Quote:
Gandalf Parker said:
Agreed. There have been some threads on that. Which scales and pretenders the AI uses best. And of course some testing might be a good idea before posting it. Some fantastic options open up when you realize that you can not only build the god and scales, but you can select a full set of equipment to give that god. Surely that sparks some evil creativity to see what you can create and how that would play as an AI.
There are other things that can be done also. Adding a magic site or two. Specific ones that help to push the AI in the "right" direction. Such as, giving the AI more resources tends to push it into making more elite troops.
Also giving an extra castle or two for a head start. Extra units (kindof like giving them pre-purchased mercenaries to start the game with).
But scenario maps ARE an option, and one that we havent seen many people trying. Im all for anyone being brave enough to put one forth as "my effort to create a better badder Ermor", or Ctis, or Machaka, or Mictlan.
|
In my experience, the most important adjustment needed by the AIs is a growth scale. Also typically you pray the AI doesn't choose a large pretender, because otherwise it will blindly send him to the death match arena. Nothing is more SAD then watching a bunch of AI pretenders blindly kill each other from the arena death match. The AIs don't build SC's either so typically on a large map I'll give them a half a dozen at the start of the game. The AI almost always does poorly at research, but if you give the AI a library at its capital it can keep up with the human opponent.
In Dominions_2 I use to provide a few AIs a magic site with blood income, a few immobile commanders with blood magic, then have them start the game with specific blood spells. This allowed the AI to wisely use blood magic where it previously never even researched blood magic. I also would have them start the game with dome spells. Every weekend I have a different level_9 astral mage cast WISH so the #startspell gets fixed in Dominions_3... maybe one of these days.
