Casting = casting Buffs / Debuffs
By the way, a very important note regarding Light-Dark buffs. Since you made the cooldown not 0.5, but 0.1 (or 0.0), the gameplay based on buffs has significantly weakened; it is necessary to think about strengthening their effects, while simultaneously increasing the prices. This is precisely so that classes with low 10-15% knowledge cannot afford too many buffs in battle, ideally 2-3 at most (if they haven't specifically leveled up their knowledge). The prices need to be increased because in the buff race, the mana cost per round has decreased by half due to the buffs being cast twice as rarely, and heroes with low knowledge gain an unfair advantage, competing with those who rely on buffs with high knowledge. With the old (0.5) cooldown, mana can run out quite quickly, but with a 0.1 cooldown, it doesn't (if you don't increase prices).
First, the general idea is that Might heroes should be able to cast blessings and curses. This is how they can make their army a little stronger. The spells' effects and duration are reduced due to a lack of spellpower. Here is an example for bless and expert Light:
Might hero with 5 spellpower: +26.65% bonus initiative for 2.5 turns.
Magic hero with 50 spellpower: +41.5% for 7 turns.
Besides the obvious difference in the bonus, the low number of rounds means that every third ATB action, the Might hero will be busy casting that blessing.
This mechanic refers to our initial idea for the Might vs. Magic class.
A hero (any class) affects the battlefield in 2 main ways - passive and active.
- Passive is in the form of skills that benefit the native army or negatively affect the enemy army during the duration of the battle, where all the benefits come from skills and stats that do not reduce the ATB.
- Active is in the form of spells and abilities that improve the native army, negatively affect the enemy army, or cause direct damage. Any of those skills require hero ATB usage and can be burst or spread over a period of time.
Might classes are more prominent in passive effects, Magic classes are more prominent in active effects, and Balanced classes thrive in between.
In a battle:
- Passive effects - they just work, no need to explain.
- Active abilities - This is the interesting part that I hope may change your vision for some of the changes. For mages, it is obvious that they have a vast range of spells to cast during the whole battle. The player is overwhelmed with choices, especially when playing with 2 or 3 schools at the same time. From the Mage perspective, the battle is interesting. But from the Might perspective... Might heroes have restricted usage of spells (dark or light), and even with the schools, the range of spells in those schools can be halved for them because some of the spells require spellpower.
So, the Active part for Might heroes is not that interesting. Basically, you cast some spells for a lesser effect, and there is not much thought put into it. But with the addition of Combat, a Might player can actually choose another path. Depending on what Combat path you take, the Might hero can adjust his battlefield strategy significantly. Yes, Combat does not offer much in the way of booster skills, but it offers tactical adjustments before (preparation) and during ("Chain strikes," "retaliation," Avenger") battle.
Not once or twice can it happen that you have a good shot against a vulnerable enemy stack but have to cancel your offense with a T5, T6, or T7 unit just because the enemy hero put a "retaliation strike" on that unit, and damn, it will hit hard. This is a concept that cannot be justified by numbers and statistics.
Also, I see that you are trying too hard to make a note, and I understand that, but keep in mind that the changes you want come with a huge investment of time.
If you open the game, get a mage hero, and visit a witch hut, it will give you 2 suggestions for skill increase, for example, Occultism and Sorcery. Do you know what happens if you visit the same witch hut with a Might hero? You will not be proposed Occultism and Sorcery. It will most certainly be Combat and Leadership. All those things are done manually for every skill.
By the way, you said that Sorcery looks really strong. Sorcery is the Magic class variation of the Might class Leadership. As you can see, a hero that has Sorcery never has Leadership.
Expert Leadership gives 50% for 0.4 initiative increase. This means that the army's average initiative is increased by 20%.
Expert Sorcery, on the other hand, increases the Hero, who only acts once, by 36%.
Once again, these are numbers that cannot be measured because questions such as "is the hero as strong as 2 unit stacks? or less? or more?" have to be answered. And the answer depends on the stage of the game, the luck factor of artifacts, and so on. Final line - it cannot be determined.
Combat and Occultism are also Might and Magic counters.
Combat increases the Might hero's army's tactical advantage.
Occultism increases the Magic hero's spell effect/damage.
Occultism tree is all in numbers; Combat has barely any numbers. How do you measure that? Although unmeasurable, they are both accepted equally.
There is no universal way to equalize heroes' primary, secondary skills.
It is like measuring man and woman... it just does not work that way.
We did it the other way around. We started with how we want it to be as an idea and then started iterating propositions. The core skill system was iterated a lot of times. There were major changes in every edition up to RC8. You can check the moddb article history. I believe the mod was put online around version RC3 or RC4.
Added 19 minutes ago
P.S.
To give some information on how skills that cannot be compared with numbers are made equal. You make the skills and then try the best way to exploit them. Then you try to find the best counter (if possible). Then the effort is calculated -
Does the same amount of effort is needed to take the exploit path then to counter it?
The good thing is that we got a lot of people to test, so blind spots are reduced as much as possible. The final result is not just my decision or Magnomagus's decision but an average of all opinions, with regards to logic, of course.
Cheers