LIFE Realm

[Seravy]

I don't think it's only the armor although that is a big part of it. But healing a unit with higher attack strength is also a better effect than healing one with lower - if each heal allows the unit to survive another turn, there is a big difference if that's a turn for your Great Drake or your Stag Beetle.
While I prefer not to go that way, if we want variable healing, it should be based on unit cost, probably including the buffs and levels. Which is unreasonably complex.
One big problem about this, and using armor especially, there is armor ignoring damage in the game. Quite a lot of it.
I don't like random - unreliable spells are one of the worst thing to have in strategy games.

[Bahgtru]

Yes, but face it - you're still interested in something as old as mom because of flavour and challenge, and the latter - unless you consider the AI smart - is its randomness.

Making variability depend on armor/to defend simplifies the calculation as all the buffs are naturally included in the end result which is displayed on the unit.

[zitro1987]

If randomness is associated with the basic healing spell, could it use the offensive spell calculation and thus be a 25-strength healing attack? (where armor applies)

[MrBiscuits]

Of course healing is more effective on better units, that's the same as in any game that has health points.

As the game progresses units can deal more damage too (eg. warlocks dealing 10 doom damage), so if you scale healing by armour you would never use it in the end game.

Also scaling healing is confusing and unintuitive. 

I guess it is impossible to include a no hero option for people who want to play without them.

[Seravy]

I think healing scaling up isn't really the problem by itself. It still costs one spellcasting turn, and spells do scale up in how much they do per use as the game progresses. So as you say, Healing becoming better only means it can keep up with the game progress and remain useful.
The problem starts to surface when either turns no longer matter (AI ran out of magic, having no ranged units or ammo left, your units moving faster or theirs sitting behind a wall, etc) or when you have units capable of casting the spell.
Unlike other spells where the cost scales with the effect - a Doom Bolt deals more damage than a Fire Bolt but for 4 times the MP and skill - healing cost is constant. So when you have the chance to actually use it 4x more times than you could the equally powerful damage spell, then we have the problem.

Considering AI casting skill is reasonably high in the late game, this can be narrowed down to 3 cases :
1. AI has no mana crystals left. Happens on the first few turns of a massive endgame war when dozens of attacks are done per turn, but not really outside that situation (unless you are against a wizard you have already mostly beaten).
2. AI has no spells that can hurt your army so you can afford wasting time.
3. You have Caster units in the battle.

Out of these, 1 is not a real issue. If you can still cast spells and the AI can't, you're winning that battle anyway, it doesn't matter if it's 4 times more effective or not.
2 is again a non-issue. If the AI has no spells that can hurt your unit, you can't possibly lose that battle. It doesn't really matter if you kill the AI's army using your own spells, or let the invincible unit do it. On top of this, having this situation happen means you can make the AI spend like a thousand mana crystal on these battles without having to do anything, which by itself is a big enough effect to win. The only real way to address this is to make sure every realm has combat spells that can be relevant in battle, no matter what sort of invincible unit the human player throws at them. Currently this is true for Chaos and Nature, but Life, Sorcery and Death can easily find themselves in situations where they can't really cast anything meaningful against a very heavily armored/high defense/invisible etc unit. However Sorcery can dispel and those units are typically buffed, as well as casting Counter Magic which will at least reduce the cost effectiveness of healing, so it's mostly Death and Life who are vulnerable to this tactic.

So in the end the only real problem is units that can cast Healing on their own I think. Out of those, Magicians and Shaman are very fragile and will likely not make it out of the battle alive which makes it reasonable. So we can further narrow it down to heroes, angels maybe Saints, and archangels (and possibly the new Spellzerker unit).

[Nelphine]

I think number 2 is a much bigger problem than you describe. Specifically, it doesn't need to be no spells that hurt you, it's just, no spells that significantly hurt you.


Even something like lightning bolt against an endurance archangel (who casts invulnerability on itself while you cast prayer) isn't that good (10.8 average damage, reduced by 4 for armor -2 more for invulnerability= equals less damage than the wizard can heal with one cast of healing).


In fact warp lightning also ends up doing less damage than healing.

So I'd argue that there's a good chance chaos can also be affected. Given the weaker damage spells involved, even after dispelling endurance and invulnerability, this also means sorcery is affected.

Now I realize that's a very rare summon, but that implies that any high armor unit will have the same sort of thing - imagine a stone giant with elemental armor and healing, or even a dragon turtle with healing against non armor piercing spells. Even if those only come close to breaking even against lightning bolt and psionic blast, that potentially means all 5 realms have problems dealing with these units.

Putting it another way, you are effectively giving Regen 5 to one unit until your casting skill runs out. Regen 2 is a very rare spell, and we've recently reduced many units innate Regen specifically due to this problem. And healing can be switched to a second unit if the first one does die.

[Seravy]

Certainly but my point is, if you're in that situation, healing the unit isn't what's winning you the game. The AI losing 1000 mana crystals every turn per battle does. You could do the same with a 20 armor spearmen assuming you can somehow make one, even if that spearmen has no hope of ever actually beating any enemies in the garrison. Healing does help in making this situation happen and might be one of the greatest enablers for it but there are still plenty of cases where it happens without you ever needing to cast a single heal - and those are the cases where this strategy is the most efficient as that means your effective cost per battle is zero. If your armor is high enough you don't need any heals. If the unit has enough Regen, same. If it's invisible and the AI can't target it, same. So in this case, Healing is merely one aspect of a greater problem, not the problem by itself.

The "fair" solution would be if healing cost scaled up based on the value (armor, cost, whatever) of the unit, but that's impossible. You cast first, target later. I guess we could make the spell subtract an additional cost after being cast but that would allow you to use it one last time when you couldn't normally afford it. Still the additional cost required is impossible to properly balance - depending on what causes your unit to be hard to damage, some of those causes will not be covered.

Here is an idea, although really hard to implement, but might work for solving "3" : What if each unit was only allowed to get healed once each turn? ...but that would really undermine the playability of Caster heroes which rely on getting 2 heals to stay alive for most of the game, so probably a bad idea. Besides, the enemy might have Caster units and deal more damage per turn, too, so limiting only healing wouldn't be fair.

[Bahgtru]

Seravy you're struggling but the answer is making it random despite not liking it - no one likes a challenge until they've overcome it... If someone tells you otherwise, what's happening is that they like the foreshadowing of the satisfaction of overcoming the challenge.

If randomness is associated with the basic healing spell, could it use the offensive spell calculation and thus be a 25-strength healing attack? (where armor applies)

That's very clean, the number should reflect firebolt to add flavour to life as anti-chaos.

An uncommon version could be armor piercing, reflecting lightning bolt. Exaltation already accurately reflects doom bolt quite closely (less heal but some bonus)

[Nelphine]

That, if the uncommon version was added, would actually be a good solution. There would be a reason to go deeper into life to get better healing spells, life would get some of the same 'obsoleted' spells that other realms have with superior versions, and the higher tier spells would specifically be much better for life units, particularly life buffed units. Overall, I agree that would be a very clean solution.

And calling it random means things like ice boltbor psionic blast are random, so unless you're going to remove or change them, this is a good solution. It's still going to be close enough to average most of the time to be reliable. Random is cracks call where nothing results on a miss, and even that we don't consider actually random because it's designed for multiple casts, over multiple battles.

I'd think healing would be strength 24, and an uncommon armor piercing version would be strength 30 and cost 25.

[zitro1987]

I'd rather not add an uncommon healing spell (we already have raise dead, resurrection there).

If healing becomes a random 24-25 strength spell, would that apply to 'mass healing' or not? I'm wary of changing mass healing - I consider it perfectly balanced.