Nargott
@Dredknight
One could argue that Destruction magic is the most problematic, but I see it as a fundamental school that other schools should be built upon.
I think that buffs/debuffs should heavily depend on spell power, as well as spells from the Destruction school. That is, the effect of buffs should primarily scale with spell power, while duration is less important. This opens up the possibility of using these schools of magic most effectively with magical heroes, thereby compensating for the lag in passive stats.
Summoning is effective in a prolonged battle, so each effect it creates should ideally be long-lasting. A wall should not be destroyed in one hit (more durable, but with less damage), a phantom should not be destroyed in one hit (I plan to make incorporeality 1/3, that is, every third hit lands, while the first and second are guaranteed misses), a hive should not be destroyed in one attack, and so on.
In addition, for Destruction spells, spell power is the key parameter, and for summoning, knowledge should be the key parameter, because this school is focused on prolonged combat.
This also means that summoning spells can be made more powerful and expensive than Chaos spells, specifically to enhance the role of knowledge and mana.
The Arcane Crystal plays more of a role as an area-of-effect damage spell rather than an obstacle, and with a sufficiently high damage output, summoning has no problems with the relevance of this spell. It has been thoroughly tested (even in early versions of the mod) that with very high damage, the crystal becomes overpowered, to the point that it is spammed every turn and wins. The same can be said about mines - they are more situational (the enemy needs to move), but high damage can also compensate for this. Even better, if not so much with damage, but with the number of mines themselves - the more there are, the more reliably they work.
The problem with Destruction magic is that it is optimal at a certain point in time - if the armies are too small, it wins, but if they are too large, it is weak. Both effects can be mitigated:
1) If spells are made with a smaller multiplier to SP and a larger constant value, then the fluctuations in power will be lower - this is especially true for the most powerful spells. I think the ideal option is when the multiplier does not scale with the skill at all, but the constant value does - this allows you to play Chaos even with low SP, if you invest in the skill.
2) This effect is mitigated by strong abilities with additional effects (freezing, stunning, reducing defense, etc.). But an even more effective way is to motivate players not to gather a huge army with one hero. If a powerful army is less effective than many smaller armies, then this will work. In addition, the player should have several vulnerabilities - cities that cannot be defended by one army. Under these conditions, the problem of huge armies is solved, and therefore, Chaos will work better even in a prolonged game.
It turns out that huge armies are the key problem for Destruction magic. Because it can be balanced for a specific point in time, and then it inevitably weakens if the armies continue to grow significantly. Another means of curbing army growth in the late game is active exchanges, when players have several battles with each other, and troops are consumed rather than accumulated.
Added 15 minutes ago
If we remain within the model of "one large army accumulates", then a possible solution here is to introduce abilities/artifacts that enhance damage, depending on the hero's level. And/or to incorporate the hero's level into the spell formula itself, if possible. How sufficient this measure will be depends on the specific numbers and the number of such abilities.
Added 7 minutes ago
Regarding:
You just defend with your troops and spam spells.
A good defense against spell spam is mana scarcity - that is, powerful spells should be very expensive. In this case, it will not be possible to cast them throughout the battle, and you will have to switch to cheaper and weaker ones. Powerful Chaos spells, even if very expensive, are also effective against larger armies - when the player still has the opportunity to spend a lot of mana, but inflict significant damage, despite the size of the troops.
If the army is in defense, and the player has enough tempo and mana to defeat the enemy only with magic, this means that either he has too much mana, or the physical damage is so weak that the troops themselves cannot inflict enough damage. In the 20/80 model (20% damage from troops, 80% damage from the hero), the player has no motivation to leave defense. In order for the troops to play a role and not just stand in a defensive position, the damage from the hero and the damage from the army should be comparable. In addition, if the hero has a mana shortage (too expensive top-tier spells), then the player will have less incentive to prolong the battle with passive defense and more incentive to engage in an exchange, trying to inflict additional damage with troops.