Game mechanics

[zitro1987]

I like option A

[Domon]

It's probably paying those lungimirant warmongers that just happened to have a paladin unit to sell when you needed it

(I like a game with the right mood and that does not make too little sense, but these little details, i could live with and abstract a justification. Gameplay should come first if possible)

However, i'm slightly worried that this change will slow down the pace of the game a lot. Maybe you could make razing better again, so that there are incentives to burn and pillage instead of conquering if you want to rush a lighting war?

Btw, if you have to do it, i'd go with option B, maybe even limiting buildings contruction somehow. And the time needed to pacify the city could be related to garrsoned troops strenght and unrest relations with your capital.

[Nelphine]

Razing and not trying to defend is actually the benefit, I think. Its how I play almost all my games. Rush around destroying things. Nothing to defend, nothing for them to take back, doesn't slow me down if I can't complete the war because I'm overwhelmed on another front, they don't get to simply recover everything, they have to rebuild from scratch which normally means they can't really threaten me again.

I don't think any further incentives are required. Conquering is a huge economic boost, but requires a much more favorable economic and military position.

[Seravy]

Razing and not trying to defend is actually the benefit, I think.

Agreed. Destroying an enemy city is a huge thing even on its own, getting to keep it is bonus that doesn't come with any "cost" right now - the cost is supposed to be that you have to defend it otherwise the enemy gets it back, but if the city is able to produce units instantly, it's just a matter of "do I have enough gold to defend a city". Which you do because conquering the city gives you gold.
I'm not saying it always works out that way, I have lost games to being unable to defend the cities I took before, but those were typically cases where I used a single hero stack to take over half a dozen enemy cities far from my own territory in the early game when those cities had no means to produce strong units and/or I had no gold income yet to do so.

[Seravy]

Meanwhile, I realized there is one more approach to this whole thing.

If the AI is at a disadvantage because they can't buy units specifically where it is beneficial to do so...what if it learned doing that?

Like, usual unit buying priorities remain, but if some conditions are met, buying units becomes a higher priority? If we could make good rules for this...let's see :

1. If the city is located on a continent with no or very few other cities belonging to the wizard, but a large amount of enemy cities (enemy meaning hostility is on towards the wizard)
2. If the city has adamant, or a lesser extent, orihalcon and mithril.

If we code this as a subroutine, and call it from production decisions as well as buying decisions, we can even connect those to allow the AI to both select units more often and hit "buy" more often in cities with ores, and possibly build less buildings and more units in cities on enemy continents.

This will be difficult to code, but I think I should be able to do it now.

[Nelphine]

You know if you keep working miracles like teaching the AI where units are most helpful, we're going to expect impossibilities like the AI actually remembering what its scouted. (I actually have some ideas for general rules that might make that work.)

And in case I wasn't clear. Just saying you can possibly teach the AI that is fantastic. Keep it up, and thank you.

[Domon]

"This will be difficult to code, but I think I should be able to do it now."

wow. just wow.

[Tlaloc]

I agree the best solution would be to leave the unit-building costs unchanged and make the AI capable of using smarter unit-building tactics like players, if possible. That would be great.

[Seravy]

I agree the best solution would be to leave the unit-building costs unchanged and make the AI capable of using smarter unit-building tactics like players, if possible. That would be great.

Working on it. Buying part is done. Building part is being coded. It's quite a lot of work but I try to finish it in the next 20 hours...tight, but I want to see Hadriex going into the all Sorcery game using the new AI

[Tlaloc]

That should be worth watching.