[NFP] AI crazy about Science

[Infixo]

Anyone wondering why AI is mostly focused on science for the last... let's say... half-a-year?
It is because of this line, added in Victories.xml file with September 2020 patch:
<Row ListType="ClassicalYields" Item="YIELD_SCIENCE" Value="150"/>
It increases the internal value of science yield by 150% and this stays for the entire game. It overshadows other yields, like Gold or Faith. AI almost stopped building Commercial Hubs. No more missionary carpets. And poor Great People are wandering around not even in Russia (they always did so in Peter's empire) but basically in every civ - because there are not enough Theater Squares to put their Great Works into.
Change this value to a reasonable 30-50 if you want to have a bit more diverse game.

 

[Infixo]

Also, in April Patch Campus was made a favored disctrict for every civ.
<Row ListType="ClassicalDistricts" Item="DISTRICT_CAMPUS" Favored="true"/>

 

[MrRadar]

<Row ListType="ClassicalYields" Item="YIELD_SCIENCE" Value="150"/>
It increases the internal value of science yield by 150% and this stays for the entire game. It overshadows other yields, like Gold or Faith. AI almost stopped building Commercial Hubs.

Oh, my! And I was wondering why it was so hard to find a CH for my spy training…
The most sloppily coded instalment in the entire franchise

 

[Aristos]

Even using the word "code" within the Civ 6 context is offensive for real coders.

 

[kaspergm]

People complained AI cared too little for science, making the game too easy. "If only the AI would spam campuses like the human player, the game would be much more challenging ...".

Well, we got what we wished for.

 

[Haig]

Hmm not too bad then.

 

[Fluphen Azine]

People complained AI cared too little for science, making the game too easy. "If only the AI would spam campuses like the human player, the game would be much more challenging ...".

Well, we got what we wished for.

I haven't played the game for months now.
Everyone still wins right?
It just takes longer now?
Is that the current state of the game?

 

[kaspergm]

I haven't played the game for months now.
Everyone still wins right?
It just takes longer now?
Is that the current state of the game?

Pretty much yes, except if you play with Corporations mode, in which case you're likely to win super fast cultural victory.

 

[MrRadar]

People complained AI cared too little for science, making the game too easy. "If only the AI would spam campuses like the human player, the game would be much more challenging ...".

Well, we got what we wished for.

A headless chicken put on steroids is still headless.
There was a thread about AI disbanding their military, probably due to bankruptcy. So it might be happening because of this. AI spam campuses, ignore CHs and lose troops. And the game is even less challenging, as AI have little idea what to do with all that science.

 

[The Highwayman]

I haven't played the game for months now.
Everyone still wins right?
It just takes longer now?
Is that the current state of the game?

Wins don't currently take longer in my experience. The Ai is a little better at science, but the player still slingshots past around T100. The compete hard for the scientists, especially early on, but pretty much give up the rest with a few exceptions. It's been easier to found a religion, as it has been for a couple years now, and I'm thankful carpets of missionaries and apostles are almost non existent. Ouch...just got a flashback of vanilla when religious units occupied the same space as civilian and military.

 

[Aristos]

A headless chicken put on steroids is still headless.

Don't be THAT mean. Firaxis' coders are real people, you know, They have feelings.

 

[kaspergm]

A headless chicken put on steroids is still headless.
There was a thread about AI disbanding their military, probably due to bankruptcy. So it might be happening because of this. AI spam campuses, ignore CHs and lose troops. And the game is even less challenging, as AI have little idea what to do with all that science.

It's fun and to a certain extent also justified to make fun of the AI in Civ6 - the military issue you mention is really disheartening, if the theory behind that is true - but to an extent, many of the issues with the AI stems from the fact that it's written to play the game the way the developers intended for it to be played, and then later - and mostly unsuccessfully - tweaked to try to play the game the way the humans play it. Which is a problem, because often it seems that tweaking the AI has been used as an attempt to solve problems that are rooted in underlying balance issues, rather than fix those balance issues themselves.

The science issue mentioned in this thread is a great example of that. Basically AI was programmed to do what developers intended the player to do, namely build a balanced mix of districts. However, as players soon discovered, it was favorable to spam campus districts in every city. Imo. the appropriate fix to this issue would be to look into campus and science generation in general - in particularly, to turn balance away from flat yields and adjacency free-yields and rather towards pr. population and/or pr. specialist yields (as discussed at length here). But as showcased above, instead what we got was an AI overwrite to make AI spam campus in every city. Of course I acknowledge that a large-scale district rebalancing scheme would be a major undertaking that would tap into the core design aspects of Civ6. But my point is, the problems the AI have don't necessarily arise just from the AI code, but reflect back on underlying (poor?) design decisions as well.

 

[Pfeffersack]

@Infixo opening this kind of a thread is a good sign

The recent sience obsession of the AI is indeed a problem (or at best a very double-edged sword). It prevents them from falling too quickly in that department and might open them the door for a science victory on the higher difficulty levels. But the overall price payed for this is high - all are AIs behaving the same isn't good for fun gameplay and the game pace (which was always too fast anyway) is further increased. It also is IMO a reason the AI can't cope with Dramatic Ages even on high levels - they expand with their extra settlers and speed up era progress by their science focus only to lack time and infrastructure to earn enough era score to prevent cities from defecting. And if they drop in a dark age, their weak military (low unit building focus, perhaps the disbanding issue on top) prevents them to take their cities back from free cities (which never have trouble to pump out quantity anyway, but now also get quality because of the overall high science development level...)

 

[aieeegrunt]

Wow that is lazy

 

[Aristos]

Wow that is lazy

Or incompetent.

Or both.

 

[Mesix]

I have taken to turning off both cultural and diplomatic VC since they come too quickly.

 

[PSG]

For me the fastest way to complete the tech tree is by not building many campuses. Faith + one promotion of the elder gods governor. The AI should do more of the same...and culture too my latest game I was running over 4K culture (about 2500 science with 4 or 5 Universities) around the year 1100AD on Marathon and I didn't build a single culture center though I did inherit one from a city flip.

 

[Pfeffersack]

For me the fastest way to complete the tech tree is by not building many campuses. Faith + one promotion of the elder gods governor. The AI should do more of the same...and culture too my latest game I was running over 4K culture (about 2500 science with 4 or 5 Universities) around the year 1100AD on Marathon and I didn't build a single culture center though I did inherit one from a city flip.

If Secret Socities are active, an AI has a faith/culture leaning and picks the Voidsingers, yes. But diplomatic, scientific, industrial, warmongering or trade focussed AIs should choose different paths.

 

[MrRadar]

To think about this again, this could also be the reason behind Barbarian Men-at-Arms in the Ancient era.

By the way, @Infixo, I've glanced into the Expansion2_Victories.xml and saw that both laser projects are not listed among AI favoured projects, unlike other SV projects. Could this be reason that AI tends not to build those projects after it launches Exoplanet Expedition?

Spoiler :

Would adding those two lines here be enough to change anything?

 

[aieeegrunt]

This hilarity brought to you by the Department Of Zero Playtesting