*Sigh*

And this is why you can't start making conclusions about how magic changes warfare if you don't know how it works in the first place. A standard medieval army doesn't have 75,000 soldiers, but even one that did would never be in any "formation" where that many could possibly be hit. If you cast a cloudkill at a standard medieval formation, you cut a hole through it. If they're standing in 5' squares (which they aren't, because packed formations pack closer than that, but that's not how DnD works), you get a 40' wide slice, which means 8x the depth of the formation. How deep is it? Maybe a dozen, couple dozen at most? You kill 100 soldiers.

Sure, the rest flee because instant death makes people flee, but the maximum number of 5' squares affected means literally nothing to what actually happens. Except the rest quite possibly don't flee, because what you actually did was kill one unit, while the neighboring units are unaffected and hold, until they see a cloud coming at them (or they hold even if they do see a cloud coming at them, because this is DnD and they could very well be fighting against the end of all things, or supernaturally boosted in morale, or the DM just isn't using morale rules). Because ancient warfare does in fact involve maneuvering of separate units, not just two blobs of guys running at each other. Though even if it was that, a Cloudkill would still barely hit anything relative to its supposed maximum, because the total area it passes over is only 8 squares wide and it takes minutes to hit that entire area while moving at a speed I'm pretty sure a formation could just walk away from once they turned. In short, that is possibly the most nonsensical calculation I could expect for mass casualties with DnD spells.

But this isn't new news. If (if) AoE magic is present on the battlefield, that just means that formations and shock combat are reduced in significance or adapt, and ranged weapons and fortifications that can resist them and mobile units move up in significance (these are actually aided massively by DnD's ranged weapons, which deal full damage regardless of range, and HoB's volley of arrows rules let you even ignore armor). Unless it doesn't, because the defending army has their own casters countering and dispelling such spells, so battlefield conditions are zero-sum with only a few spells from one side getting through. Unless that doesn't, because one side has a wand to burn down. And really, how does one hold territory in a world where the DM says lol magic just killed any amount of troops anywhere? And what even is the point of holding territory when the power of food and money pales so in comparison to the power of Bob the Wizard, who can only consume so much food and money?

The ridiculous variability of DnD magic and its being completely tied to the choices and presence of specific individuals completely uproots the very concept of warfare, to the point that yeah, it might make more sense to replace the whole thing with calling out the enemy's casters and champions, if no one wants to go through the headache. Because they know that they don't have the information required to figure out how to fight an unidentified caster, despite any given caster absolutely having limits that could be exploited. But the cessation of war in the face of such "caster supremacy" would remove that outlet of demonstration of said supremacy, so people prefer to act as though people would still mass up to be killed over a change in tax collectors. Indeed, any caster who publicly proves their ability to slay whole masses of troops at once, should prevent the very formation of an army without their approval anywhere within their reach. Which is entirely on point for the sorts of stories with casters who have DnD style magic- and if we wish to speak of honor, it would likely earn fear and censure for such a caster to actually blast an army they've decided to oppose without a proper warning which would invariably cause them to back down. But then we circle back to the rarity of that sort of magic, because it is rare, by the base rules of the game, which means normal armies would still be raised to fight other normal armies when casters aren't around or getting involved, etc.

As the phrasing goes- do you really want to get an actual university minor in pre-modern warfare in order to answer the question, or would you maybe rather just assume that magic is rare and balanced enough across the land that the fantasy game can have medieval-ish troop combat with an occasional spell from the PCs or their opposites? 'Cause unless you actually define everything from the the socio-political situation that begets strategic aims and the prevalence and level of magic and technology and supply logistics and available manpower and taxes and usable income and what everyone knows and doesn't know about all of those things and other things all the way down to the individual spell selections and disposition of each of the finite casters in the regions you've drawn up. . . there's a reason that there are entire genres of detailed computer game simulations about warfare, even without magic, which are still blunt abstractions, and yet involve dozens and dozens of variables which DnD has not assigned, let alone a particular DM for a particular campaign.


Incidentally- the Philospher's Stone fixes also put a heavy crimp in the battlefield madness of DnD spells as well. If you have to be well within bowshot and the standard spell is only 10' wide, that's a doozy. I mentioned it in the War of the Lance thread for "tier 1" worries, but it also applies heavily to battlefield conditions, simply by limiting the solo caster's ability to target an entire battlefield with impunity.