Deity AI's behavior is inconsistent


Jan 26, 2023
After 1000 ish hours of playing civ 5, I have tried some heavy customized stuffs with XML/LUA and IGE as well, the idea would be to create an epic deity game scenario where you are some kind of Atlantis city civilization in the middle of a donut/pangaea kind map, surrounded by Zulus/Huns/Mongols/Rome/Japan etc... still working on the settings and map.

I noticed this the 1st time in 1 vs 1 on a 4p map with no city states and marathon in a standard game, was playing Shoshones, and got very lucky with ancient ruins ( got over 15 due to the bigger/empty map ).
This allowed me to get, a lot of free techs, free writing, great library and most of the ancient/classical era wonders+fast expands with liberty, and also a tech lead very early vs Spain.
I decided to forward settle Isabelle because there was a nice coastal spot near Madrid, and I had 3 composite bowmen from upgraded scouts while she was still with warriors and maybe a regular archer, so I was confident in my defenses.

From this point, she litterally stopped growing, and never tried anything nasty, except spying and sending prophets/missionaries.
She really had a poor army, especially for deity standards ( despite the fact that It was a 4p map, so she had lands with luxuries and coal on the west of the map, but never used them ).

This ended with domination at early industrial with no surprise, but triggered my curiosity about the fact that deity AI underperforms when you do well.
When i got gatlings and frigates, I decided to go behind her capital to explore the rest of the map, she was actually struggling with barbs on this side, which never happens in deity games when you are behind as you should be, the map is fully covered by AI cities at industrial era usually.

Then I tried Spain myself, until I got a very good map with GBR/Victoria/Solomon's mines, and this time only my neighbor, Arabia, underperformed ( he had riflemen when all of us were building bombers or spaceship ), but Germany gave me what I wanted, a "fair war" at medieval era, when you consider Germany's eventual carpets of doom, and my insane start.

This week I wanted to try something fun as England on pangaea quick map, with the deity opponents mentioned above Shaka/Oda/Rome/Assyria/Attila etc...
When I give myself some crazy start with IGE, not only my neighbor will underperform, but all the AIs in the game, even the ones I haven't met.

The most intriguing was Shaka, he was my neighbor, the guy was litterally on standby, no units built at turn 45 quick ( he had only 142 military score and only 9 or 10 pop ), no expand, very slow to no growth, he was last in almost everything, and other AIs were doing barely better, except Oda who was doing his deity stuff as usual, and rapidly got more military than my 5 full promoted composite bowmen, and was annexing city states on his half of the pangaea, which is the norm for deity AI.
I played an other game vs deity Shaka as neighbor, and he had 5 cities + 2 settlers built, and an army of swordsmen + catapults at turn 48.

So I am wondering why is the AI becoming suddenly shy when you have a big lead, and why do they underperform so much once they are behind, especially your neighbor?
They should get science/growth/expand/production at the same pace,if not more, because they benefit from my tech lead ( give myself writing and masonry ), but once you have a lead, they become even "slower" for most of them ( I even revealed the map to see what they are doing, and they are doing really poorly compared to normal games, It looks like the AI needs to be leading in order to cheat, otherwise they play like prince AI for most of them, It's like Deity Ai is ragequitting when I get GL before them and then prince AI takes control lol ).

I get that having army prevents from DOW, but they should still grow/expand/build military etc... which is not the case when you have dreamlike scenario with Spain/Shoshones or even IGE.
Last edited:
Top Bottom