[1.2] Looking at Victory Conditions
Having more differentiated victory conditions would really help with replayability. Right now conquest seems like the only real way to win. The other ways to win all require you to outpower the other opponents anyway. Even if you choose to do them instead of conquest they are usually less challenging because conquest is always somewhat different but the others usually play exactly the same. My main issue with all the side victory conditions is that none of them are competitive. You don't compete with other players to accomplish them and the AI never threatens you with them. Unless this is fixed they will never be more then tacked on side options. They are also not strategies that require you to plan ahead. After you build up enough tech or military power you just do them, and voila you win.
Conquest Victory
Not much to say here. This is the default way to win, and the only real way the AI wins by. Mass troops and take cities to conquer all comers. You win when every city is yours, or you enemies have surrendered. It can be fun but there should be other ways to win. Massing military power is a no brainer because this is the only real way you can be destroyed.
Alliance Victory
Right now in order to ally with anyone you need a late game tech and if you ally with all the remaining players you win. This has several problems. Firstly the lack of early and midgame alliances prevents a lot of strategies and really limits your interaction with other factions for significant parts of the game. Secondly it causes a lot of weird situations when it comes to alliance victories. For instance if you go on a rampage and kill 5 players but have one tiny ally you overlooked you get an alliance victory, because apparently you are a master diplomat. Secondly, you can't unally people to get conquest victories, you have to wait for the treaty to expire.
I would suggest splitting the alliances tech into 2. A earlier tech that unlocks alliances and the later one that unlocks the victory condition. This would allow players to fully interact with AI's much earlier but still prevent games from ending too soon.
Secondly I would suggest changing the condition for alliance victories slightly. You should be able to get a Diplomatic victory even if not every single player is allied to you. So say if your alliance has more then 60% of the faction power you win. Forcing players to destroy every non-allied player is pointless and not very diplomatic. Also there should be a wonder called the Sovereign's Council, with a high influence cost, that you have to build before you can get a diplomatic victory. This would prevent you from getting diplomatic victories by accident. With these or similar changes alliance victory would become more then just a side category of Conquest and would require it's own strategies and game plans. Also AIs forming alliances would be a real threat no matter your military power.
Magic Victory
Magic victory right now is just a generic civilization or wonder win. It really is what you do if your winning but can't be bothered to try and conquer the other players. All you do is get a certain amount of tech and then wait to build a bunch of buildings, and then cast a spell which is just even more waiting. That's really really boring. Magic victory should be spiced up and made much more competitive and riveting.
To do this I would suggest adding a new rule to the Spell of Making. This new rule would be that the caster has to have more Mana then any other player. The spell would be cancelled and uncastable if the player does not have the most mana. Thus players going for a magic victory would have to plan it from the start and compete amongst each other to amass the most shards, magic tech, and unique buildings. This would be much more organic and interesting then the current system. Like military conquest it would be organic, challenging, exciting, and require planning.
You could also make the spell not require the towers except for the forge, move them to different magic techs, and and make them wonders. That way instead of just taking up time and delaying an automatic win the towers would be late game boosts to Mana production players would be rushing to build them before the others to secure the most mana. Thus adding an additional layer of strategy.
You could alternatively use Mana produced per turn, total elemental shards, or a completely new resource instead of Mana as the spell constraint.
Quest Victory
I like the effort that has gone into this and the dedicated assets, however it is somewhat poorly implemented. It can be fun to do once or twice but gets old quickly and the AI can't win this way. To complete this victory condition you have to find the Pit of Lost Voices and complete the resulting quests. Again this can be fun but it feels tacked on and has nothing to do with the rest of the game at all. There is no competitive aspect to it.
I suggest stealing some ideas from HoMM 3. They had a "quest victory", where players raced to collect all the pieces of a map by visiting obelisks to find the treasure. Instead of the current system I think it would greatly increase re-playability if you re-added the dragon statues to the last quest, and seeded the random maps with several statues. Every statue would give the player a quest which when completed would reward them with a key part. In order to face the Dragonlord you would need a certain amount of key parts. Not necessarily all of them because players would be competing to get enough of them first. This way players going for this type of victory would have to go out exploring and race around the map to complete each quest first.
Alternatively you could have it so each statue would not be destroyed upon quest completion, and be able to be completed by every player once. Or have normal quests have a random chance of producing key parts. So actually investing time in questing has a reward and leads to a unique victory distinct from military might. Also the key could reveal and allow teleportation to the Pit.
