Well, Naga's Temple of Tides can train Mur'gul Reavers then again the Naga race is considered overpowered. I will consider it.
Naga were never supposed to be a melee faction. This was added mainly for the AI and the few missions you control it.
Also this unit is cursed because why training a light unit where you can have T1 abominations with ensnare ?
So your example is not valid. Having a second 0 military unit available is very broken because it become the only strategy available.
Right, I can slow down bandit hero training by using Dependency Equivalents (Tier 2 - Tower Ruins, Tier 3 - Dragon Breeding Pit).
That would only fix part of the problem, because these buildings are built faster than teching up. Furthermore they also are their own requirements themselves. I got numerous solutions possible but I'll let you cook something from your own mind since it's more enjoyable for you !
Mercenary rank upgrades increases mana pool of the mercenary if its have mana. Should I make it separate from the mercenary rank upgrades?
Also, that requires editing the vanilla creeps which I prefer not to. Requiring mercenary spellcasters caster upgrades like a vanilla race to be able to cast their spells is a no go.
It would be safe to seperate them, or even create your own mercenary variants. Also why the allies would buy overpriced mercenaries which doesn't benefit from your upgrades ? Very situational too. I mean, it has space to be improved.
I can halve the hp bonus from 300 to 150. Also good catch I forgor them.
If you really want the map to be playable at previous versions, you might make alternates where it gives per level +25HP for light units and +50HP to heavier units. Or even separate the upgrade in two (three ?) to spread the power boost across light/heavy/ground/air/melee/ranged (cross the unconcerned out). Else I would definitely replace it by a % bonus instead.
The thing is it has a cursed dilemma : either boost the light unit a lot or the heavy units.
Compare the price and the power of your units with other melee ones, taking the upgrades & abilities in account.
In the late game your units are way too much cost effective. The dice stacking would orient the player to only spam the units with the highest damage.
For the caster upgrade, since the vanilla mercenaries already have so high mana, I would increase the mana regen instead.
Else it is free real estate for mana burns and feedback.
Hidden Cache right? Since you didn't give a number, I think I will increase the gold cost from 120 to 125 and for Seedy Pub from 140g to 150g, 40l to 50l in the future update.
My numbers would be
215 30 70 for the Hidden Cache, and
225 90 60 for the Seedy Pub, with a limit of 3 upgradable somewhere.
Yep. They are really that broken. And that's only for those two. The tents are much too easy to spam. If your really want cheaper tents, reduce their HP but keep the 20 wood cost and the build time of the farm.
Well the bandits already have Spoils of War upgrade and the orcs already have that too. I refrained from using Pillage for the bandits to make it more unique.
This requires more thought, I will revisit this in the future. I feels like I delayed my new altered melee races long enough, so I will prioritize working on them for the time being. Anyway thanks again for the indepth balance report.
+5 gold per level is almost fine, but +5 wood per level is a bit broken.
I would prefer a % cost based (i.e. with the natives GetUnitGoldCost and GetUnitWoodCost, iirc).
If you really want to keep the flat per level, I would do 4 gold per level and 4 wood per level-3, so you would not spawn free wood from killing a bunch of low level units ; or keep the 5 gold but take the upkeep in account. You call. The thing is spawning lumber from thin air is very strong as it is supposed to be a main economical bottleneck for strategy.
Building pillage would be a plus, but if you see it as a redundancy it's fine too.
One last thing, the cost reduction for mercenaries is almost too strong alone, so imagine with all of the other upgrades.
I would make it timed somewhat, with charges reloading somewhere in the UI. So the player might be able to scrap some benefit with strategy instead.
Or even apply it only to the gold cost and not the lumber, with a reduced effect.
[EDIT] After more tests cheesing the game, I got more feedback
☼ Chaplains have much better mana, mana regen plus having access to inner fire while being cheaper than forest troll priest.
☼ Spoils of War actually loot from summons and illusions, which makes it even more broken
☼ Custom mercenary heroes are globally underwhelming, but it's just a matter of some stat tweaks here and there
☼ Albedo's poison cloud perma-freeze the game, I suspect at higher level and/or with many units around
☼ Gawr Gura stormclouds disapear at death, they only appear again when learned again. You might need to respawn the clouds when resurected
☼ Shaltear
► Bloodpool seems not working. You might need bribe's damage engine as it is the most stable, complete and modern.
► Berserk requires berserk research
☼ Erufu ultimate doesn't increase the number of target since it doesn't add barrage nor any kind of multi attack. It is just attack speed right now