left, but faint chance for the human player to get one. Mostly when I play Liberty I already have a religion by the time Liberty closes out.

The first time this happened, I thought the game was bugged or I wasn’t paying close attention. The second time, I came to this forum to confirm this particular caveat of the Liberty finisher!
 

Thank you for providing the source code! I never knew what Per Era Modifier meant exactly...

I am probably going off topic here, but I am with you on the relative value of the Shrine over Monument.

Sure! I think the opportunity cost for rushing the shrine is quite low...even if you don't get what you want from it, you still didn't lose that much. But if things go even moderately well, you get a lot of reward from it. I usually go Scout-Scout-Shrine (even lose some food to hurry 1 turn), unless ToA seems like the better choice, and in that case, I buy the shrine with my first 250 gold. :)
 
Last edited:
Sure! I think the opportunity cost for rushing the shrine is quite low...even if you don't get what you want from it, you still didn't lose that much. But if things go even moderately well, you get a lot of reward from it. I usually go Scout-Scout-Shrine (even lose some food to hurry 1 turn), unless ToA seems like the better choice, and in that case, I buy the shrine with my first 250 gold. :)
Agreed. I am long time advocate of Scout-Scout-Shrine. I found in the majority of my Deity games, and having a Religion adds a whole dimension of play. And even when I don’t found, I am able to buy faith buildings (or even foreign missionaries) while they are still cheap. Because of the compounding nature of early faith, by the end of the game it is unusual if I haven’t purchased two GE and two GS (assuming Tradition/Rationalism). But on occasion, an AI will enhance before I have a pantheon. That means I have wasted eight early turns in my cap!
 
Agreed. I am long time advocate of Scout-Scout-Shrine. I found in the majority of my Deity games, and having a Religion adds a whole dimension of play. And even when I don’t found, I am able to buy faith buildings (or even foreign missionaries) while they are still cheap. Because of the compounding nature of early faith, by the end of the game it is unusual if I haven’t purchased two GE and two GS (assuming Tradition/Rationalism). But on occasion, an AI will enhance before I have a pantheon. That means I have wasted eight early turns in my cap!

Cool tables and analysis of those SSM SMS MSS scenarios. I read some posts from that thread to see what the dominating views were. I find that more often than not, on a standard pangaea or even on a continents map, scout-scout is pretty much the standard way to play regardless of what follows after. They are so cheap and are capable of doing so much, even on standard map size.. Ethiopia is maybe the exception but otherwise, I don't get why monument would be such a hurry. If you find a contested natural wonder or amazing spot, you'll build a settler without collective rule anyway, case in which monument actually delays it.

About religion, sure, in a 8 player game if the rolled rival civs are, Celts, Maya, Ethiopia, Byzantine or whatever other civs like to focus faith, it will be extremely hard to get a religion (I'll play my next games with this setting to see what happens). But, like you said, even in the worst case scenario, we still get some faith at least, which can be used to get something from it, eventually. What is even more frustrating is when you abandon attempting to obtain a religion, thinking you won't have enough time, and then, for some reason, there's still one available in turn 100 and you think, "damn, I would have had it by now if I tried". Similar to how you can skip trying to build a wonder only to see it's still available 15 turns later.

Monument or shrine boils down to whether somebody considers that a small amount of culture values more than a potential pantheon that can lead to a potential religion, similarly how you well pointed out that a few turns of extra scouting should be worth more than that small amount of culture.
 
The thing where you ignore monument as liberty is interesting to me, as it seems inefficient to go for collective rule so slowly as liberty. Whenever I play liberty depending on the map size my build order is usually scout monument shrine scout or scout scout shrine monument. As tradition, I generally agree that monument early isn't worth it.

Of course, if you play celts you can just ignore everything and go for like 3 scouts and a monument as your early builds, which is certainly strong. One of the reasons I love that civ.
 
The thing where you ignore monument as liberty is interesting to me, as it seems inefficient to go for collective rule so slowly as liberty.
Look at the tables in that thread I linked too. Since the next-policy-cost is fixed (and relatively low) and you are getting some Culture from your capital, Monument relatively late in the queue has only a very modest delay to CR. Eight turn latter at worst, and it’s a zero turn delay if you happen to hit a Culture ruin between turn 22 and 42. Meanwhile, you have 8-16 extra turns of effects from the Scout or Shrine (whatever you built instead of the early Monument).
Of course, if you play celts you can just ignore everything and go for like 3 scouts and a monument as your early builds, which is certainly strong. One of the reasons I love that civ.
Celts or Ethiopia gets me out of my Scout-Scout-Shrine opening routine. Which I why I favor them. If I go Tradition, it is rare for me to build a Monument, regardless of the Civ. I would for Ethiopia, but not Celts.
 
I just find that the consistent play on small map size is monument before scout#2 I won't prioritize it over shrine, pantheons are FAR too valuable for that
 
A friend recommended this game called Deity Challange Lineup with Zulu which can be found here: https://forums.civfanatics.com/threads/the-deity-challenge-lineup-game-1-the-zulus.528421/

I'll put a spoiler since this is a public game that others might be interested in playing. I think it is one of the best examples of how early shrine is better than monument. I would be open to see being proven wrong with a practical example, but I just don't see it happening.

Spoiler spoiler :
I lost 3 turns seeking river or fresh water but I saw the option of settling on hill with 4 wheat in capital. I think Sun God is very strong here and I researched Pottery first. BO: Scout-Scout-Shrine-then a few more scouts. I was unable to steal workers from neighbours but I found Uluru relatively quickly and decided to go for it... since it's very hard to get a religion on this map. 7 scouts in total, one found weapon upgrade, I used them to generate XP to prepare to defend the exposed Uluru city.

How did early shrine help? Well, because I built it so early, I could have Sun God early (2nd pantheon), which in turn allowed me to have good enough population and production to produce the first settler quickly, faster than the fastest Collective Rule. Good early population also allowed Pyramids, and now after the initial expansion phase and with a religion secured, the empire is starting to bloom.
Had I built Monument instead of Shrine, I would have made a pantheon much later, which means the religion would have been delayed or even lost. I did take Collective Rule eventually but it was a mistake because I will stay on 4 cities and it would have been better to invest some extra production needed for the 4th settler instead of wasting a policy.

Morale of the story is that when a good pantheon is available, especially an instant food one, the shrine is really good! It just occured to me that there might be another case where shrine is much better than monument - that is with a lot of jungle with Sacred Path. Monument gives 2 culture, but with only 3 jungle tiles you can get more than that, for the same production cost. And of course it's possible to have way more than only 3 jungle tiles. Even if it's temporary (jungle sucks long term), the initial culture boost is pretty good to have.

turn68.jpg




 
I just had the hardest and craziest recovery ever, where I won a game against 2 runaways, after losing 3 out of 4 of my cities, including capital, around turn 130-135. The game is the same from the post above, the zulu challange.
The capital was lost to France for more than 40 turns.

After a good start, I made a few strategical mistakes (got greedy and wanted Heroic Epic before i mass units, wasted 2 policies for republic+collective rule when it made no sense) and from around t90 I found myself in this situation:
Spoiler :

t93zulu.jpg


which continued to this...

t133.jpg



t134_capital.jpg


I'm thinking of writing a more detailed write-up about how this game unfolded in the future, so for now I'll just show how it ended - a very hard worked t 327 Domination Victory!

Never give up!!

t327_victory.jpg


 
Last edited:
do you trade (extra) luxuries with the AI 1 on 1 (for example wine for whales) or do you prefer the 7 gold per turn to sell them?
 
do you trade (extra) luxuries with the AI 1 on 1 (for example wine for whales) or do you prefer the 7 gold per turn to sell them?

I prefer 7 gold per turn early in the game. but 4 happiness can be good when close to Natural golden age.

Later after ideologies, happiness can be also good to offset Public Opinion.
 
I prefer 7 gold per turn early in the game. but 4 happiness can be good when close to Natural golden age.

Later after ideologies, happiness can be also good to offset Public Opinion.

Thanks. a few more questions:

1) With all that gold you make what do you usually buy? Rush Buildings, Units, or get allied city states?
2) How often do you make the war/wait 6 turns/collect tribute rotation? on cooldown?
3) What turn or era do you start the war/wait 6 turns/collect tribute rotation?
 
Top Bottom