Research costs

[Seravy]

Zitro raised a good point btw with spells in treasure.

They were already greatly undercosted (2400 treasure budget for an average 6k value very rare) but with the new costs, they'd be massively undercosted in treasure.
And very rares from treasure breaking games was always a thing (like that 1408 Suppress Magic AI on Myrror once...) so idk, any suggestions for that?

(Not that books or retorts for ~1000 are anywhere fair, those things are priceless. So I guess we can just say treasure is sometimes way better than expected.)

[Seravy]

I wonder if I should put Unicorns into Life realms "cheapest uncommon" slot. It's not a very strong creature, but I don't think it's a spell that should be rushable early? Although that would actually help out life AI...so maybe do it? Or are they good enough now with their buffs?

Other spells I could put there are True Sight or Exorcise. But both of those are fairly decent counters to certain enemies even if not strong spells overall.

[Seravy]

Spells assigned to the cost table :

160 Earth Lore Floating Island Hell Hounds Star Fires Skeletons
200 Nature's Eye Nagas Fire Elemental Guardian Spirit Weakness
260 Resist Elements AEther Sparks Disrupt Holy Weapon Cloak of Fear
320 War Bears Guardian Wind Warp Wood Heroism Summon Zombies
320 Fairy Dust Phantom Warriors Fire Bolt Healing Black Sleep
320 Water Walking Confusion Wall of Fire Holy Armor Life Drain
320 Sprites Resist Magic Corruption Bless Darkness
380 Earth to Mud Psionic Blast Warp Creature Endurance Ghouls
440 Call Centaurs Focus Magic Shatter Heavenly Light Wraith Form
480 Web Blur Flame Blade Just Cause Mana Leak
420 Giant Spiders Aura of Majesty Gargoyles True Sight Lycanthropy
640 Land Linking Vertigo Chaos Channels Planar Travel Drain Power
960 Change Terrain Aether Binding Raise Volcano Exorcise Syphon Life
1280 Crack's Call Spell Lock Mystic Surge Astral Gate Blood Lust
1280 Ice Bolt Water Elemental Fire Giant Resurrection Wall of Darkness
1280 Cockatrices Flight Fireball Unicorns Night Stalker
1280 Nature's Cures Dispelling Wave Immolation Raise Dead Reaper Slash
1600 Construct Catapult Counter Magic Lightning Bolt Stream of Life Possession
1920 Transmute Phantom Beast Chimeras Divine Order Black prayer
2140 Great Lizard Spell Blast Fire Storm Prayer Shadow Demons
2500 Earthquake Flying Fortress Chaos Rift Prosperity Warp Node
3000 Elemental Armor Stasis Chaos Spawn Exaltation Evil Presence
3500 Stone Giant Invisibility Magic Vortex Mass Healing Famine
4000 Petrify Storm Giant Doom Mastery Lionheart Dark Rituals
4000 Iron Skin Banish Warp Lightning Altar of Battle Terror
4000 Blizzard Mind Storm Blazing March Angel Wrack
4000 Gaea's Blessing Uranus' Blessing Doom Bat Holy Word Cloud of Shadow
4500 Gorgons Air Elemental Efreet Invulnerability Wraiths
5000 Survival Instinct Wind Walking Doom Bolt Inspiration Wave of Despair
5500 Earth Elemental Magic Immunity Flame Strike Incarnation Zombie Mastery
6000 Earth Gate Creature Binding Chaos Surge Enlightenment Summon Demon
9000 Seismic Mastery Spell Ward Hydra Consecration Pestilence
12000 Herb Mastery Great Unsummoning Warp Reality Holy Arms Death Knights
15000 Regeneration Djinn Disintegrate High Prayer Annihilate
15000 Behemoth Haste Apocalypse Supreme Light Massacre
15000 Call Lightning Sky Drake Doomsday Call to Arms Animate Dead
15000 Great Wyrm Suppress Magic Great Drake Life Force Evil Omens
18000 Colossus Mass Invisibility Meteor Storm Crusade Final Wave
21000 Entangle Spell Binding Call the Void Arch Angel Eternal Night
24000 Fairy Ring Time Stop Armageddon Charm of Life Demon Lord

[Seravy]

By the way one more advantage of this system I just realized.
Even if the AI does have excessive power advantage (to support its crazy combat spending), that does not translate to such a large research advantage anymore if they are at peace, for the same reason as the human.
It's also less vulnerable to being held back in research too greatly by wars, as a larger percentage of power comes from buildings.

Research went from 19 to 27 in cities, while power went from 24 to 28. Nodes went 25-> 30 or something around that.

So in the old system an AI empire's which owned an entire plane with 35 cities was 665 RP and 1215 PP, now it's 945 RP and 1430 PP. So Research went from 35% to 39.7% in the mix. Well, not that huge but it matters. Since the intended spending is 50% and the AI is guaranteed to have 40 of that, if they spend evenly from the remaining 60% they actually end up overdoing research, but fortunately wars will drain mana and hold it back. In the old system the AI appeared to be slow on research merely because they also got close to none of it done in the first 5-10 years like a human. Sage's Guilds are not very high priority and cities still developing...

(I actually wonder how such a small increase of ~+20% overall resources will support the triple RP costs but the math says it does. Which can only mean we have been actually spending a fairly minor portion of our assets on research without realizing. Let's see...overall RP cost of an entire realm in the old system was 81000 RP. 35 cities fully built plus nodes produced ~2k. So...40 turns to breeze through the entire spellbook once controlling Arcanus. That's 3 years if you have peace. Or can actually last 3 years on gold to mana conversions and treasure even during some less severe wars. No wonder it didn't feel right. Even with half of a plane you could get it done in 6 years...when the intended timeline for a game was about 25 from start to the victory screen so research for like 20 of that. And let's just not even think what happens on a Huge map with 50 cities/plane. Maybe I should have started with this calculation, lol. It's so painfully obvious if I look at it like this.)

[Nelphine]

Things I would change:

Commons:
Nagas should be more expensive; I would swap for guardian wind.

I don't like the chaos arrangement at the high end of common, but I'm not really sure what to put there. None of their spells really deserve the extra cost. I would swap hell hounds and disrupt though.

I would swap resist elements and water walking.

I'd swap wraithform and mana leak.


Uncommons:
I'd swap cockatrices and construct catapult. I'd swap land linking and nature's cures.

I'd move phantom beast to 1600, aether binding to 1920, spell lock to 960 and counter magic to 1280. (I don't like vertigo at 640, but I don't like any sorcery for 640. I'd also like dispelling wave at 1600, but not as much as phantom beast).

I'd swap fireball and raise volcano.
More to come soon.

[Seravy]

Ok, will do the adjustments tomorrow morning.

One more thing I realized that should have been alarming :
This is a strategy game. Basic gameplay is pretty much : "I see you have X. I could do Y to counter that. So I produce Y. Now I strike at your X and have one more city. Or two. Or however many. Unless I assessed the situation wrong and Y is not effective vs X or I'm incapable of making Y in a timely manner in which case I lose." Or at least it should have been like that. But my experience was more along the lines of "I see you have X. If I had Y now, I could beat you but I can't afford waiting to produce Y because by the time I have enough Y to attack your city, you'll know Z and I can't win against that. So I have to take the risk of attacking with what I have and...damn, I lost again, maybe I should not play Expert anymore, it's too hard." (where Z might be fire storm, a city defense spell, a combat spell, a new creature, anything higher tier than mine, really.)
Well no, the difficulty wasn't too high (or who knows...), but the game was so fast on research by the time I could have executed any strategy whatsoever, it got outdated. Cool, I finally have my 9th Shadow Demon, let's attack...what, they have Holy Word already, damn. I could beat their fortress with 2 stacks of 9 ghouls but I can only summon one stack. By the time I get a second, they'll have shadow demons in their capital. So let's just not beat this guy and focus on the others... I could annihilate that guy with two stacks of 9 Efreets but then they have time to research Famine or whatever and I lose half my empire. So let's strike with only one stack...arrgh it didn't work, now I'll get both the famine and have no armies anymore, but I had to try. And this kept happening.
Generally the AI already knowing a spell wasn't the problem, since I could deal with that otherwise I would have lost to it before getting to attack them. The problem was losing to a spell they were expected to "know soon".
And obviously this "strategy inflation" hit normal units the hardest since they were lower tier in the hierarchy than spells. "By the time I produce these 9 Paladins, the enemy gets Flame Strike" sort of thing. Or "By the time I have my 9 Cavalry, they'll have....<Insert any uncommon here>".

[zitro1987]

The treasure points of spell treasure currently at 1-4-9-16 times a constant.

It would be better to have a more exponential pattern like 1-3-9-27 (times 3 each time) or 1-4-11-22 (+3, +7, +11)

*Not a bad concept discussed in thread though the rp numbers are so extreme, I struggle wrapping my head towards not choosing 'mana' in normal scenarios, especially early on. An impossible suggestion would be to increase the first uncommon or rare or v rare you research by a number (or reduce the rp speed for that first spell)... making it feel like you hit a new tier.

[Nelphine]

Uncommons continued:

Firestorm to 1920; Chimeras to 1600; Lightning Bolt to 1280

Prayer to 1280; Astral gate to 1600; Stream of Life to 1920; Divine order to 2140

Swap Black Prayer and Possession


Rares:

Earth Elemental to 2500; Earthquake to 4000; Blizzard to 4500; Gorgons to 5500

Swap Mind Storm and Air Elemental

Swap Warp Lightning and Chaos Rift; Swap Doom Bolt and Efreet

Invulnerability to 3000; Exaltation to 4000; Altar of Battle to 4500

Wrack to 2500; Terror to 3000; Evil Presence to 3500; Warp Node to 4000; Famine to 4000; Swap Wave of Despair and Zombie Mastery


Very Rares:
Swap Herb Mastery and Regeneration; Swap Fairy Ring and Entangle

I don't like Great Unsummoning at 12000 (nor Spell Ward at 9000), but, I don't really have anything better for either slot, so it seems the best it can be.

Swap Call the Void and Meteor Storm

Swap Call to Arms and Holy Arms (although, I'm not really sure about this, so, probably doesn't need to be done.)

[Suriname]

Is cost going to be used as AI trading desirability as well? Or is there a different table for that?

[Seravy]

Nagas - GW : ok
Resist Elements - Water Walking : ok
Wraith Form - Mana Leak - Disagree, we want Weapon Immunity and transportation early, if anything WF should be faster but it's a really good spell...on the other hand Mana Leak is useless for a while as nothing early does magical ranged except ghouls.
Cockatrices - Catapult : Unsure, I feel catapults are better. In fact I have a "consider raising casting cost" note on that. Also want at least one of the three summons in a middle slot.
AEther Binding : This is something I often rush research for. So I want it to be reasonably accessible, even if not early, but not too late either. Being able to do that and get ahead in skill production from ~turn 50-60 onwards is one of the merits of playing Sorcery.
If we want cheaper Spell Lock, I can swap it with Vertigo. For cheaper Counter Magic we can swap with Dispelling Wave. We can also move Beast down 1 slot if needed while doing that.
Fireball - Raise Volcano : Cheaper Fireball would definitely be nice but not sure I like more expensive Volcano. What if we swap with Chaos Channels instead?

Firestorm to 1920; Chimeras to 1600; Lightning Bolt to 1280 : You didn't specify what goes to the 2140 slot which is probably the reason for this whole change?


Prayer has always been the most expensive Life spell to research and I rather keep it that way, it is that good. That however prevents us from shifting the other 3 upward.


Black Prayer - Possession : unsure, I think this change would hit AI a bit too hard, Possession is their first pick so it directly defines how long it takes for AI to reach UC in death. Might even be better to swap Possession with Syphon Life (as they are competing spells the AI won't do the SL if they did get Possession to show up) 

Nature - Earth Elemental is very powerful and this would put it really close to the other combat summon (Catapult). For more expensive Earthquake we can swap with idk, Gaea's Blessing maybe? Human is unlikely to pick up Earthquake at full price though.

Mind Storm - Air Elemental : Ok

Warp Lightning - Chaos Rift : Don't want it that close to Lightning Bolt, also it's a VERY powerful spell. Meanwhile Rift takes time to be effective.


Doom Bolt - Efreet : Ok

Life : Isn't Invulnerability way too good for the cheaper slot? Rather swap Invulnerability with Altar of Battle.

Death : We agreed on city curses being in the cheap slots in the old system. Changing that would be a major shift in the game, and would mean we need to delay AI research of these things most likely. Currently they are all early picks done before the second half of uncommons.

Herb Mastery - Regeneration : Ok
Call the Void - Meteor Storm : Ok, this will help the AI too, great idea!
Call to Arms - Holy Arms : Rather not swap these.

Is cost going to be used as AI trading desirability as well? Or is there a different table for that?

AI trade desirability is a separate value and I have no plans to change that. How useful the spell is for the AI should be the major factor there, not the cost.