Alternative Map during 1.18

[KdNseer]

Then what new measures are you proposing?

Set a standard year for every tech line.Higher cost for these techs which are beyond the current era.
For example,before ad500,the players will cost 2 times to research feudalism,4 time to research campass and 8 times for finance.

 

[Cypriot]

I made a lot of progress regarding the settler AI in recent days. Like I said before, my two main problems with the previous state were:

1. The AI only poorly followed the settler maps, often preferring tiles with settler value 1 over easily available 10+ sites
2. The AI often got "stuck" not founding any cities even though there was easy room for expansion, e.g. Carthage not settling North Africa and Rome not settling Iberia

I addressed both problems without "forcing" these outcomes and by mostly fixing bugs and adjusting the logic in the existing game logic. There include:

Can it fix strong reluctance of Polynesian and Harapan AI to expand?

 

[Leoreth]

What is the issue with the Harappan AI?

With Polynesia, I am not sure. Maybe it is worth trying to enable them again with the recent changes. Although at first glance it does not seem to me that the issues I addressed apply for Polynesia.

If I had to guess for them it is also the insistence on sending defenders. Maybe it is worth not sending a defender if the destination is a one tile island? Or a land mass with no enemy units on it?

However that would not be trivial to implement because currently the logic for filling cargo ships happens without taking the city sites into account.

 

[Logoncal]

My question to this is: Would you want a City Sprawl of Harrapa in a game, where they just collapse when India spawns?

 

[Leoreth]

Progress has been slow lately, both due to the recent Victoria 3 update and because I was stumped by an unexplainable bug. Fortunately I have now discovered the cause of the bug, and it's pretty likely that it is present in 1.17 and previous versions as well. The bug can lead to city sites with very high AI found priority to flip to a negative found priority, which causes the AI to never consider the city site for founding. This could particularly occur with sites designated as high priority in the settler maps. It might explain why the AI stubbornly refused to settle some locations no matter how highly they were prioritised in their settler map (and as it turned out, even more so if it has highly prioritised). Cities like Pyongyang come to mind.

I have done some other things to help the AI along with settling (in addition to what I talked about before) but the AI settler behaviour looks in a very good place right now - not just in the sense that it's doing a lot of settling but also that it's much more responsive to configuration via settler maps than in previous versions.

I have some other minor things unrelated to settler AI that I want to address. After that the plan is to return to 1.17 for a bit and fix its open issues.

 

[Matara]

How is it reflected in current 1.17 version? I'm curious because nothing seems too weird for now.

 

[Leoreth]

I don't know, I did not investigate the situation in 1.17, it's just a guess. As you said it works with the bug present, so I will not worry about it.

 

[SultanRedSnake]

Progress has been slow lately, both due to the recent Victoria 3 update and because I was stumped by an unexplainable bug. Fortunately I have now discovered the cause of the bug, and it's pretty likely that it is present in 1.17 and previous versions as well. The bug can lead to city sites with very high AI found priority to flip to a negative found priority, which causes the AI to never consider the city site for founding. This could particularly occur with sites designated as high priority in the settler maps. It might explain why the AI stubbornly refused to settle some locations no matter how highly they were prioritised in their settler map (and as it turned out, even more so if it has highly prioritised). Cities like Pyongyang come to mind.

I have done some other things to help the AI along with settling (in addition to what I talked about before) but the AI settler behaviour looks in a very good place right now - not just in the sense that it's doing a lot of settling but also that it's much more responsive to configuration via settler maps than in previous versions.

I have some other minor things unrelated to settler AI that I want to address. After that the plan is to return to 1.17 for a bit and fix its open issues.

Ahhhh this is a pretty big catch. Niiiice

 

[Askii2D]

Dear Leoreth,


Are new civ's stability maps ready?

 

[Leoreth]

I shared them earlier in this thread.

 

[Dracosolon]

Alternative Map during 1.18

This is the follow up thread for this one, signifying that we are now out of development for v1.17 and to clear the log and relaunch the discussion. I will keep using this thread to post updates on what is going on, especially while the development happens mostly behind the scenes.

forums.civfanatics.com

Here's the link.

Just noticed Marseille is outside France's core. Was its core too big?

 

[Gen1_]

And why is there an Indonesian stability map ?
Won't it disappear in 1.18 ?

 

[Leoreth]

The stability maps are from before the new civilizations were added.

 

[Leoreth]

Some bugs you only discover by playing: my invasion of Egypt stalled out because the restriction that desert is only passable when culturally controlled was a bit too strict - it actually only permitted the owner of the tile to enter, not friendly civs with open borders or enemies.

I think that is another reason why Egypt has been doing so unexplainably well in my test runs. Literally no one else can attack them from Asia.

 

[Dracosolon]

Set working overtime I see.

 

[Matara]

Let's add winter general effect to deserts, rainforest and marsh.

 

[Leoreth]

By this metric, the ancient era gained 10 turns, the classical era gained 49 turns, the medieval era gained 5 turns, the Renaissance era gained 7 turns, the industrial era gained 17 turns, and the global era gained 7 turns. So the early game has been significantly extended.

The classical era gaining 49 turns means that you have ACTUAL action between the Argead empire, Persians and the Romans.
How would that impact UHVs, such as Rome?

Actually an update on this: after playing through it as the Celts, the classical era does drag a bit currently, so I decided to revert the change here that introduces a period of 5-year turns before returning to 10-year turns. This shortens the classical period by 20 turns on normal speed (still considerably longer than on the old calendar). I have not "found a home" for those 20 turns yet so currently the game length is 580 turns, but I probably will get them in somewhere.

Another topic I want to bring up is health from resources. As a reminder, here are the relevant resource and building changes in 1.18 that impact health:

Changed buildings:

• Granary: removed health from Corn, Wheat, Rice, added +1 health from Salt
• Harbour: added +1 health from Fish, Clam, Crab, Whale
• Pharmacy: removed +1 health from Spices, Banana, added +1 health from Incense, Opium
• Wharf: removed +1 health from Fish, Clam, Crab

Removed buildings:

• Smokehouse

New buildings:

• Tannery (Tanning): +1 health from Deer, +1 happiness from Fur
• Grocer (Guilds): +1 health from Spices, Banana, Citrus, Olives, Dates
• Abattoir (Refrigeration): +1 health from Cow, Pig
• Grain Silo (Labour Unions): +1 health from Corn, Millet, Wheat

New resources:

• Millet: +1 food +1 health (1 city), +2 food with Farm
• Potato: +1 food +1 health (2 cities), +2 food with Farm
• Citrus: +1 food +1 health (2 cities), +3 food with Orchard
• Dates: +1 food +1 health (2 cities), +3 food with Orchard
• Olives: +1 food +1 health (2 cities), +3 food with Orchard
• Salt: +1 commerce +1 health (3 cities), +1 food -1 production +2 commerce with Quarry

Specifically I am thinking about grains and Granaries. I think 1.17 has a problem where a lot of health buildings were frontloaded and it was too easy to gain health early on, without having to actually invest in e.g. Baths.

I addressed this by removing Smokehouses and removing the health effect from Granaries, shifting it all the way to the industrial period with the new Grain Silo building. To compensate this a bit, and to remove clustering of health benefits in the medieval era, I moved the health benefits of seafood resources from Wharfs to Harbours.

However, I am wondering if I went overboard with this change. A landlocked city now can only get health from Baths before the Middle Ages, if you ignore the +1 with Granary salt. Also, Granaries now feel a little weak. You do not always have salt, and there are often better things to get than +25% faster growth. Areas like Egypt and Babylonia generate enough food either way.

The value of grain resources is also quite low - they only supply one city, and for most of the game there is not much else to get out of them. I mean, they still provide the food to their tile, and that was my original thinking behind limiting their utility. In turn, grains also have a very low trade value (the AI is aware of how many cities will benefit from a resource).

After thinking about this a bit, I think my preferred solution is to have grains provide no base health at all, but instead add +1 health from wheat, corn, rice to Granaries, so that they provide +2 health with both Granary and Grain Silo. I would do the same for potatoes, with the difference that they do not get health from grain silos but instead happiness with distilleries.

This makes Granaries give +1 health from wheat, rice, millet, potato, and salt. Seems like a lot, but for the early game most civilizations only have access to 1-2 staples plus maybe salt.

 

[Cacaso]

Actually an update on this: after playing through it as the Celts, the classical era does drag a bit currently, so I decided to revert the change here that introduces a period of 5-year turns before returning to 10-year turns. This shortens the classical period by 20 turns on normal speed (still considerably longer than on the old calendar). I have not "found a home" for those 20 turns yet so currently the game length is 580 turns, but I probably will get them in somewhere.

With the large size of the map, I think that the "age of navigation" for the colonization of the Americas and then Asia+Africa should take a little longer to be executed, so it may need more turns, or higher ship speeds.

And I also always had the impression that in the industrial era, the game "accelerates" and so the CIVs (especially Britain) start to have a very large production capacity and then end up developing too quickly, perhaps trying to give more time in this part can be interesting for gameplay. so that this phase can be appreciated a little more.

So I think it should be a good place to put these 20 turns. to let the end of the renaissance and the beginning of the industrial era be a little longer

But we have to see, as this balancing of UHVs occurs, it will be possible to discover which parts of the game are "surplus" or "missing" turns.

 

[Leoreth]

All ships (almost) have already gained an additional move, for example galleys have four moves now. The travel distances across the Ocean don't feel meaningfully larger now.

I also already increased the length of all other eras by approximately the same amount. You are probably right that those turns should go into the late third of the game. It's not so much a question of finding the right place for them but rather to make the math work out so that 600 turns = 5020 years.

 

[Wessel V1]

2020 was still pretty far away when DoC started but since we're in 2024 now, an option might be to simply move the end year to 2030 or so?