AoE vs Single Target Damage - Balance Discussion

[Wavelength]

@Tai_MT Good point you make - there are some uncommon design paradigms that even a "full damage on each hit to X random enemies" skill would work in. A good example I can think of offhand would be a skill where the main function is to "open up" an enemy to further combos - in a case like this, you really do get more benefit to hitting a spread of enemies than you do by wailing on one enemy X times.

I'm not sure I see the logic in your elemental example clearly, though - an X Random Hits would be dicey against an assortment of elemental strengths/weaknesses, whereas if there was only one enemy left and that enemy was weak to the skill, and you'd be hitting them X times, that would be pretty broken.

 

[Tai_MT]

@Wavelength

If you had an assortment of enemies, let's say 8 enemies on the field, and six of them have the skill be "effective" on them while 2 are immune and 2 in the "effective" category also take 2x damage from the element of the "random hit" skill, then it becomes a "risk vs reward" issue. If your skill will hit 4 random enemies, there's about a 1/8 chance it will hit an enemy immune and a 1/8 chance it will hit an enemy that is weak to it and kill them instantly. But, there's a 1/4 chance of hitting an enemy neither weak nor immune and deal regular damage. With 4 hits, those are pretty good odds of maybe getting two or more kills a cast, but also a high risk of the skill hitting the immune members.

The player could then have the other party members kill those immune to the elemental skill in order to eliminate the risk portion of the skill usage.

But, the dev would have to be very careful in the actual troop compositions they are employing. They would also have to be very careful with the stats they give the enemies in most instances.

After all, you're trying to avoid the two extremes when using these types of skills (unless those extremes is why you have the skills in the game to begin with and you're balancing around that). You want to avoid the "Multi-Target skill kills all the enemies quickly, so there's very little need for the single target skills" as well as the "Single-Target Multi-Hit skill mows down bosses and individual difficulties quickly and easily without much thought".

Balancing around those two extremes is likely going to be tricky for any dev. So, in the long run, it just might not be worth having such skills in most cases.

That's not to say such skills can't be incredibly useful for building strategy or interesting combat dynamics. It's just that they have a lot of "moving parts" that need to be considered very carefully when using them. It's also rare that I've seen these skills executed well outside of games that let you take them as "part of a character build" (usually dungeon crawler type games and some MMOs, with maybe one or two PvP type games in there as well).

 

[Wavelength]

@Tai_MT I think your odds are off by half (should be 1/4, 1/4, 1/2, right?), but I see your point - that a diverse enemy composition can make those "X Random Hits" skills interesting, and even if whittling down the troop until it's only enemies that are Weak to the element might push it into OP territory, and least that requires an interesting and knowing tactic from the player.

Piggybacking on what you said about "instant kill", it could be especially nifty to assign these X Random Hits skills a hidden special element where instead of, say, a 1.5x or 2x Weakness to Lightning, enemies with such a weakness to Lightning would take, say, 10x damage from the "Random Hits" Lightning skill. That would introduce a big element of Reward to the Risk/Reward tradeoff even with a low number of Random Hits (like 1 or 2), and would also open up some interesting tactics where you can eliminate most other members of the troop before finishing the last one off easily with the right element. Maybe an unusually tanky enemy you just didn't want to deal with.

You want to avoid the "Multi-Target skill kills all the enemies quickly, so there's very little need for the single target skills" as well as the "Single-Target Multi-Hit skill mows down bosses and individual difficulties quickly and easily without much thought".

I think this sets up my original objection to the "X Random Targets" skills well - the issue I have is that to compensate for the randomness of the targeting, the skill must be more effective (more powerful, or at least less costly) than an actual single-target, multi-hit spell. But once you're down to a single enemy, or you're facing a boss without minions, now it is acting as a single-target, multi-hit spell, except it's much more effective than it should be. This assumes that you have an "X Random Targets" spell with an appropriate element - as you alluded to, a careful designer might be able to avoid this situation happening often by giving characters such skills but only in elements that fragile enemies tend to come in, and I guess that could be a really cool dynamic.

I have generally preferred to lean into these spells' uses against large groups of enemies, though, by only allowing each target to be hit by them once (per spellcast). Subsequent hits can still target that enemy but will do nothing. The more enemies there are, the more unpredictable the targeting becomes, but the more average hits you score against the enemy troop as well.

For general AoE skills, a solution that I really like to avoid making "trash mobs" too easy and brainless is to use an actual Area of Effect (usually circles or rows) rather than just hitting every enemy in the battle. It's fun, it's tactical, and it makes each skill/spell (and each enemy troop layout!) feel a little different than the last. Yanfly had a good Area of Effect plugin for MV, and I think that @Tsukihime is fairly far along in developing a good one for MZ.

 

[Tai_MT]

@Tai_MT I think your odds are off by half (should be 1/4, 1/4, 1/2, right?), but I see your point - that a diverse enemy composition can make those "X Random Hits" skills interesting, and even if whittling down the troop until it's only enemies that are Weak to the element might push it into OP territory, and least that requires an interesting and knowing tactic from the player.

To be honest, I kept flip-flopping on the correct ratios. Each "hit 1 random enemy" would effectively have a 1/8 chance of hitting any of the enemies. There's a 2/3 chance of hitting an enemy that doesn't outright resist the skill.

I essentially found it difficult to express the ratio I was thinking of. Yes, 2/3rds of the enemies would take damage from the skill. Yes, about 1/4th of them are weak to the skill and would probably die instantly. But, in reality, you probably have a skill that is, "targets 4 random enemies" and there can be "repeats". So, if one hit of the skill kills one of the enemies, it increases the odds of the next hit landing on an enemy that can fully resist all damage from it. At any given time, there is probably a 25% chance of landing on one of the enemies that would die in a single hit or a 25% chance of hitting an enemy completely immune. But, for any one specific enemy to be hit is a 1/8 chance. You would also have weird odds skewering where computers really don't handle RNG well anyway, and you might have 3 of the 4 hits landing on an enemy completely immune, or landing across both immune enemies.

Anyway, I just didn't want to get into the mud and weeds in trying to debate percentage chances. I also wasn't really sure whether I wanted to use an individual's chances of being hit, or just the overall chance of hitting the correct enemy. I decided on the "individual's chances" because that felt more like the perspective of the player rather than the dev. You have a skill that can hit 4 times, but each target is random and there's a group of 8. There's a 1 in 8 chance of hitting any specific enemy. Each hit is going to feel like that 1/8 chance rather than an overall 25% chance, or 50% chance.

But, maybe I'm just pulling that perception out of my rear. It tends to be the way I think about multi-target skills as a player, anyway. I'm just assuming at this point other players "feel" the way I tend to in such instances.

Piggybacking on what you said about "instant kill", it could be especially nifty to assign these X Random Hits skills a hidden special element where instead of, say, a 1.5x or 2x Weakness to Lightning, enemies with such a weakness to Lightning would take, say, 10x damage from the "Random Hits" Lightning skill. That would introduce a big element of Reward to the Risk/Reward tradeoff even with a low number of Random Hits (like 1 or 2), and would also open up some interesting tactics where you can eliminate most other members of the troop before finishing the last one off easily with the right element. Maybe an unusually tanky enemy you just didn't want to deal with.

This would definitely be something worth implementing in such skills to make them more balanced with the rest of the skills in a game. A character may get lucky and land one of these hits on the tanky enemy while missing the other 7 and it kills them outright. Or, maybe they just get really unlucky and hit everything except that target. The player could then improve their odds by eliminating some of the other enemies. It would essentially force "careful" players to wait a turn or two to whittle down enemy ranks before using such a powerful skill that relies heavily on chance. They could mitigate risk for the same reward, or take a high risk and hope to get that reward. It might even feel pretty good to blow the skill as the first action in combat and get lucky enough to wipe out the big beefy tanky enemy without having to whittle down the other forces. Especially if that tanky enemy has some "wind up" or "set up" that makes him a bigger threat as the battle goes on.

Personally, I'd probably equate the feeling on par with the "SSSSMMMMMAAAAAAASSSSSSHHHH!!!" critical hit notifications from Earthbound. At least, as a player. "Yes! I got you! Eat it! Now to mop up!"

I think this sets up my original objection to the "X Random Targets" skills well - the issue I have is that to compensate for the randomness of the targeting, the skill must be more effective (more powerful, or at least less costly) than an actual single-target, multi-hit spell. But once you're down to a single enemy, or you're facing a boss without minions, now it is acting as a single-target, multi-hit spell, except it's much more effective than it should be. This assumes that you have an "X Random Targets" spell with an appropriate element - as you alluded to, a careful designer might be able to avoid this situation happening often by giving characters such skills but only in elements that fragile enemies tend to come in, and I guess that could be a really cool dynamic.

I have generally preferred to lean into these spells' uses against large groups of enemies, though, by only allowing each target to be hit by them once (per spellcast). Subsequent hits can still target that enemy but will do nothing. The more enemies there are, the more unpredictable the targeting becomes, but the more average hits you score against the enemy troop as well.

For general AoE skills, a solution that I really like to avoid making "trash mobs" too easy and brainless is to use an actual Area of Effect (usually circles or rows) rather than just hitting every enemy in the battle. It's fun, it's tactical, and it makes each skill/spell (and each enemy troop layout!) feel a little different than the last. Yanfly had a good Area of Effect plugin for MV, and I think that @Tsukihime is fairly far along in developing a good one for MZ.

Your objection with it tends to be mine as well. It can be done, but it is difficult to do. Generally speaking, I don't think a dev should attempt to do it unless they want to put forth the effort to design the gameplay around making such skills work (and they would have to dedicate some time and effort into doing this). I don't see these types of skills pulled off well all that often because they tend to be difficult to manage and balance as devs.

I'm not saying we shouldn't ever have them in our games. But, it's difficult to have them in games where the combat isn't specifically set up to handle them and to allow players to capitalize on their unique gameplay mechanics.

Random hits for the sake of "it does something different than other skills" just doesn't work. Without a clear goal in mind for how the skill is meant to be utilized, it will easily fall flat and render a massive chunk of gameplay unbalanced (or the skill utterly useless).

 

[Wavelength]

Sounds like you weren't saying "don't hate on X Random Hits skills" so much as "don't hate on all X Random Hits skills because really good design can find a way to integrate it strategically". And after this discussion I must agree. It's quite rare that a battle system is designed specifically with these skills in mind, so I tend to think that "subsequent hits on the same target do nothing" is the best way to balance them in general (in a combat system that's not specifically designed around them). But yeah, I think I can see their value now in a system designed specifically around the ability to load 4 "random" hits onto the same single enemy.

 

[Frostorm]

How would you guys go about "Chain Lightning" type AoE spells. Basically, the kind that jumps and perhaps decreases in dmg per jump. For example's sake, let's say we have 3 Lightning type spells:

I'll create a few possibilities...

Damage as a lever:

• Lightning Bolt
  • scope: single target
  • cost: 100mp
  • dmg: 200
• Chain Lightning
  • scope: 3 targets
  • cost: 100mp
  • dmg: 200 (91+64+45 aka -30% per jump) or ~67 avg dmg per target
• Thunderstorm
  • scope: all or 5+ targets
  • cost: 100mp
  • dmg: 200 total but only 40 per target assuming 5 targets

MP Cost as a lever:

• Lightning Bolt
  • scope: single target
  • cost: 100mp
  • dmg: 200
• Chain Lightning
  • scope: 3 targets
  • cost: 200 219mp
  • dmg: 438 (200+140+98 aka -30% per jump) or 146 avg dmg per target
• Thunderstorm
  • scope: all or 5+ targets
  • cost: 500mp
  • dmg: 1000 (200 per target assuming 5 targets)

A more realistic blend of levers:

• Lightning Bolt
  • scope: single target
  • cost: 100mp
  • dmg: 200
• Chain Lightning
  • scope: 3 targets
  • cost: 150mp
  • dmg: 300 (137+96+67 aka -30% per jump) or 100 avg dmg per target
• Thunderstorm
  • scope: all or 5+ targets
  • cost: 200mp
  • dmg: 400 w/ 5 targets (80 dmg per target)

I guess what I'm trying to say w/ this example is: Which lever(s) do you emphasize more/less for AoE skills/spells?

 

[duty]

@Frostrom How do you want the Chain Lightning spell to be used in combat? Do you want it to be something the player uses infrequently to great effect or something that's used quite often?

Which encounters would this skill help solve? As in, where is this a highly effective tactical choice?

 

[Frostorm]

Which encounters would this skill help solve? As in, where is this a highly effective tactical choice?

This would only apply to game's w/ "range" but basically I'm gonna make it so Chain Lightning has a range of 3 tiles, which is the same as Lightning Bolt but that range is just for the primary target. So if you have a target within 3 tiles, but 2 more targets even further than that but within 3 tiles of the primary target, then you can potentially hit targets that you normally couldn't reach. That would be it's intended use, imo (along w/ having more chances to cause Paralysis due to hitting more targets).

Lightning Bolt: (gray = caster, green = possible targets)


Chain Lightning: (theoretical furthest reach scenario)


But it doesn't have to jump in a straight line:

@duty I hope this answers your question.
I can see how a "Chain Lightning" spell would be far less tactically useful in a default battle system though.

 

[Grunwave]

In my project, I use row attacks and PBAOE in addition to AOE and single target. I also have a couple of spells with more complex targeting.

But my goal is for the player to open their spells and have to debate which is best in the current situation. I want them to use all of their spells and be rewarded for efficient choices.

Part of that is making AOE cost more than Single Target.

Another instance from that mathematical approach: Instant Spells. They do less damage and/or cost more than their castable counterparts, but you can take more/faster actions.

 

[lianderson]

(stabs nearby car mechanic)

Very rarely do encounters go beyond three, so this one uses half damage! Good day humans.

 

[duty]

@Frostorm if you're using a location based system, then the advantage of chain lightning is obvious. You get to hit more foes from further away.

It seems like a low cost, lower damage version of the spell is the way to go.

The ideal scenario for chain lightning seems to be a series of enemies that queue up in a not-so-straight line to attack you. The spell becomes a method to soften up multiple targets as they approach, so it makes sense the player should be able to cast it often.

Lets say there are three enemies approaching the player: A, B, and C. A is the closest, B is the second closest, and C is the furthest.

The player has enough time to cast chain lightning once before enemy A has closed to melee range, twice before enemy B closes to melee range, and thrice before enemy C closes to melee range.

This means that, by the time enemy C has gotten close enough to be a threat, it's been struck 3 times by the chain lightning spell.

 

[TheoAllen]

I don't know about everyone else, but my experience tells me that
70%multiplier for AoE feels useless.
100% multiplier for AoE feels not bad.
120% multiplier for AoE is broken, kinda.

Strictly speaking, a single damage skill may always be better if it deals more damage. At least in my design.

Instead of area damage, I'm more inclined to mix both. It is called "splash damage".
It goes something like, "Pick a target to take 140% damages and the rest takes 60%"
It feels like single damage with a bonus of area damage to the rest.
You feel satisfied that one particular target took a lot of damages.
You feel satisfied that the others also took a slight portion of damages.

I don't like "x random target" skills. It does not feel rewarding for area damage and only feels rewarding when all hit the same target. So I never made such a skill.

 

[Frostorm]

Instead of area damage, I'm more inclined to mix both. It is called "splash damage".
It goes something like, "Pick a target to take 140% damages and the rest takes 60%"
It feels like single damage with a bonus of area damage to the rest.
You feel satisfied that one particular target took a lot of damages.
You feel satisfied that the others also took a slight portion of damages.

This is actually my favorite version of AoE spells, except I don't know how to implement it (for LTBS)...

 

[Milennin]

Generally something along the lines of single target does like 2-3 times more damage than multi-target, but there are other factors to consider, such as the MP cost and secondary effects or conditional triggers. Splash damage is also a good variant option that sits in between those two.

 

[Frostorm]

A good example I can think of offhand would be a skill where the main function is to "open up" an enemy to further combos - in a case like this, you really do get more benefit to hitting a spread of enemies than you do by wailing on one enemy X times.

Ah, great idea! Each discipline tree in my game has an "opener" move, which usually applies a state or sets up some kind of mechanic for other skills to take advantage of. These openers are currently all single target skills that deal minor damage, but I wonder if I should make them AoE...

 

[Aesica]

I try to use the following things--not all of them necessarily for a given skill or spell, of course:

• AoE skills are generally weaker. This is the captain obvious solution that I suspect just about everybody else does, too.
• Formula-based damage dispersion (damage weakens for every target over 1)
• Randomized damage dispersion. This is where a skill hits 1-3 (depending on the skill) random targets with each strike, and the spell strikes several times. Here's what I mean. Note that this particular skill would deal full damage to 1-2 targets, but with more than that, it has a chance to strike other targets, meaning its unlikely one target will take the entire brunt of the spell.
• Skill is a single target + aoe hybrid skill. Meaning it hits a selected target, then deals an aoe pulse to all targets.
• Pure aoe skills typically are less cost-efficient to use against 1-2 targets than single target skills of similar strength. So if Flare (single target) and Ultima (aoe) both have a 6x damage modifier in their formulas, Ultima might cost 120 mp while Flare only costs 45.

Edit:

I don't like "x random target" skills. It does not feel rewarding for area damage and only feels rewarding when all hit the same target. So I never made such a skill.

Yeah, I'm not a huge fan of RM's default 1-4 random targets functionality, so I made a plugin to extend it so you can deal damage to say, 2 random unique targets with each hit, x number of times. It feels a lot more rewarding against larger groups of enemies without feeling broken against just a single boss or something.

 

[Wavelength]

Ah, great idea! Each discipline tree in my game has an "opener" move, which usually applies a state or sets up some kind of mechanic for other skills to take advantage of. These openers are currently all single target skills that deal minor damage, but I wonder if I should make them AoE...

Good question. I think either can be fine, but the idea of having these "opener" moves be AoE is appealing, especially if most enemies appearing in AoE troops are the type to die in one or two hits ("trash mobs") anyway. Of course, if you find that these moves are just providing way too good value as an AoE (they're starting to overshadow your other skills), that's a reason to avoid making them AoE.

• Formula-based damage dispersion (damage weakens for every target over 1)

I've always found this to be quite an effective way to avoid making a character "too good" at handling large groups of enemies, though it can also be a little unsatisfying because you never really get to see the big number hit a large group of enemies and feel like you just made the battle way easier on yourself.

Getting the numbers is always a bit tricky because if you add too much of a drop-off, it actually becomes a very poor tool against troops of 5-8 enemies! I think if I were to make a spell like this, I'd probably do something like calculate total damage as 100% + 20% for each additional enemy, and then divide that total damage by the number of enemies to determine how much to deal to each one. For anyone who has made a spell with this design (AoE spell with falloff damage for each enemy), how did you calculate the damage it should deal?

Another take on this would be to add some sort of conditionality to your AoE so that you can't just spam it to win every "mob" battle. For example, the AoE hits for single-target level of damage, but only affects enemies who are at 60% or less HP.

I also like your Single + AoE pulse suggestion a lot @Aesica.

Yeah, I'm not a huge fan of RM's default 1-4 random targets functionality, so I made a plugin to extend it so you can deal damage to say, 2 random unique targets with each hit, x number of times. It feels a lot more rewarding against larger groups of enemies without feeling broken against just a single boss or something.

Always nice to see a kindred spirit in the disdain for that functionality! I was looking at my old Ace script for dealing with this (it basically ignores additional hits on the same enemy when using an "X Random Targets" skill, and also lets you select which enemy will take the first hit if the designer wants it that way), and I forgot what I had written in the intro:
"Hit me once, shame on you. Hit me four times with a '4 Random Enemies' spell... still, shame on you!" - Succubus

Think I'll start remaking this one for MZ soon. I'm so tired of seeing these supposedly-AoE skills that are actually much better off used on a single target!

 

[Frostorm]

I just had an idea! But damn, if only I could get SRD's SkillExtender plugin to work w/ my current setup...I'd be able to create a system where selecting a spell brings up a 2nd list of choices (this would be where SRD's plugin does its thing). This list of choices would be to select the size of AoE you want your spell to be, ranging from single-target to a 3-tile AoE or even a 5-tile AoE. Each size would have appropriate resource costs of course. They could even have different dmg coefficients since they would actually be different spells database-wise. The base effect of the spells would be the same though. In other words, it would seem like a single spell from the player's PoV, albeit w/ choices for different AoE sizes.

Thoughts? Would this simplify things (reduce to potentially reducing the total number of spells in the game)? Or would it overcomplicate combat instead?

Edit: This wouldn't have to apply to all skills/spells ofc.

 

[TheoAllen]

For anyone who has made a spell with this design (AoE spell with falloff damage for each enemy), how did you calculate the damage it should deal?

I'm making it simple.
1 enemy = full damage
2 enemies = falloff damage (per enemy)
x enemies = capped at a certain value

For example, base damage is 150%
The reduction is 30% per enemy
2 enemies = 120%
3 enemies = 90%
Minimum cap at 75%
So, if there are 5 enemies, you will still deal 75% to each of them.

 

[Wavelength]

I just had an idea! But damn, if only I could get SRD's SkillExtender plugin to work w/ my current setup...I'd be able to create a system where selecting a spell brings up a 2nd list of choices (this would be where SRD's plugin does its thing). This list of choices would be to select the size of AoE you want your spell to be, ranging from single-target to a 3-tile AoE or even a 5-tile AoE. Each size would have appropriate resource costs of course. They could even have different dmg coefficients since they would actually be different spells database-wise. The base effect of the spells would be the same though. In other words, it would seem like a single spell from the player's PoV, albeit w/ choices for different AoE sizes.

Thoughts? Would this simplify things (reduce to potentially reducing the total number of spells in the game)? Or would it overcomplicate combat instead?

Edit: This wouldn't have to apply to all skills/spells ofc.

This is going to sound weird because in a lot of ways what you're proposing would be good (and I'm generally a fan of allowing a lot of player freedom), but I've generally disliked "flexible" scopes when I've seen them in games I've played. I think what feels lousy to me is that it means there's no reward or penalty for setting up good or bad situations - you always have the right tool at your disposal at the right time. And this feels strangely unsatisfying because it takes some skill expression away.
I feel like combat is the most satisfying when characters' kits are more highly specialized (maybe one is great at taking down clusters of enemies in a small radius quickly, while another can attack large areas at once but has needs to get close to the action to do so), and you have enough tools to set up those situations if you play skillfully.

That's just my opinion, though. I don't think your idea is bad.

I'm making it simple.
1 enemy = full damage
2 enemies = falloff damage (per enemy)
x enemies = capped at a certain value

For example, base damage is 150%
The reduction is 30% per enemy
2 enemies = 120%
3 enemies = 90%
Minimum cap at 75%
So, if there are 5 enemies, you will still deal 75% to each of them.

That sounds pretty good! Linear falloff wouldn't work well if it went forever, but the "house minimum" ( ) means it will always be at least somewhat effective. I like it.

I remember Stella actually had a skill like this in your demo and I couldn't figure out how severe the falloff was.