They can always be sent into battle.
well, that would be better than a bunch of adepts with all combat. Let them build axemen for whacking, for crying out loud, and save the casters for a-casting!
Ok, let them two argue that out.
Im all for this concept of making the AI use magic better
I agree.
If you want more interesting games you definitely need a more interesting game setup... easier than tweaking the AI
I disagree - different game setups really don't change much. The initial 100 turns (depending on game speed, map size, etc.) all revolve around getting techs to make use of the nearby resources, and building enough units to defend your cities at the same time as expanding your civ. This is what ALL setups have in common - sometimes this will be a very short period (especially if you start with techs, or premade scenarios with multiple units/cities), and sometimes it can be a very long period (huge map with only one other civ).. This is the period known as masturbation - you play with yourself.. the AI has almost no input into your gameplay.
The next period is where the interesting part of the game really is - once borders have been set, and civilizations have set their tech choices for what type of win they are going for, then comes the point where good games are made. Sure, you can give the AIs a bunch of stuff so that the player never gets past the first period - but that doesnt make for an interesting game. Interesting games are made by a multitude of roughly even - but very different - civilizations meeting and trying to win from this point.
But since the AI is unable to effectively use Magic (which is a big component of this Mod) - the second period is stunted because of the player's ability to properly use magic, thereby giving the player a HUGE advantage once they get to the part of the game where Mages and Archmages are being built - which almost always takes place during the second period of the game.
The final period of the game is either the Player running so far ahead that the computer doesn't stand a chance - or the player was beaten up so badly during the first period that the second period was almost nonexistant and the player does all they can just to hang in there until the time expires (or more likely, they just exit the game and start over).
Two answers -- A) Currently the AI almost never trades until very late in the game, resources available for trading are incredibly scarce.. it would seem to me that a Fire mana (arguably the most valuable) should be worth more than the cow+corn that you can get for it when someone finally is willing to trade. B) Apparantly, the team is working on revamping the trading system - so the above issue is likely to be modified... then we'll see how valuable that extra fire mana is.
>>>>>>>>>> Long post - I know.. But I really think that The AI CAN be taught to use magic better. but it'll take time because the mod itself is still being developed and because of the complexities involved. It may be that simply changing the AI Weights involved in which units to build could free up resources - which could speed up research - which could enable the AI to develop more magic based units, and that could increase the chances of seeing the AI use magic properly. Or maybe each spell needs to be handled completely seperately. Or maybe the AI for the Adepts needs to be modified so that they know what promotions to take. Or maybe it is a combination of all these things.
EDIT -- Oh by the way, I think that Tsunami should go away - I doubt that the AI will every be able to use that spell properly.
But spells like Fireball and Meteor are, to me, incredibly important that the AI learns how to use them - That first water spell is pretty good too. In the next few days I'll take a look at the AI for the Mages to see if I can see anyway to make them more likely to cast fireballs -- but for the time being I'm still testing my changes to the actual fireball to make sure that, once cast, they continue to attack as aggressively as I can get them to.
