NATURE Realm

[Nelphine]

Editted my last post a bunch. Will respond later, but it's called a wish list for a reason.

[Nelphine]

So with the 3 costs you've described you already use what my wish list wanted, except with the 'modifiers can increase base cost', which I think should be removed.

[Nelphine]

Re: spell blast and turn order, I think one of the most common scenarios where the AI uses spell blast is late game (where the human can have casting skill of 200+ or even 500+). This means that any time the human casts a spell that takes 2 or more turns, a spell blast 'should' by spell description be costing the AI (if the spell blast comes on the last possible turn) at least a full turn of casting skill worth of mana - in other words, hundreds of mana, which could actually be noticeable.

Also note, most spells by that stage of the game will be cast in 1-3 turns. If the spell blast is not charging the ai for that mana, that means, the vast majority of the 'cost' of spell blast is literally never being paid by the AI.

Further most times the ai spell blasts (and they spell blast a spell that costs more than 1 turn for the human), it will be because the human saw the AI casting a long spell (such as a 1000+ cost global or, especially, spell of return) and the human misjudged how long the AI would take to cast it, thus enensuring there was a full turn if spell casting skill being 'used' prior to the spell blast.



That's my reasoning why the AI should pay the full cost.


Now to address the technical problem.

If the turn order can't be addressed, can we add the cost calculation directly into spell blast: 

So, if an AI is casting spell blast, call the amount of casting skill the target has, increase the spell blast mana cost by that much, and then set targets leftover casting skill to 0.

[Seravy]

This means that any time the human casts a spell that takes 2 or more turns, a spell blast 'should' by spell description be costing the AI (if the spell blast comes on the last possible turn) at least a full turn of casting skill worth of mana - in other words, hundreds of mana, which could actually be noticeable.

On the other hand, if the player does not spend the mana into the spell, they also keep it.
So it's the question of "both players lose one turn skill worth of mana crystals" vs "no one loses any mana crystals".

Considering AI resources, not losing any is better. (the AI has more mana to spare 90% of the time, and in the remaining 10% they are already too vulnerable to mana loss.)

By spell description, Spell Blast should cost as much mana crystals as the other person used, and that is exactly what happens. And skill is lost regardless of the mana use, either you already spent it, or the next turn starts without you getting another chance to use it, so Spell Blast's cost never considered that.

the vast majority of the 'cost' of spell blast is literally never being paid by the AI.

Spell Blast has no relevant cost, that's why it's one of the top spells in the game. You get to cancel any amount of enemy skill for 50 of yours. Paying the same amount in mana crystals is at best a minor annoyance.


By the way, turn order is inherently an unfair thing that provides different advantages to different people.

Being first means you have the advantage of moving before others, but not being able to react as well to their moves. Going last means you can react to everyone well, but obviously, you have the least initiative in any situation. Spell Blast is reactive, so it favoring people later in the turn order is a natural consequence of how turn based games work.

Also, the human gets another advantage, the AI skill modifier. Even if the AI casts 2000 mana worth of spells, if they do it for a price of 1000, you can spell bast it for 1000 as well. Despite it meaning you are stopping a 2000 value spell. Furthermore the AI is generally dumber at playing against Spell Blast, meaning you will get a lot more chances to use it meaningfully. And since we are talking about a spell that cancels enemy skill at no skill cost of your own, that is a big advantage, even if it comes with a mana cost.

To begin with, the AI is holding back on using Spell Blast greatly. Losing less mana crystals is a minor unimportant detail compared to the advantage that brings to the human, as they are not required to hold back at all.

So, if an AI is casting spell blast, call the amount of casting skill the target has, increase the spell blast mana cost by that much, and then set targets leftover casting skill to 0.

Based on how difficult it was to make it remove AI skill advantage from the cost, I'm pretty sure I couldn't do that even if I wanted.

[Nelphine]

*shrug* I don't think it's the end of the world if it stays as is. I think it makes things a bit harder on the human, as they are an individual player, so overall their enemies get a bigger benefit than they do (by also casting it on each other), but you are right that's it's balanced per player.

And the technical side seals the deal. So I accept it not being changed.

[Nelphine]

Can AI use of transmute become any more sophisticated? (I know change terrain already has been discussed, and is probably as good as it gets, although it's got issues like changing the terrain that's 2 diagonal from the city and so doesn't actually help the city)

Specifically, seeing AI cities with 6+ adamantium ore always drives me nuts. Particularly when (as in this game) there are no wizards with corruption or raise volcano - only 1 chaos book in all 5 wizards, and that chaos book is owned by the wizard casting transmute.

[Seravy]

Probably can but it would be a lot of extra work.

[Seravy]

They also can't recognize shared tiles, can't use it to remove enemy adamantium, and can't use it to make adamantium for allies.

[Nelphine]

Making adamantium for allies is advanced enough I wouldn't worry about it.

The shared spaces is probably the most important. 

Removing enemy adamantium is also pretty advanced so is not super important.

Whether doing any change at all is needed, particularly without shared spaces, I leave to you.

[Seravy]

Well, I most definitely won't be doing this for this particular update, releasing the other stuff is priority right now. (and after that I'll need a break and focus on doing something else while also playing test games with the new sawmill). But some good news, it seems Transmute targeting is done in the same AI segment as most other stuff so there is a lot of free space. We will be able to make it more advanced, eventually.