An idea on how to hard-code tile selection for border growth

[doublex55]

Border growth happens the entire game. It really shouldn’t be concerned with ancient eta techs.

 

[doublex55]

Going for floodplains over iron hill is laughable. At least ill get my farm triangle

 

[Gazebo]

You know what's more evil than burned popcorn? Hard-coding.

 

[LifeOfBrian]

Hard-coding is just one idea. Since you're the expert here, I'd of course support alternatives to hard-coding that you prefer and that would still address the problems of the current system. That is if in your opinion you agree that there are problems with the current system?

 

[DukeBrani]

What if the border growth algo would take the city tile manager into account when deciding where to expand the border? For example, when a city is founded the player selects Production focus (and locks tiles of choice, just in case) to indicate preference for production-heavy tiles in the future. Does that make sense?

 

[Stalker0]

You know what's more evil than burned popcorn? Hard-coding.

And even more evil than that? The border growth algorithm

 

[doublex55]

Why is it growing to stone 4 tiles away before it claims all the workable tiles -_-

 

[azum4roll]

Why is it growing to stone 4 tiles away before it claims all the workable tiles -_-

For the +10% wonder production, obviously!

I've posted the following on Github, just copying it here:

If performance allows, each potential tile (neutral tiles) in radius 5 should be considered and given a score based on the following criteria:

• Tile yields. Culture, science and faith should be worth more than other yields. Border growth/golden age points should be worth less. 0 points on this for tiles outside radius 3.
• Potential improvement yields. We can keep it simple: farm on flat tiles, mines on hills, the respective improvement on resources. Only consider base improvement yields + pantheon bonus. 0 points on this for tiles outside radius 3.
• Civ UAs. Add some score on mountains for Inca. Not sure if we should consider trees for Iroquois and rivers for Songhai since that opens up lots of other possibilities to complicate matters.
• City Connections. Prioritize growing those tiles first, slightly.
• Natural Wonders, Strategic Resources and Luxury Resources. Add some more points for them even if they're out of radius 3.
• Distance from city. Prioritize tiles closer to city first, slightly.

After that, each tile on the path from the city to tile X should get bonus points based on the score of tile X. Doesn't have to be much, just a slight nudge to grow towards great tiles not adjacent to the borders.

Lastly, zero out non-adjacent tile scores and sort all the tiles based on score.

Tell me if I've missed some points or you find my choices questionable.

 

[LifeOfBrian]

Another example of the woeful state of current mechanics for tile expansion.

 

[Bromar1]

View attachment 580831

Going for floodplains over iron hill is laughable. At least ill get my farm triangle

I think that being across the river from the actual city tile makes it harder for borders to grow naturally in that direction