Current v1.13 Development Discussion

[Leoreth]

So to keep you updated, I'm currently working on religion spread. The goal is to make automatic spread more accurate while avoiding determinism.

Right now I'm in the process of taking apart and understanding the code. One interesting observation is that it was completely impossible for natural spread to bring more than one religion to a city. I wasn't aware of that before, so that'll probably be removed.

I will also introduce region-specific spread modifiers that are intended to partially replace civ specific modifiers. That should allow realistic spread even if the "correct" civ in the area is collapsed or has been conquered.

The code is also really inefficient as far as I can tell, maybe I can achieve some speed improvements here too.

 

[NerfCothons]

On marathon in almost every game I play starting at 3000 BC these days Buddhism spreads to several European cities and Zoroastrianism spreads to twice as many. I understand not wanting to create a deterministic setting but maybe there should be an early game limitation to natural spread based on something like distance from the holy city. Or maybe the spread rate is just too high on marathon. Either way once you remove the one religion cap from natural spread we're going to see weird stuff like this all over.

If you are trying to remove location determinism from parts of the game, such as colonization, perhaps pulling from religions found within the founder's other cities would accomplish the same goal better. Presumably when a fully Catholic Spain founds St. Augustine Spaniards are moving in. So, there's not much reason for it to have the protestant bias that it does in the code for when the same tile is settled by England as Savannah. Although nearby religions should also move in, Dutch settled Colombo should not only pick up the dutch Protestantism but also local Hinduism, Buddhism, and Islam.

Obviously I am am not offering any solutions within the framework of what is possible since I really don't know a thing about the mechanics of this part of the game, but I'm glad you're taking a look inside.

 

[Shadow Warlord]

Maybe there should also be a way for religions to spread for civs using Secularism (outside of the emigration mechanic)? That way missionaries from surviving monasteries wouldn't be the only method for late game religion spread.

 

[Imperator Knoedel]

I'm pretty sure what NerfCothons is experiencing is specific to Marathon speed.

 

[Tigranes]

The code is also really inefficient as far as I can tell, maybe I can achieve some speed improvements here too.

It is amazing how after so many years there are parts of code that never been touched by community (otherwise this would be reported and could become a public knowledge by now). Just how many people does it take to write a brand new game? I read somewhere that Civ 3 AI was written by 1 person.

Going back to the development of the things that already been implemented I need once again stress the very unbalanced situation for Persia. 8% goal was set with previous stability and previous UP, with those new ones you simply cannot avoid your cities go Indy. Unless you create many large cities in your core real fast, which is not an option, given the time it takes to get to Calendar and Currency, build plantations and UBs. Also it seems that religion hurts Persia with her goals being Pantheon Wonders.

 

[Leoreth]

On marathon in almost every game I play starting at 3000 BC these days Buddhism spreads to several European cities and Zoroastrianism spreads to twice as many. I understand not wanting to create a deterministic setting but maybe there should be an early game limitation to natural spread based on something like distance from the holy city. Or maybe the spread rate is just too high on marathon. Either way once you remove the one religion cap from natural spread we're going to see weird stuff like this all over.

Now that you mention it, there is a chance for religion to spread every turn and I don't think there's any mechanic that counteracts the fact that more turns on Marathos means more chances.

However, one of the region specific things I am going to do is that there are some regions where certain religions cannot spread naturally at all, unless it's the owner's state religion. That should prevent random Confucianism in Europe.

If you are trying to remove location determinism from parts of the game, such as colonization, perhaps pulling from religions found within the founder's other cities would accomplish the same goal better. Presumably when a fully Catholic Spain founds St. Augustine Spaniards are moving in. So, there's not much reason for it to have the protestant bias that it does in the code for when the same tile is settled by England as Savannah. Although nearby religions should also move in, Dutch settled Colombo should not only pick up the dutch Protestantism but also local Hinduism, Buddhism, and Islam.

Yeah, both of these things should be possible and need to be balanced in some way against each other. In principle both elements correspond to civ and region specific spread factors.

It is amazing how after so many years there are parts of code that never been touched by community (otherwise this would be reported and could become a public knowledge by now). Just how many people does it take to write a brand new game? I read somewhere that Civ 3 AI was written by 1 person.

To quote Soren Johnson: "Most of the AI in Civ 4 really is High School level programming".

To be fair, with the restriction that only one religion can spread to a city the algorithm is fine, I think. They can simply skip cities that already have a religion. And usually programming works on a "good enough for our purposes" principle.

But the algorithm does iterate every city on the map for every city per turn, a classic case of quadratic complexity. I think I can easily drive this down with some caching, while also tweaking the algorithm do behave more in a way that is desirable.

 

[ales_]

I will also introduce region-specific spread modifiers that are intended to partially replace civ specific modifiers.

You'll have to divide the map on provinces, won't you?

To quote Soren Johnson: "Most of the AI in Civ 4 really is High School level programming".

This means that it's rather simple, doesn't it?

 

[Leoreth]

You'll have to divide the map on provinces, won't you?

It already is. (Don't tell anyone)

This means that it's rather simple, doesn't it?

Yeah, although we was only semi-serious.

 

[NerfCothons]

Maybe there should also be a way for religions to spread for civs using Secularism (outside of the emigration mechanic)? That way missionaries from surviving monasteries wouldn't be the only method for late game religion spread.

I think that once monasteries become obsolete the method of spreading religion should be civic driven. We'll see what kind of changes come from this batch to decide what's appropriate.

 

[Shadow Warlord]

It already is. (Don't tell anyone)

must...have...province...overlay...

 

[ales_]

What for?

 

[citis]

I think he means the cores. They are almost everywhere, thus practically they work like RFCA provinces.

 

[Leoreth]

Regions are not cores, and you never interact with them directly. So an overlay would be confusing rather than helpful.

 

[mrrandomplayer]

I'm just going to point out that one of the letters in OCC is for challenge. I'll never balance the game around it.

Fair point, that was a poor example for the tech problem.

Now spies are too important. The only way to play big is mass armies and spies early, do not research by yourself until you got a big land basis.

This is a much better example, when one of the best DoC players around says that spying is more viable for getting techs than research. I've found that after the end of the classical era (where the tech costs were balanced to make sure that the ancient civs wouldn't get out of control), European advancement is crippled because the tech costs are too high. Whenever a new civ spawns, eg Poland with Civil Service, they are in the lead with tech because the rest of the civs are too slow in research because the techs cost too much. Rhye reduced tech costs at 600 AD (and I think other times as well) for a reason.

 

[Malchar]

If you play BTS and set the commerce slider to 100% espionage you can do quite well by stealing techs rather than researching them yourself. In many ways, espionage and research are interchangeable resources. If you are putting penalties on the research rate, then you should also put penalties on the tech stealing rate to maintain the symmetry.

As it stands, the current system gives a strong penalty to the tech leader who is researching new techs, and it's easier for the civs who fall behind to catch up... if they're smart enough to use espionage.

 

[Leoreth]

Espionage is completely unrelated to the tech costs discussion. The cost of stealing techs scale directly with their research cost.

 

[CaterpillarKing]

I have question, what's the purpose of Harappa? As in big-picture. What do they add to the world as a whole? They are very interesting to play and fun but I fail to see the reason they exist for the first bit of the early game and then totally collapse and crumble to the point where even my capital was destroyed by instability.

 

[merijn_v1]

I have question, what's the purpose of Harappa? As in big-picture. What do they add to the world as a whole? They are very interesting to play and fun but I fail to see the reason they exist for the first bit of the early game and then totally collapse and crumble to the point where even my capital was destroyed by instability.

It's human player only. You gave the reason for that yourself.

 

[CaterpillarKing]

I didn't know it was human player only. Carry on

 

[mrrandomplayer]

I started a 600 AD game as Germany, and when I spawned, I had an absurd tech lead. Only HRE had Rifling, and everyone was lagging behind like crazy.

Spoiler :

I just rolled over Poland and probably could beat most of the other Euros, backing my theory that techs cost too much.

The base cost of techs for me is regular though, so now I'm really confused about what you did to change the tech cost adjustment. It seems not to readjust for Ancient civs and by the tech level in this game medieval civs.