Test games played

[Nelphine]

Yeah the 'too much damage overall' is why I wanted to change wave damage so that it isn't a constant increase s you get less targets. Make 1-2 targets artificially lower so that it peaks in the 3-5 range. (At worst, just make it do the same damage per target against 1-2 targets as it does against 3 targets, although ideally it would actually be lower). That way, your 3-9 damage, which IS well balanced, doesn't change.


And yes, the fact the human can just focus a doomstack of very rares means they win (except chaos). That isn't restricted to Demon lords. Put 9 archangels against the AI, or 9 sky drakes, or 9 colossus, and you win. And just like demon lords, you probably take 0 damage doing so (even if demon lords lose doom bolt, they still don't take any damage, even with the extra round) although unlike death you might need to buff your very rares (invisibility and iron skin), but that isn't really an issue since demon lords at so hugely expensive to start. And AI demon lords still need the doom bolt because of lairs and the AI using only 1-3 demon lords in a doomstack.

The problem isn't doom bolt, or demon lords, or slingers. The problem is that the human can pick units that do high damage AND target ai units that can't survive it AND the AI has no idea when units become obsolete. It's the same thing as me saying the AI should stop using magicians when I've researched sky drakes; the human understands when units are obsolete, and the game design is such that units DO become obsolete - but the AI has no idea.

So unless you a) teach the AI how to recognize game changers or b) make the variance between units FAR lower so that no units do become obsolete, these discussions will crop up every time you happen to encounter this state where some units are obsolete.

[Sapher]

Uploaded part 3. August 1406, Myrran is conquered
https://www.youtube.com/watch?v=xZiylZVXcPQ

[Seravy]

a) teach the AI how to recognize game changers or b) make the variance between units FAR lower

a. is impossible, b. would make it a boring game I don't want to play.

But Demon Lords aren't about this - they deal enough damage to first turn kill ANYTHING. (except a doomstack of 6+ high health very rares or a full 9 rares.)

We want to give the AI at least 3 turns in these battles so they can kill one or two units with spells and ranged attacks each battle.

I mean, "obsolete" might be the correct term, but if every unit you can produce is obsolete, producing something else isn't going to help at all. There are no city troops that can survive a demon lord stack for a turn, except maybe War Mammoths.

[Seravy]

Here is an idea :
We can make Wave of Despair cost more if it's not a "nerf", but a feature.
"Wave of Despair costs 72 to cast, and this cost is unaffected by cost modifiers".
At first sight this just makes it more expensive but...throw in Divine Orders and suddenly it looks way better. Throw in Evil Omens and it's an amazing bargain. Throw in Specialist or 10 books and you suddenly want to use other spells instead as those benefit from your cost reductions. So the spell becomes more expensive but still have good uses, and most importantly, those good uses overlap the intended role of the spell (counter heroes or unbeatable buffed very rares = likely Divine Order in play).
A cost of 72, with no cost modifiers is equivalent to a cost of 90 reduced by Specialist and 10 books. So for all sorts of cost effectiveness comparisons, it works as if the spell costed 90, even though it does not.
At this cost, using the spell 4 times is still viable - but only if you haven't used any other spell. You absolutely cannot do something like using black prayer+massacre first, and then use 4 waves to kill the remaining 3 great wyrms. That would require like 350 skill total which you definitely won't have at the time very rares start to appear. (You might have it later though, but you either managed to deal with them already or lost the game, so that's not a problem.)

[Seravy]

As for Demon Lords, they are fine as is EXCEPT the Doom Bolt. Yes, it's amazing and makes the unit feel more awesome and powerful, but...it's a chaos spell. A high end Chaos spell. On a Death creature. Doesn't really makes sense or belongs there at all. I would rather have a different spell there, unfortunately we can't really make a new spell ability, but let's play with the thought. Demons already cast all the spells you can have in the realm so either they get a very rare, or there is no point. Demon Lords with free Massacre would clearly be unreasonable. Free Annihilate...isn't bad. It does mean they'll still do too much front loaded damage, but at least, it'll be limited to only work on non-very rares. Finally, Animate dead. Now that is something we can't use to do front loaded damage, but fits the theme of "very durable, long term combat potential" perfectly. Unfortunately, it also defeats the purpose of this change, allowing the AI more turns to damage the stack - you can reanimate any lost demon lords. Sure it is limited to once per unit but still.

So I think the best move might be to just get rid of that Doom Bolt entirely and add no replacement. It's STILL a unit with a ranged attack on par with colossus, twice the ammo, three summons, better immunities, flight, power production, and life steal. (Although the life steal is already quite nerfed, it still allows the demon lord to heal by targeting weak units. Even if they "only" heal about 10-20 total health from the 8 shots, that's still huge.)
So even without Doom Bolt the Demon Lord is already the best fantastic creature.

...but that's for the human player. The AI probably doesn't really have much use for Doom Boltless Demon Lords. Sure, it still is a flying healing colossus, that summons demons, but heroes can cut that down easily. Or not? We've added Supernatural since then...

Now, if we really want to insist on Doom to counter heroes and high defense/unbeatable targets, which Death does have a hard time fighting against, we can just make it deal doom damage by default. Without changing the stats, that would be 10 ranged, or 12 melee damage, fairly balanced? A bit overlapping with Death Knights though, which piece armor. (I'm unsure if we want that though - high armor is already weak against many realms and units. It's probably better to leave them without Doom, if absolutely necessary, giving them high attack strength.)

[Nelphine]

Quick casting a doom bolt and a ranged attack is the only thing that makes demon lords lhreats when used by the AI. It's not the doom damage - it's the raw output, because demon lords are most often found by themself with garbage units supporting them, like werewolves or wraiths, and your ranged city units destroy them. They need to kill multiple units in the first turn to be a threat.

Doom bolt is better than annihilate specifically because it doesn't kill very rares or rares - but it does hurt them. The ai needs this to be able to matter.

Change the game, us a stack of any very rares. The human wins. Take away doombolt don't give demon lords anything to replace it.. and the human still wins without losses. Humans with very rares is the problem. Doombolt is NOT the problem.

So if removing diombolt entirely won't fix a stack of Demon lords in the hands of a human, then we look at what else that will do. It will make neutral demon lords too weak. It will make ai doomstscks with 1 demon Lord too weak.

The aI NEEDS to be able to do consistent damage to units who are immune to resist affects, AND have the option of killing 2 weak units on the first turn. That's what makes demon lords s challenge in a lair, whereas great drakes and sky drakes are pushovers that we ignore.

We want all very rares like demon lords and djinn - we don't want any like hydras and great drakes.

[Seravy]

Change the game, us a stack of any very rares. The human wins.

So what? Ban the human from using very rares? :D

But yes, you are right. I keep forgetting the human is supposed to be able to win battles without losses. (regeneration is for that, too) The AI wins through attacking many places simultaneously, so even if such a doomstack exists, the human goes down first because he is losing at a faster rate than winning.

Although Demon Lords in particular, even if you are down to your fortress only, maintain themselves, so you can still defeat everyone and win the game. In theory. But in practice, you probably still suffer losses anyway, as the enemy gets the first turn and will have enough ranged units.

...so no change to Demon Lord then.

That leaves Spell Ward. Slightly reducing the penalty on summoned units is probably a step in the right direction but we still have to deal with the "what if the human did not get their top very rare creature?" problem, as the weaker one usually won't do the job. Although the heroes are supposed to be the answer for that, and yes, that is why we have heroes who can cast Flight on themselves or fly naturally. I was really dumb for not remembering that, unfortunately a 44 hour long game against Sorcery is frustrating enough to forget how to play. So maybe we are good then? With Heroic Heart, using such a hero is even more viable...although the city might have units in them the hero can't possibly take (multiple storm giants with true sight? demon lords? No invisible hero can fight that...) so we likely still need to do something else.

[Nelphine]

Agreed. Just because heroes can do something, I don't think we should base game balance around that. Heroes realistically break the rules and let you do awesome stuff, but if the rule are based on them, they don't feel as special anymore.

So, maybe look at what spot spell ward fills for sorcery? Start from basics, like we've done in so many other cases? What does it really actually need to do?

[Seravy]

It fills the role of the ultimate city defense spell, pretty much. It ensures your garrison can't be killed by combat spells, and your city can't be destroyed by curses. That leaves armies as the only thing the cities are still weak against, but since armies are no match for spells (which is why you need this to counter the enemy spells) and you do get spells to fight theirs, the city is pretty much invincible.
On a more technical side, it counters the two spell types you can't otherwise counter, but need to to keep "control" of the game - spells that damage your city and garrison but can't be dispelled (Call the Void, Earthquake, Blizzard, Fire Storm), and combat spells (Flame Strike, Massacre, Magic Vortex). It also counters doomstacks which are another thing you can't otherwise stop from ruining your city but since these are less commonly available to players, this is a weaker counter.

Problem is, if such a spell does do its job, it basically says "You can't lose the game", since not losing city resources ever means you can't lose the game. Now, the city curses part is likely OK, but the combat part is problematic. Combat spells is a binary thing - if the enemy manages to get even one of those high end spells through, you are going to suffer very significant losses and possibly lose the whole city (it only takes one Flame Strike to wipe out most of a garrison). If not, then we get what we have now.

The problem might even be more fundamental : Sorcery counters buffing and counters summoned units. But unbuffed normal units are not powerful enough to fight Sorcery. So it literally is "heroes and combat spells only", unless you happen to have access to one of the 14+ resistance illusion immune fantastic units. (every realm has one except Chaos, but Demon Lord is only 13 resist, still close enough.)

[Nelphine]

Right, you've described what it IS. But I mean, go back like spell binding. What does it NEED to be? Yes sorcery is all about control - but do we NEED to control every last thing? I don't think so. Instead, decide what it actually needs to do. Particularly in comparison to other very rare spells - in the same way that spell binding may have made sense conceptually, it was brutally unbalanced. Spell ward is in the same place. What role does it actually need to play for the human to be worthwhile as a very rare ? (Countering city curses is plenty) what role does it need to play for the AI to be worthwhile as a very rare? (Countering city curses may not be enough, but that's all consecration does, and it's still on the list.)

Another example: I don't think we need it to fill the role of 'if sorcery gets to very rare, sorcery wins'. That's a role we've been actively removing from the game.