Magic Spirits are OP

[Nelphine]

Right and I think this would be a perfect example of why the AI gets upkeep reduction. Because we can't make them smart enough to know how to utilize certain units. So when units might be too expensive if used incorrectly, the AI has the reduction so that it can afford to use them incorrectly.

So I don't think an increased upkeep would be an issue for the AI at all.

[Seravy]

I had a feeling I read about this exact topic a long time ago somewhere so I tried to dig it out...
and behold, I found this idea :

For the time being, I have another suggestion. Let's give both Spirits 0 scouting (I can do it I guess). They will not reveal any terrain, so they'll be useless for scouting.

If we decide spirits shouldn't be used as scouts, this is probably the best option. (I'm not sure that's needed, honestly I'm used to doing that and taking it away would feel really weird - yet it's a fact scouting is really easy.)
The downside of this solution, if you move the spirit towards an unknown tile, you might attack someone, so Spirits will be literally unusable for scouting, not just ineffective.

btw if we do the research option, it comes with the risk of the AI not researching it early enough - though I can adjust the research priorities to make sure they do, preferably after other priority research such as werewolves, spiders and gargoyles.

[Nelphine]

So, one if the big reasons I'm concerned about this, is that (aside from cities) AI already know the whole map. Sure its a simple form of cheating but it does give them an advantage (especially for early settling)

Magic spirits are early enough to counteract this.

I would love to make scouting more meaningful, but I'd also like to keep the AI roughly equal with the human.

Would you Seravy be interested in looking into making the AI care about scouting? As I mentioned I have some ideas for how to do that, but its a big project, and no matter how well meaning I am, I don't know how the various pieces of code will interact so some (or all) of my ideas are bound to come across as either stupid or impossible given the existing code.

[Seravy]

Would you Seravy be interested in looking into making the AI care about scouting? As I mentioned I have some ideas for how to do that, but its a big project, and no matter how well meaning I am, I don't know how the various pieces of code will interact so some (or all) of my ideas are bound to come across as either stupid or impossible given the existing code.

I'd like to hear your ideas but I'm 99.9% sure it can't be done. To begin with, an AI cannot work with incomplete information, so if they had to obey normal scouting rules, they couldn't decide what to do at all.

[Tlaloc]

I like having something to cast and explore with right away. It increases game speed and interest level. I wouldn't want to see it delayed. If something needs to be done, just increase the maintenance by 1 mana or 1 gold. The AI units know where your capital is from the start, so I don't see this as imbalancing.

[zitro1987]

I don't see the problem myself - a lizardman spearman is far cheaper for scouting. I think the 1 upkeep mana is fine as it is.

However, I like the idea of not starting with a researched magic spirit is quite interesting, but I'm not sure how I feel about allowing a player to randomly start with more advanced ones like dispel, detect magic, etc
*Maybe making magic spirit, summoning circle, heroic shout, enchant item, detect magic, and dispel magic all cost around 75-125 research points would make it more balanced. I don't think any of these spells are inherently superior to each other.

[Catwalk]

Scouting 0 would also do the trick. I see several good solutions to this problem.

Tlaloc, if you like having an option available to you right away you can simply choose starting summon spells.

[Catwalk]

zitro, a magic spirit is much better for scouting. It costs mana rather than food, always moves two and can walk on water. No other means of exploration are necessary the way things are.