NATURE Realm

[Seravy]

I wanted a different name but...they are taken.

Planar Gate - implies planar travel.
Astral Gate, Earth Gate - existing spells.

I know!

How about Fairy Ring?


https://en.wikipedia.org/wiki/Fairy_ring

[Nelphine]

Ehhh, the higher end nature units don't seem like they'd come from a fairy ring. I'd much prefer gates of hell.

As to what to summon I'd suggest 25% uncommon, 50% rare, 25% very rare, preferably with a modifier based on manager spent by the triggering spell. So the numbers I'd suggest would be for 700+ and for each 20 mana lower, you'd have a 1% reduction in very rare and a 1% increase in uncommon. (So a 320 mana triggering spell would have a 6% chance of very rare and 44% chance of uncommon.)

Reasoning: AI cast a lot of rare summons that will trigger the summon. Don't want this to cause natures wrath to summon very rares in most cases, or the dispel priority ends up being just as bad as the current natures wrath.

[Seravy]

Gates of Hell sounds too much like Chaos or Death, same problem as Gates of Heaven being too "life".
A Fairy Ring sounds very Nature-like though - and you don't always get the high end units.

For the rarity, I came up with this idea :
For each 50 casting skill, roll a D2. For each roll of "1" increase the tier by one. (common is tier 0, uncommon is 1, rare is 2, very rare is 3)
So at 100 skill, the average result is an uncommon, at 200 it's a rare, at 300 very rare.
We can tweak the number to higher if necessary - 60 or 75 might be better, resulting in 360 or 450 for averaging a very rare respectively.

The spell is meant to help maintain your advantage, and not "catch up" to someone much stronger, so it should reward the stronger user and penalize the weaker user through the tier of creatures, and Skill is probably the most accurate measure of a wizard's power. (And AI is usually ahead in it, which is important - AI is less efficient at using the creatures and needs more. Note that this means raw skill, without AI bonus - their advantage comes from building towers faster and having more power to spend on SP.)

Do note I don't want the result to scale based on what triggered the spell.
This indirectly penalizes the player who is behind because the enchantment's owner gets more relative benefit compared to what that player was casting - and that is a perfect match for the intended role of "maintain position of advantage".
As for the cost, NW's 301 was carefully selected and spell costs adjusted to trigger or not trigger it, so it's perfect. Every rare or very rare triggers it, except Stone Giant, but I think it's fine for Nature wizards to have one rare creature that can avoid this triggering it. In fact they need that exception, unlike other realms, their focus is on summoning.

[Nelphine]

I said gates of hell because I think this should be a chaos spell

On average, if you have to research this spell (as opposed to having it in treasure), you're going to have 300 casting skill when you cast it. That means its going to get very rares the nahoeiry of the time. That means the AI opposing the spell, cannot afford to summon rares, and if the spell is owned by the human, the AI can't afford yo summon very rares. That means the spell has to have a dispel priority equal to the current natures wrath. That means while the unfun gameplay of dstroyrd building is lost, the unfun gameplay of 'yay dispel wars' is still there, defeating the main purpose if the change.

Any time this spell averages a results in a very rare for the human when triggered by an AI summon, the dispel priority will need to be that high.

[Seravy]

Npte that "averages a very rare" still means only half of them will be very rares, those that do have the average or higher roll. Below average rolls will be worse.
It's only an average in theory - in practice, rolls that are "better than very rare" can't balance out "rare or worse". But we can change the number any way we want to.

[Seravy]

But I understand, it might be better to use a higher skill amount, and use skill modifier by AI casting advantage.

[Nelphine]

I don't think you can ever afford to average very rares. No AI can afford to summon very rares for the human, ever, without a numerical advantage on their side. (Which is why my numbers end up with at most, 1 in 5 being very rare when the triggering ai does any sort of summon); if the human is getting very rares when the AI casts globals, that's OK because that doesn't occur frequently.

And even against another AI, the triggering ai can't afford to give them very rare summons, when the triggering AI is only getting rare summons - the very rares just win hands down.

[Seravy]

Or, instead of the AI casting advantage, it's easier to add the difficulty level directly.

Let's try setting goals.
The human player should probably perceive a 100% increase to their summoning capability with this in play at, idk, 400 skill maybe?

So, we need to get 400 worth of creatures per turn at 400 human skill - which equals ~0.8 very rare.

Assuming there is a total of 800 casting skill among surviving AI (this late in the game it's either one big AI or 2-3 smaller), and we play Expert, that's 1600 worth of spells. That will be an average of 4 triggers if the AI only casts big spells, but they sometimes don't so let's assume 3 triggers. (Note that on lower difficulty it will be far less)

Unfortunately, the stronger the opponent, the more triggers we get, so the stronger this spell becomes - this scaling is not desired but it can't be helped - fortunately the need of own casting skill at least makes sure it won't work from a very underdog position.

Assuming 2 rares = 1 very rare and commons/uncommons are worthless, we want 0.8/3= 0.266 very rares per roll. So a 13.3% chance for very rare, and a 26.6% chance for rare, with the rest being common and uncommon.

13 times (D10=1) will have that percentage chance on very rares. For rares it results 38% which is higher but we can get away with it if we say 3 rares are worth a very rare, not 2. This still leaves 50% for common and uncommon.

So we need a formula where

(Skill/X)=13 at 400 for the human.
That means X=30 - each 30 skill allows an extra D10 roll where 1 raises the creature tier.

Now, the AI.

The AI will generally get a similar amount of triggers - slightly less because the human player has no casting advantage but other AI do. So for the AI we can assume 2 triggers a turn.

For the AI we'd like the spell to be more efficient - it should raise their creature output by 100% at as low as 200 (raw) skill.

So at 200 skill the AI gets to summon 400 worth of creatures like the player - so they will want 0.8 very rares from 2 triggers.

0.8/2=0.4 so 20% chance for very rare and 40% for rare. 16 rolls yield a 21% which is close enough.

(200/30)+AI advantage=16

Rounded, that means AI advantage = +9 rolls. This would be for Expert and higher difficulty. Expert is level 3, so +3 per level of difficulty sounds good, meaning the complete formula is

(Skill/30)+if AI (Difficulty*3) number of rolls with a D10 where each roll of 1 raises the summoned creature's tier.

[Nelphine]

I'd rather not see it summon commons at all. At that stage of the game they just use up unit slots for no reason. I'd like a condition that if no 1s are rolled, it still gives an uncommon.

Those numbers look good.

[Seravy]

That makes sense, yes, I can do it that way.