Civilization VII Update 1.2.0 - April 22, 2025

[stealth_nsk]

wrt 3, that may also refer to resources required for civ unlocks

Probably. I just hope treasure resources are included as major ones.

 

[Kenshiro70]

Good lord. Understanding the output and the impact of changes requires both micro- and macro-analysis plus a healthy dose of statistics. For the type of work I do, that's when I'd call in an enterprise architect. I'm nowhere near smart enough to see the totality much less forecast the impact of a single major change like randomizing or doubling the area. And doing BOTH at the same time? Fuggedaboudit.

What on earth were they thinking?

 

[Gorbles]

Good lord. Understanding the output and the impact of changes requires both micro- and macro-analysis plus a healthy dose of statistics. For the type of work I do, that's when I'd call in an enterprise architect. I'm nowhere near smart enough to see the totality much less forecast the impact of a single major change like randomizing or doubling the area. And doing BOTH at the same time? Fuggedaboudit.

What on earth were they thinking?

You have to understand, enterprise-level SaaS often works on far larger scales than any games development team (even an established, mature team like Firaxis). You barely even have things like SLAs in video games; servers often have scheduled, multi-hour maintenance at regular intervals (which completely breaks something like the six nines standard most commercial SaaS tends to ride on).

It's a very different relationship - you have individual consumers / the market buying a standalone product (en masse) vs. something that's (typically) a recurring subscription, with contractual definitions to safeguard both the vendor and the client. This has a knock-on effect on the structure of the non-development parts of the business.

Think of it as SaaS with a small, dedicated team. Heck, think of it as more "startup" kind of mentality. It's not, but as someone who's been through the startup-to-enterprise SaaS pipeline, basically you need two things (especially for a release like this, where turning public opinion round counts for more than the visualised end state of the product):

• Responsiveness.
• Experimentation.

The best way to experiment is (assuming something isn't literally on fire) is to get it out to as many players as they can. The amount of pseudo-RNG and similar work in the algorithms used for VII is significant (and historically, the franchise - I delved into the infamous assignStartingPlots.lua back when Beyond Earth was new), and this means developers are limited by the physical hardware they have in-office. I have a similar problem - our software is for web and mobile, and the range of mobile phone hardware is immense. There's only so much testing we can do, and only so much load testing we can scale to (in non-prod environments).

 

[protocol7]

Depends on what their goal is, if they think the game is fun and needs to be polished up they should slow down and move methodically to build trust and show they are making their already great game more stable. If they aren’t sure if people actually like the foundation, they should rapidly iterate on those fundamental systems to try to find that match as soon as possible. I’ve heard arguments here that both possibilities could be true, probably it’s just different groups of people who are happy or not with the game as-is, and I’m not sure which Firaxis is focused on. But it would be very hard to try to please both at once (move fast and don’t break things?)

 

[stealth_nsk]

Depends on what their goal is, if they think the game is fun and needs to be polished up they should slow down and move methodically to build trust and show they are making their already great game more stable. If they aren’t sure if people actually like the foundation, they should rapidly iterate on those fundamental systems to try to find that match as soon as possible. I’ve heard arguments here that both possibilities could be true, probably it’s just different groups of people who are happy or not with the game as-is, and I’m not sure which Firaxis is focused on. But it would be very hard to try to please both at once (move fast and don’t break things?)

I think there's a big intermediate group, who are not very happy with the foundation, but could enjoy it once the game is polished. Taking them in mind, Firaxis strategy seems legit - by iteratively improving the game they could gather large enough audience, while any radical changes could alienate everyone.

 

[Kenshiro70]

You have to understand, enterprise-level SaaS often works on far larger scales than any games development team (even an established, mature team like Firaxis). You barely even have things like SLAs in video games; servers often have scheduled, multi-hour maintenance at regular intervals (which completely breaks something like the six nines standard most commercial SaaS tends to ride on).

You want a job? Cause you're talkin' my language.

What's interesting is that there's starting to be more overlap. It's less about scale and more about complexity of systems and their attendant interactions. Civ VII is a collection of interdependent systems, and it's hit the tipping point where the Butterfly Effect come into play. You could make a compelling argument that Civ VII is a collection of good individual ideas that do not work well together - for example, constraints on individual pieces like Distant Lands being inaccessible during the Ancient Age end up having wide-reaching impact on the traditional idea of a 4X game being a sandbox.

It's a "boiling the frog" situation - game companies have slowly drifted into this level of complexity and don't realize when they've passed the tipping point. The results are problems like this resource one, as well as some of the other regressions seen in the patches.This is why I love watching game development - it provides some very visible lessons that I can take and apply back in my enterprise work.

But back to your point, Firaxis doesn't have the development resources to approach this like enterprise SAAS would. However, what they can do is create a Steam beta channel and let all of us in these forums bang on new code for a couple of weeks to smoke out any issues.

I wasn't kidding about the job by the way - I'm interviewing for a CPO position and would be looking to assemble a team to go in and rapidly do some good. PM me if you're interested.

 

[Gorbles]

You want a job? Cause you're talkin' my language.

What's interesting is that there's starting to be more overlap. It's less about scale and more about complexity of systems and their attendant interactions. Civ VII is a collection of interdependent systems, and it's hit the tipping point where the Butterfly Effect come into play. You could make a compelling argument that Civ VII is a collection of good individual ideas that do not work well together - for example, constraints on individual pieces like Distant Lands being inaccessible during the Ancient Age end up having wide-reaching impact on the traditional idea of a 4X game being a sandbox.

It's a "boiling the frog" situation - game companies have slowly drifted into this level of complexity and don't realize when they've passed the tipping point. The results are problems like this resource one, as well as some of the other regressions seen in the patches.This is why I love watching game development - it provides some very visible lessons that I can take and apply back in my enterprise work.

But back to your point, Firaxis doesn't have the development resources to approach this like enterprise SAAS would. However, what they can do is create a Steam beta channel and let all of us in these forums bang on new code for a couple of weeks to smoke out any issues.

I wasn't kidding about the job by the way - I'm interviewing for a CPO position and would be looking to assemble a team to go in and rapidly do some good. PM me if you're interested.

Hah, I'm flattered but just a developer, not really much good at the higher-level product management stuff. I offer my input from my corner, when needed, as the walking library of institutional knowledge (having been around so long).

r.e. Firaxis, it's definitely interesting seeing (admittedly without any real insight into day-to-day operations) how different developers can vary so wildly. Firaxis seem to be at the more organised end of the scale (we've seen rapid changes in patch scheduling, hotfixes, even things like public comms). Other developers I follow haven't changed in that regard in close to 20 years. Others are only just taking advantage of community platforms like Discord and the like.

Would be cool to see some public Steam betas, if they have longer term plans in the works. I hope they do - despite how rocky public perception has been to VII so far.

 

[Morgawr]

Old World game has had a beta channel for at least 5 years now, they update it weekly and then update the main game once a month. Works really well.

(If you’re hiring any data / financial / business analysts with decades of management and strategy experience I am on the job market as well! )

 

[uppi]

Good lord. Understanding the output and the impact of changes requires both micro- and macro-analysis plus a healthy dose of statistics. For the type of work I do, that's when I'd call in an enterprise architect. I'm nowhere near smart enough to see the totality much less forecast the impact of a single major change like randomizing or doubling the area. And doing BOTH at the same time? Fuggedaboudit.

What on earth were they thinking?

Having worked with semi-statistical algorithms myself (basically, you start with a statistical approach and then learn that it does not work, because some assumptions you'd have to make for the mathematically correct solution do not hold. Then you start trying to "fix" the statistics and it invariably becomes a mess), my experience is that you don't try to forecast, but instead you tune it, let it run, collect data and then analyse the data whether the result matches what you want.

The longer I think about it, the more I think that in this case the problem is unsolvable without design changes. You want

• The resource density to remain roughly the same
• A variability in the resource types, so you don't have 4-sugar-and-nothing-else starts
• Some resources to always appear
• Symmetry between the Homelands and the Distant Lands so that all play by the same rules
• Have enough treasure resources to fulfill the legacy path in Exploration

These contradict each other, so an algorithm that fulfills all of these might not be possible. In my opinion, design changes are needed. Either increase the value of treasure resources, reduce the amount needed or have other sources of treasure fleets (maybe from trade?)

 

[stealth_nsk]

Having worked with semi-statistical algorithms myself (basically, you start with a statistical approach and then learn that it does not work, because some assumptions you'd have to make for the mathematically correct solution do not hold. Then you start trying to "fix" the statistics and it invariably becomes a mess), my experience is that you don't try to forecast, but instead you tune it, let it run, collect data and then analyse the data whether the result matches what you want.

The longer I think about it, the more I think that in this case the problem is unsolvable without design changes. You want

• The resource density to remain roughly the same
• A variability in the resource types, so you don't have 4-sugar-and-nothing-else starts
• Some resources to always appear
• Symmetry between the Homelands and the Distant Lands so that all play by the same rules
• Have enough treasure resources to fulfill the legacy path in Exploration

These contradict each other, so an algorithm that fulfills all of these might not be possible. In my opinion, design changes are needed. Either increase the value of treasure resources, reduce the amount needed or have other sources of treasure fleets (maybe from trade?)

I don't see them contradicting:

Resource density is handled separately from other requirements as first part of the algorithm generates resource spots before distributing resources. Poisson geometry they use seem to do the trick (although I don't have enough math myself to follow it fully)

For the rest I could make an algorithm which would distribute resources exactly according to those rules (plus hemisphere specialization), it's not that difficult. First, sort resources between those to appear on each hemisphere, when distribute them among resource locations avoiding clusters with, for example, the same Poisson geometry.

However, Firaxis went with just random distribution within small set of rules and this shows one more priority you haven't mentioned - enough randomness to make different games distinctive and fun.

 

[Morgawr]

For the record, I do like randomness and prefer that.

Nearing the end of my EXP game now, with the whole map revealed the treasure resource generation looks fine in this particular game. I will do a thorough count by mainland / distant lands before transitioning and post.

P.s. been posting about this extensively on another thread - the impact of resource changes has been MASSIVE, making deity it much more challenging which, surpassed my expectations

 

[Kenshiro70]

The longer I think about it, the more I think that in this case the problem is unsolvable without design changes.

You've hit one of the real underlying problems with Civ VII. . Some of the systems, as currently implemented, can only handle so much change. This is probably one of the reasons why there aren't large and huge maps available out of the box. Those will need to be rewritten, which is going to take time.

Which brings us to the second underlying issue - console limitations. They made a decision to create one version that works on all platforms without understanding the full impact on testing and performance, and it's starting to really limit what they can do. The console UI is only a surface-level problem; performance and memory is the real killer. The requirements you just outlined might not be tenable on the Nintendo Switch's ten-year old processor. Notice how 1.2.0 didn't release for the Switch - even this change is too much and needs optimization work.

The first clue I had of this was that prior to 1.2.0, most of the changes were simply parameter changes, which is another clue that they are trying to avoid code changes as much as possible. I'd bet that parameter changes are much easier to roll out on consoles - those are "inert" and probably get through the certification processes much more easily.

In my opinion, design changes are needed. Either increase the value of treasure resources, reduce the amount needed or have other sources of treasure fleets (maybe from trade?)

That's the third and most vexing problem - they don't seem very flexible on their design principles. That's going to have to change in order for things to get better.

 

[uppi]

I don't see them contradicting:

Resource density is handled separately from other requirements as first part of the algorithm generates resource spots before distributing resources. Poisson geometry they use seem to do the trick (although I don't have enough math myself to follow it fully)

For the rest I could make an algorithm which would distribute resources exactly according to those rules (plus hemisphere specialization), it's not that difficult. First, sort resources between those to appear on each hemisphere, when distribute them among resource locations avoiding clusters with, for example, the same Poisson geometry.

However, Firaxis went with just random distribution within small set of rules and this shows one more priority you haven't mentioned - enough randomness to make different games distinctive and fun.

I very much doubt that it is a Poisson geometry. Instead it is going to be some variation of a Poisson distribution with some additional rules. I don't think the function is in the TypsScript, so we can only speculate.

But anyway, the point is that before 1.2, some latitudes were almost entirely treasure resources, so there was a high chance for settlements with multiple treasure resources in range. This was to the detriment of the AIs in the distant lands, but since the player did not start there, this did not matter too much. Now that the generation is supposed to be symmetric, you cannot do that anymore, because you need to mix in non-treasure resources for civs which start there (which could be the player(s)). So, if you keep the same overall resource density, you will have a lower density of treasure resources. And no algorithm could ever change that, even if you handpicked the resources yourself according to those criteria.

 

[evonannoredars]

If I were to try and draft a workflow for better resource generation (ie more treasures), with no changes to current resource spawn requirements and assuming this had to function independent of the code determining previous era generation:

Spoiler :

- Multi-era resources that did not generate in a previous era get blacklisted, except for treasure resources. Not really important for this but should be how it works anyway imo
- Depending on how many did not generate last era, allow one or two more to be generated
- For all remaining treasure resources, identify all tiles currently with resources that they could go on (also considering hemisphere identities)
- Establish number of treasure resource substitutions needed to bring proportion of TRs up to a certain number, divide number up between TRs with some slight +/- randomisation (so not equal numbers of all TRs)
- Determine competitiveness of each substitutable tile (how many TRs could go on each tile) then evaluate sum competitiveness faced by each TR
- Begin with the most competitive tile, swap it for the TR facing the most sum competition
- Continue in order of competition, with a chance to skip over claimed tiles (reduce amount of treasure generation in already claimed land)
- Don't swap out a resource if the quantities of it on the map get too low, remove all of those endangered resources from the substitution pool
- If all substitutable tiles have been depleted, use existing function to scatter remaining required TRs across the map
- Also generate some other resource types in unclaimed areas for variety

I'd like to restate my coding experience is v limited so i won't be offended if anyone points out a major flaw in this

 

[Morgawr]

As promised I did a complete count of all treasure resources on my current Exploration map (Continents, all standard, only UI mods. Special thanks to Moxl map search mod)

There are more than enough of them - 73 in total, 32 in distant lands, 41 in mainland which are 'empire' resources for mainlanders.

I understand some of your games generated way fewer than this, thus the complaints?

 

[MutilationWave]

As promised I did a complete count of all treasure resources on my current Exploration map (Continents, all standard, only UI mods. Special thanks to Moxl map search mod)

There are more than enough of them - 73 in total, 32 in distant lands, 41 in mainland which are 'empire' resources for mainlanders.

I understand some of your games generated way fewer than this, thus the complaints?

View attachment 730426

Thanks for the data. I only had access to one ocean, and I found eight there. One in the islands and seven on the coast of distant lands. I'm starting to explore the other ocean but yes it's not looking good. I'm currently razing a city state suzerained by my enemy so I can resettle one tile south and get two instead of one. I'm in progress taking the town to the north of it which can reach 3, it was settled in the perfect spot by my enemy right before a distant lands civ could settle it.

After this I'll have 6 of the 8 treasure resources, I may finish the path but we're already at 46%. I'll leave Napoleon one city so I don't advance the age, hopefully nobody else wipes him out.

 

[Lonecat Nekophrodite]

Is it just me seeing Mexicans getting regular Linear Infantry? what happened to Soldaderas? no longer replacing regular infantry or now limited number units that requires specific conditions to train? or nerfed? (they're more of Partisans and less of regulars IRL) but this isn't really their uniforms. too British to my likings. and actually US Army uniform of the war of 1812!

 

[Noble Zarkon]

Notice how 1.2.0 didn't release for the Switch - even this change is too much and needs optimization work.

It's possible that was certification related with the Swtich 2 coming out, that process seems to take a lot longer.

It's less about scale and more about complexity of systems and their attendant interactions. Civ VII is a collection of interdependent systems, and it's hit the tipping point where the Butterfly Effect come into play.

But back to your point, Firaxis doesn't have the development resources to approach this like enterprise SAAS would.

Definitely seems like they have a lack of people doing the big picture Enterprise level Software Architecture / Technical Leadership especially with the cross platform complexity they have introduced.

 

[Pfeffersack]

As promised I did a complete count of all treasure resources on my current Exploration map (Continents, all standard, only UI mods. Special thanks to Moxl map search mod)

There are more than enough of them - 73 in total, 32 in distant lands, 41 in mainland which are 'empire' resources for mainlanders.

I understand some of your games generated way fewer than this, thus the complaints?

View attachment 730426

To me that might be an indicator that the perceived scarcity is indeed related to getting the total treasure resource amount now distributed to two hemispheres instead of "new world only" before the patch. The increased number of different resources probably reduced instances of each individual resource. While that doesn't bring the total number further down, it might contribute to the feeling of less clustering...and the player has to watch out for more resource types "at the right locations"

 

[MutilationWave]

To me that might be an indicator that the perceived scarcity is indeed related to getting the total treasure resource amount now distributed to two hemispheres instead of "new world only" before the patch. The increased number of different resources probably reduced instances of each individual resource. While that doesn't bring the total number further down, it might contribute to the feeling of less clustering...and the player has to watch out for more resource types "at the right locations"

I understand what you're saying, and I agree with your logic, but they have become seriously scarce. If you don't have an empire coast to coast at the end of antiquity, you're not going far on the legacy path unless you get lucky or play very fast and well. Likely both.

I was able to get 6/8 I have access to. We're already at 50% so I'll try to remember to say in this thread if it was enough to get the path done. I hope so, because I hate losing my cities.