Early game summoning over-represented?

[Nelphine]

Exactly.

[Seravy]

Let's try to think about how much the buildings should cost.

Barracks - 1 additional level is fairly significant, +1 attack, defense and resistance. That actually is sometimes better but at least as good as the Hit from the Alchemist's Guild. However, the AG also provides magic weapons, generates power, and unlocks special ores so I'd say Barracks should be cheaper.

Alchemist Guild - This I'm unsure about. On one side it should cost more - the effect is powerful even for a higher tier unit - but on the other side it should not : if it's not cheap enough then early weapon immunity becomes overpowered. We can raise it a little as those weapon immunity spells will take longer to use and production builds this faster, but we can't go overboard. Probably have to keep it in the 150-300 range.

Library - This is part of AG's cost so raising it is something we can consider. Doing so should probably mean a higher research output : 2 research/turn takes 50 turns to get a common, 200 to get an uncommon spell. Of course, more cities mean more of these but even with 5 cities it's 10 turns for a common, 40 for an uncommon from Libraries, that's ultra-slow. 2/turn feels irrelevant even compared to the starting power of 10-24. If this produced 3 research but for the price of 60 production that might be better?

Wizard's Guild - Aside from requiring the Alchemist Guild in addition for making Magicians, no other change necessary but if we raise Library, we probably wan to reduce research on this by the same amount that got raised.

Stables - This I'm unsure about. For most races this merely unlocks cavalry -  A unit only slightly better than swordsmen - so a cost raise doesn't feel necessary. (in fact, I think cavalry is one of the underused unit types as is.)
But for some races this unlocks really amazing units that annihilate swordsmen tier units with ease - Nomad Horsebowmen, maybe Dark elf cavalry (ranged with 5 moves) and Gnoll Wolf Riders coming to mind. Those rival or even exceed fighter's guild units in how useful they are yet the building is much cheaper.

Fighter's Guild, Armorer's Guild, Fantastic stables - no idea how much cost raise needed if any. Raising it will make slow races worse which is contrary to the goal, but not raising it means people can easily skip cavalry and go straight for halberdiers which counters cavalry... (although the ability to do that definitely helps the AI to avoid cavalry-rushed all the time which might become the new threat with easy early summoning gone.)

War College - I think fine as is, 450 is fairly steep.

[Nelphine]

Centaurs come to mind as an outlier. They and horsebowman perform almost the same role, but centaurs require a fighters guild.

I'd suggest either make then not require a fighters guild, or all the racial units from stables should also require fighters guild.


On topic. I would not change library. Having a cheap building early on is useful and I'd rather have AG cost in AG.

That leaves:
Alchemists guild
Stables
Barracks
War College
Fighters Guild
Armorers Guild
Fantastic Stables

Stables I'm going to skip as I think a discussion of the racial units it allows, and when we want those units, is required before worrying about cost.

Alchemists guild: as an economic building, this is covered by maintenance. In theory, 2 power should be 1 maintenance. And it should cost between 40 (2 rp) and 120 (2 power that can be enhanced by dark rituals and cult leader). Therefore, I'd guess that the AG should cost 100 as an economic building, and have 1 maintenance.
However, it doubles as the single most important military building. +1 to hit and nullifying weapon immuniy is better offensively on all non shaman/priest/magician/warlocks than a barracks, unless the unit has less than 4 attack power. The barracks offers basic defense though. So the alchemists guild is better for say, anythjg with more than 5 attack. (Note, if mithril or adamantium or orihilucon is present, the AG is flat better.) For units with 10-12 or higher attack, the AG is better than the war college (without ores).

I'd suggest a minimum 300 cost, upwards of 400, and 5 maintenance. (Alchemists historically have very high maintenance even when they did very little for their employers.)

Weapon immunity: uncommon or better will take long enough that they're fine, alchemists guilds will still exist.
That leaves fire elemental and wraith form as potential problems. Chaos is super weak at common (basically only flame blade, and fire bolt against weak things are universally good) so I have no problem with fire elemental getting an indirect buff.
Wraith form is more problematic - this spell was a huge discussion to make common when we made that change, so I'm loathe to give it more power. I'd be tempted to actually give wraithform a -2 penalty to armor to compensate the AG change. (This also indirectly helps nature against heroes, since nature almost requires cracks call to fight decked out heroes, and it's easy to get wraithform to immune it.)
More problematical is the alchemy retort. This would probably have to be changed to gold-mana conversion only. Which would be sad for taking away synergy for experienenced players, but would be more AI friendly anyway.

Barracks and war college. Barracks has no maintenance, and is cheap. It also doesn't do a lot. I'd raise to 200 at most, assuming 300-400 for AG.

War college is a lot more problematical. First of all, +2 attack and defense is as good asland linking (an uncommon spell), but its actually probably better because city troops typically have more figures than fantastic summons. Land Linking costs, what, 800 to research? And 50-70 per cast? War college + Barracks should be comparable; but you have to build it in multiple cities. So, assume 5 military cities, that each build.. 10 troops; that needs to compare to 3300 power production costs (not even including increasing casting skill). Production also isn't as valuable as power, so assuming production is worth 80% of power, and assuming an arbirtary extra 500 for casting skill, then 5 barracks + war college should cost 4750; so 1 barracks + war college should be 950. Assuming barracks is 200, that makes war college 750.

I wouldn't change the maintenance.

(Yes war college is a HUGE part of the power of my bezerkers. I would absolutely pay the cost increase I'm suggesting for these buildings, and it would make my current impossible strategy that much harder to do.)

As a note, warlord will also be indirectly nerfed by this. Warlords most amazing feature is letting units start at elite for the +1 HP. This is certainly worth more than 2 picks. But if there is no war college (or there is an altar of battle, or heroism spell), then warlord is worth less than 2 picks.

Increasing cost of war college will increase how often warlord is not bringing a unit to elite, which will lower its average power to closer to 2 picks.


Armorers guild and fantastic stables - these buildings very rarely make units that are obviously better than a war college + fighter guild unit. Different, and situationally better, yes, but not huge amounts better. I would want both of these to be around the same cost as the alchemists guild (300-500 range, 3-4 maintenance).

That leaves fighters guild. Which relates to the discussion on stables. I would also suggest comparing to common/uncommon summons, and considering the research cost of both, and deciding that way. This should be at least 325 (to make the human take the same time to build it in capital with an extra 4 population), up to 600. (Yes, I think its OK for fightes guild to be more expensive than armorer's/fantastic. The difference between swordsmen and halberdiers is HUGE.)

[Seravy]

Oh yeah, I forgot Shrine. We might also want to reconsider its cost - being the only thing that produces power early, and 2 being fairly low, 120 is quite expensive, although it does allow the pretty useful Shaman and reduces unrest. Still, compared to a Library, Marketplace or even Stables it feels a bit inferior? 1 Unrest reduction might sound good but in reality at this stage of the game it's either 1 food and 20 production (2 spearmen) or the loss of 1-2 gold a turn (dropping back tax rate 1 level), it's rarely a benefit of actually removing a rebel that can't be removed cheaper.

I mean a Library gets you 2 for 40, sure the Shrine does more and power is better but triple cost? idk about that...can't remember a time when I last built a shrine before turn ~70+.

Weapon Immunity : changing spells is not on the table, especially not Wraith Form, Lycanthropy (no, it only takes 10 turns to research it and it can be guaranteed), Shadow Demons (ok, this is unlikely to appear early but it's not impossible to go for it, takes 20+ turns with Sage Master but it IS that powerful and gets you to the other plane too!), or Fire Elemental.
We need to set a price with that in mind.
I'm absolutely not changing Alchemy either.

AG Maintenance is irrelevant : you only need it in the few military production centers which means an insignificant cost, but at higher maintenance, the other roles (nightshade, power generation) suffer greatly as those are not worth paying 5 gold for, even if ignoring the building cost. So you build even less of these outside of military cities which is the exact opposite of what I want.

Stables Units :
Centaurs should stay as is.
Horsebowmen, idk. Without barracks and alchemist these actually aren't that good. They just don't do enough damage. So more expensive barracks and alchemist = weaker horsebowmen. (this is also true for longbowmen btw)
Wolf Riders have 66% more health and somewhat more attack power than Cavalry but their per unit cost is double and they have no first strike so they aren't directly better, the problem is you can have a lot more power in your stack, enough to fortress spike using it...although producing 720 worth of units might just be too slow, even with the extra population - probably needs over 25 turns to build such a stack. Their 5 pathfinding movement is worrying though, 4 might be better, that still means they are faster than normal cavalry outside battles on most terrain.
Oh, and the cost does not stop there, you want an alchemist guild for that otherwise the first fire elemental or wraith form stops your "doom" stack.

Library : Making it 3 research without significantly raising the cost (only to 50 or 60) is an option. Although that devalues the University so maybe it's a bad idea.

[Nelphine]

Aren't wolf riders 100, so 2.5 times the cost?

And right now AG is really an extra cost on amp tower. It doesn't natter what you make the AG cost, because you need it to get the amp tower. And an amp tower is worth like 2000 production. So everyone has to get one, and they're a bargain. I already listed the cost I'd be willing to pay for an AG for its non military non-required-for-amp-tower benefits (I even forgot nightshade, and that literally makes 0 difference to be. I wouldn't pay 10 production for nightshade.)

So economically, 400 +900 for amp tower is a bargain. Militarily, 300-400 for magic weapons and ores is totally worth it. You make the maintenance high to show that people shouldn't buy it unless they need the military or they want the amp tower. The power production and nightshade use us literally complete icing in the buing so that it does SOMETHING while the amp tower builds.

Horsebowman: they kill sprites, hell hounds, war bears, naga, phantoms, chimera (if in small numbers), werewolves, earth elementals, and halberdiers. Anything with 5 or less armor.

[Nelphine]

Edit: also, how are you going to get werewolves in 10 turns and cast them?? With reduced starting skill/mana, you'll have at most ~15 power production. With sage master that goes to 25 research. So, in my 10 death sagemaster specialist game, OK, you could research them in 7-9 turns, which us perfection. And then you'd have 0 mana built up, so it would take you 6 turns to get the mana to summon 1. So yes you could do one in ~15 turns, but werewolves are only armor 9 units. They aren't invincible even with weapon immunity. And that's IDEAL circumstances.

And right now you can do it even faster, and no one builds alchemists guilds right now in the first year, so if its not a problem now, why would it be a problem with this change?

[Soviet elf]

By the way, barracks provide only +1 at and +1 res (for non-warlords), no defense bonus for first level.

[Seravy]

Aren't wolf riders 100, so 2.5 times the cost?

I thought so but the file clearly says 80. Probably got lowered when Gnolls got buffed.

400 +900 for amp tower is a bargain.

I was talking about the maintenance there not the building cost.
The building cost can't be too high because of weapon immunity and alchemy. That is not an economic matter.

Horsebowman: they kill sprites, hell hounds, war bears, naga, phantoms, chimera (if in small numbers), werewolves, earth elementals, and halberdiers. Anything with 5 or less armor.

Being able to damage and being effective is not the same. If I lose more (or even equal) value due to AI fire bolts or ranged units next to those (or their army cornering horsebowmen, with move 3 nagas, large stacks can kill some) then it's not a viable strategy. So no, being able to kill 1-2 bears or nagas while I lose 2-3 horsebowmen for it means they weren't worth producing. And that's about how good they are at level zero and normal weapons.

Besides, Nagas and Bears move 3. Attacking from 4 tiles has a ranged penalty. So a level 0 horsebowmen won't be able to hurt either of those effectively. I tried, without barracks and alchemist guild they simply don't work. (and then I haven't even considered the possibility of a city wall or guardian wind or buff on the unit...endurance, heroism, etc on a swordmen...)

Against phantoms, yes, it works. Hell Hounds, it kinda works if there aren't too many but 9v9 half your horsebowmen will die because they will reach and corner some. In a Chaos node, you can forget about it, rough terrain slows your hosebowmen and the hounds have more armor. Sprites and Earth Elementals are easy. Halberdiers, usually easy as well unless the wizard has Endurance on them but at such a low attack power I'd expect to run out of arrows before killing the entire stack (and then retreat and no experience so still level 0)

They aren't invincible even with weapon immunity.

At 36 health per unit and 2 health extra per turn they are pretty close to invincible at 9 shields against most early units.

Not that it matters. The alchemy retorts alone is enough reason to not raise AG cost to 400. And then there is Fire Elemental and Wraithform.

[Nelphine]

I know I've argued for not changing the retort myself in the past. I understand the reasons why not to do it then.

But if you make halberdiers harder to get, I think you have to make alchemy harder to get - because a mithril weapon swordsmen is better than a halberdier. And mithril isn't that uncommon (I'd say on poor minerals, at least 1 wizard has it out of every 2 games on their capital.). And with reduced spell casting, and the fix to the bug that let AI know where your capital was, you can't assume the mithril will be negated. So now instead of a wraithform ghoul, you have a stack of 9 AI mithril swordsmen. Without summons, how does the human counter that?

However, you might be saying that 400 is too high, but 300 is OK. Then we're at least at a level of agreement.

But back to the retort. +1 to hit is better than warlords offense, and it saves you 150 production (currently). And gold-mana is worth more than half what channeler does until late game when all your combats are at x3 distance.

So its around half of 2 different 2 pick retorts. Warlord is less offense than defense, and if the AG is cheap then Alchemy is only worth like 1.5 picks. But its arguable as to whether its 1 or 2 picks even right now. If you make AG more expensive at all, achemy moves to definitely worth being 2 picks.

On the other hand if you don't change AG price, then you cant reasonably increase the price of fighter guild much or everyone will just run around with magic weapon or mithril swordsmen and skip fighters guild.



Totally forgot to mention. I'm fine with shrine being 120. There are things that can boost its output, and you never think of it as unrest reduction vs Garrison reduction. Its unrest reduction beyond garrison, so it doesn't remotely need to follow the same cost pattern. Unless you want shaman, which are a military unit, so the higher price is justified in this change anyway.

[Tlavoc]

I thought the alchemy retort just gave magic weapons, but didn't use minerals like the alchemist guild did? Or am I just misunderstanding something? (of course, on top of a 1 to 1 exchange rate between gold and mana)

As in, if you have the alchemist retort and have adamantium next to your starting town, you still have to build an alchemist guild to actually have adamantium weapons - until you build the guild your units will only have magic weapons, +1 to hit, not adamantium; +1 to hit +2 melee/ranged/defense.