This is a response limited to the propositions of AI DLCs:
The primary reason why we don't make "AI only" DLCs (with only AI improvements with any GUI additions that is) is that it would be hard to sell, few people would buy it. Different people have different expectations of what constitues "good" AI, some want it to play in a very railroaded historical manner, others like min-maxing human opponents. The "skill level" of the AI as a function of the time and effort put into it tends to be logarithmic, which means heavy investment into an incredibly skillful AI doesn't necessarily yield good results and certainly isn't a good business decision.
Furthermore, it might not even be possible to make a (much) smarter AI that doesn't slow the game down to a crawl. I'm often considering improved algorithms for e.g. army movement but have to discard most ideas because they would be unrealizable due to the large number of armies and potential moves. The ideas I do think could work I have to wait with until my vacation with prototyping because it is a very large and uncertain task.
The primary reason why we don't make "AI only" DLCs (with only AI improvements with any GUI additions that is) is that it would be hard to sell, few people would buy it. Different people have different expectations of what constitues "good" AI, some want it to play in a very railroaded historical manner, others like min-maxing human opponents. The "skill level" of the AI as a function of the time and effort put into it tends to be logarithmic, which means heavy investment into an incredibly skillful AI doesn't necessarily yield good results and certainly isn't a good business decision.
Furthermore, it might not even be possible to make a (much) smarter AI that doesn't slow the game down to a crawl. I'm often considering improved algorithms for e.g. army movement but have to discard most ideas because they would be unrealizable due to the large number of armies and potential moves. The ideas I do think could work I have to wait with until my vacation with prototyping because it is a very large and uncertain task.
- 10
- 4