Armor as a bonus to saves instead of being its own thing

[Greywander]

Something I've looked at previously was turning saving throws into static defenses. So when you throw a Fireball, instead of each target rolling a Dexterity save against your spell save DC, you would make a spell attack roll against their Dexterity save DC. The math comes out the same, the only thing that's different is how it's presented.

Then I had the idea that we could make things more interesting by having you roll a single attack against multiple defenses. For example, let's say that armor can help block a Fireball. So you'd roll one spell attack and compare it to both the target's AC and their DEX save DC. If you beat both, you do full damage. If you only beat one, you do half damage. And if you don't beat either, you do no damage.

As you can see, this allows for degrees of success, but this can also be used in another way: giving a different effect depending on which defenses you beat. This in turn has an interesting effect since we're only rolling one attack roll against multiple defenses, where the range of possible effects changes depending in which defenses are higher on the target. Consider, for example, a spell that hurls an ice spike that will pin the target in place if you beat their DEX save and deal damage if you beat their CON save. Against any target there are three possible outcomes. One is that you fail to beat either save, and the spell has no effect. Another is that you beat both saves, pinning the target and dealing damage. But the in-between effect, where you beat one save and not the other, will depend on which save is higher. If a target has a higher DEX save, it's not possible to beat their DEX save without also beating their CON save. Thus the only third option is that you deal damage, but don't pin them in place. On the other hand, against a target with a higher CON save, the third option is that you pin them, but don't deal damage.

That was about as far as I got the last time I brought this topic up. I was thinking about this again, and one thought I had was if perhaps we should have other defenses more similar to AC than to saves. AC is kind of a special "armor + DEX", so maybe other types of defense that use other ability scores could exist. But then I thought, what if we go the other way? Eliminate AC altogether and instead have some saves gain an armor bonus against certain attacks. For example, a weapon attack is now against the target's DEX save, and adds the target's armor bonus to the save. Ray of Frost would target CON with the armor bonus added. And so on.

This could then lead to martials getting more kinds of attacks that target different saves. For example, if you grapple a target, you could then put them in a joint lock, making an attack against their STR save. Or if you're dual-wielding, you could do a feint attack that targets their WIS save. The tricky part is defining when to add the armor bonus or not. There's also the matter that adding an armor bonus on top of the ability score modifier and proficiency will make it much harder to beat, so perhaps the armor bonus should be instead of proficiency. Things that use armor would then not care about proficiency, and Unarmored Defense would simply allow you to use your proficiency bonus instead of an armor bonus. Alternatively, just like armor gives a bonus, maybe weapons or spells could have their own bonus. So, for example, an "armed" melee weapon attack would add both the weapon's attack bonus to the attack roll and the target's armor bonus to their save DC, while an "unarmed" attack would add neither the weapon bonus nor the armor bonus. (There's probably better names for these, especially since we don't want to mix them up with unarmed strikes, which would probably still factor in the target's armor.)

I suppose I should say that at this point we're deep into homebrew territory. This isn't something you could patch into 5e, not easily anyway. But if you're doing a big overhaul and rewriting a bunch of content from scratch, then it wouldn't be that much more effort to fit something like this in.

Thoughts on this?

[Amechra]

What if armor mostly just determined which save you got to use?

Something like...



Armor Type
Dex Save
Con Save
Other?


Unarmored/Light
Yes

Yes (disadvantage)
--


Medium
Yes
Yes
--


Heavy
Yes (disadvantage)
Yes
Strength


Shields?
--
--
Strength


Mage Armor

--
--
Casting Stat


Barbarian Unarmored
--
Yes
--


Monk Unarmored
--
--
Wisdom



This isn't perfect, by any stretch of the imagination (for one thing, it makes non-magical light armor into nothing more than flavortext), but I hope that the idea is reasonably clear — you get to pick to make a Dex or a Con save when you get attacked, with some features and equipment letting you change which stats you get to use. The "nice" thing about this is that it also works well with the "being incapacitated means you auto-fail Dex and Str saves" rules that 5e has.

[Greywander]

What if armor mostly just determined which save you got to use?
I feel like I must be misunderstanding this. If you can choose which save to use, why wouldn't you just always pick your highest save? The whole point of the save system is so that all your defenses don't hinge on a single stat. And usually it's the attacker that determines which save they're targeting, though I could see having a class feature or feat that allows you to swap one save with a different one a limited number of times or under very specific circumstances.

Now, one thing I was thinking about is if different armor might work better under different circumstances. For example, let's say you get wrapped up by a giant constrictor snake. Snek squeezes you, which would probably attack your STR save. Rigid armor like plate would help you resist, while flexible armor like mail would do pretty much nothing. The question is, how are we defining when different types of armor apply? How many different types of armor do we even have? It's not just a question of flexible vs. rigid, either; we could also see things like mystical talismans that act like armor against certain magical effects, but not against physical attacks. Heck, iron armor might add to your saves against fey charms. This... this is actually a pretty interesting idea, lots of hidden potential.

I think perhaps the best way to set this up is to create a list of different types of attacks, not unlike the list of creature types or damage types. Each suit of armor is then marked with which attacks it provides protection against, and otherwise it gives no protection. Each attack is then given one attack type, which is determined separately from the save that is targeted. So taking the feint as an example, it would attack the target's WIS save, but would still count as a physical strike, so regular armor would apply even though it's attacking a mental save.

If armor is simply added on top of proficiency (if applicable), then it could make a particular save almost untouchable. For example, a knight in full plate might be almost impervious to weapon attacks. To defeat them, you'd either have to target a save they're not proficient with, target them with an attack that bypasses their armor, or preferably both.

As for a list of different attack types, there's potentially an infinite number we could do, especially if supernatural attacks are included. That said, no reason why a magical attack couldn't be treated the same as a weapon strike. If you shoot an ice spike at someone, is that much different from shooting an arrow at them? Maybe some attacks might have multiple types, and either the best armor would apply, or all applicable armors would apply, either of which would make multi-type attacks worse than single-type attacks. What about something like this to get started?

Strike. A direct blow with a weapon or physical object.
Crush. A slow, steady pressure that crushes the target or rips them to pieces.
Splash. A spray of poison, acid, fire, cold, or some other chemical, liquid, or gaseous effect.
Divine (alignment). Holy, or unholy, power that punishes the target. Wards that protect against divine judgement are keyed to ward off specific alignments.
Supernatural (creature type). A supernatural ability of a particular sorcerous bloodline. Each bloodline (creature type) has its own method of warding.
Arcane (spell school). A spell learned through one of the schools of magic. Different arcane barriers are used to ward off specific schools.
Special. This attack is uncommon enough to not fit into a specific category, and will specify what conditions under which it can be protected from.

I'm sure there's a lot that's not covered by this. For example, any kind of thunder damage effect, though those could fall under one of the magical attacks. Same with psionic effects (though could easily be marked as supernatural (aberration), though, or even supernatural (psion)). Explosions would probably be splash attacks. I guess thunder attacks could also be splash, but that doesn't feel quite right. Given the number of damage types and creature types, we should probably have 8 to 12 different attack types, so we have some room to expand. Do we want separate attacks for different kinds of weapons? E.g. thrust vs. hack vs. cut vs. smash? We could then get more specific with which types of armor work best against which types of attacks. This might even lead to consolidating bludgeoning, piercing, and slashing into a single damage type, with the attack type determining the difference between them, but let's not get ahead of ourselves.

[Amechra]

I feel like I must be misunderstanding this. If you can choose which save to use, why wouldn't you just always pick your highest save?

Because your best save might not be available — 5e already does a little bit of this, where being incapacitated means that you automatically fail Dexterity and Strength saves. Presumably you'd increase the number of situations where stuff like that happened if you were going down the "armor determines which saves you can use" route.

As for your suggestion of defining seven different attack types and giving different armors different bonuses against different attack types... people have been trying systems like that (mostly in the form of weapons vs. armor tables) since more-or-less the beginning of the hobby, and the marginal benefits of using them have always been swamped by how much of a hassle they are in actual play, unless they're hyper restricted (like, say, having armor interact differently with guns). The only times I've seen that kind of thing work have been in video games, where the computer can automatically apply all necessary modifiers. I mean, I guess it could work if you were only going to play the game with a VTT?

...

Honestly, your biggest issue with hacking 5e to make all six ability scores equally important is that you'd have to essentially re-do ability scores from scratch, because D&D's ability score system is designed around having dump stats. The best solution I'd seen to that within the D&D framework is 13th Age's way of handling defenses — your AC is based off the middle value of your Con/Dex/Wis, your PD is based off of the middle value of your Str/Dex/Con, and your MD is based off of the middle value of your Int/Wis/Cha. In other words, someone with Dex 16, Con 14, and Wis 12 would base their AC off of their Constitution, and that would only improve if they increased both their Dexterity and their Constitution.

[MrStabby]

I don't think any of this is bad for a new system, and I don't but something being hard as a reason for not trying it...



I see think things like:
grapple as a strength attack against a strength AC
Swinging a sword as a strength attack against a Dex AC
A poison could bomb as a dex attack against a Con AC
Hypnosis as a Cha* attack vs a Wis AC
Fireball as an int* attack vs a dex AC

*I am playing with a system that moves the spellcasting ability modifier from the choice of class to the spell so its harder to be good at every spell. It works well. Not pushing you to do the same but I thought fleshing out the table would show that you could do it.



One thing I like about it is it might give a bit more room for complexity around bonuses. One think I leaned from D&D 3 was that all those situational bonuses were a pain. A +2 here and there, a modifier... It slowed the game down. I think there is a differnce between complexity that happens in sessions and complexity that happens between sessions though, and this could lean to the latter. If you have a more complex formula for calculating say Charisma AC that can have odd little bonuses here and there then it doesn't bother me and as an AC can be pre-calculated and added to a character sheet it needn't slow down sessions.

I think there are certainly a lot of blanace implications - there exist spells that (thematically at least) could target the AC of any stat and at the moment they are mostly saves ans somewhat managed by magic resistance and legendary resistance. Certainly someting worth playing around with I think.

[Greywander]

Because your best save might not be available — 5e already does a little bit of this, where being incapacitated means that you automatically fail Dexterity and Strength saves. Presumably you'd increase the number of situations where stuff like that happened if you were going down the "armor determines which saves you can use" route.
Wouldn't this just lead to stacking everything into one save, making you nearly invincible right up until you fail a save, then suddenly you're powerless? Except not every save has an auto-fail condition. Imagine a wizard stacking everything into their INT save with Mage Armor; now nothing can touch them unless it gets past their maxed INT save, and no conditions make INT saves auto-fail. We could add conditions for each save to make them fail, but then combat would degenerate into both sides taking turns rolling dice until one side gets lucky, and then they just win. That doesn't sound very fun.


As for your suggestion of defining seven different attack types and giving different armors different bonuses against different attack types... people have been trying systems like that (mostly in the form of weapons vs. armor tables) since more-or-less the beginning of the hobby, and the marginal benefits of using them have always been swamped by how much of a hassle they are in actual play, unless they're hyper restricted (like, say, having armor interact differently with guns). The only times I've seen that kind of thing work have been in video games, where the computer can automatically apply all necessary modifiers. I mean, I guess it could work if you were only going to play the game with a VTT?
This is definitely something to be cautious of, but the system doesn't need to be super complex. For example, a given suit of armor or mystical ward could have a single bonus, then define which types of attacks the bonus applies to. That isn't any more complex than a bonus that gives advantage on specific types of rolls, which is something that's relatively common in 5e.


Honestly, your biggest issue with hacking 5e to make all six ability scores equally important is that you'd have to essentially re-do ability scores from scratch, because D&D's ability score system is designed around having dump stats. The best solution I'd seen to that within the D&D framework is 13th Age's way of handling defenses — your AC is based off the middle value of your Con/Dex/Wis, your PD is based off of the middle value of your Str/Dex/Con, and your MD is based off of the middle value of your Int/Wis/Cha. In other words, someone with Dex 16, Con 14, and Wis 12 would base their AC off of their Constitution, and that would only improve if they increased both their Dexterity and their Constitution.
This is a good point, and to be honest I'm not really happy with how ability scores are handled in 5e. It just seems to push players towards cookie cutter ability score spreads, where they max their primary stat and (on MAD classes) their secondary stat, leave CON at 14 or 16 and dump everything else. It's boring and makes ability score distribution kind of meaningless since there's only a few "right" spreads.

So this is something else I'll be putting some thought into on how to improve it. One thing I'm considering is making every ability score useful for every character, so even within the same class you can fine tune how your character is specialized through their ability scores. That doesn't really solve the issue of defenses, though, especially on characters choosing to minimax their stats anyway.


One thing I like about it is it might give a bit more room for complexity around bonuses. One think I leaned from D&D 3 was that all those situational bonuses were a pain. A +2 here and there, a modifier... It slowed the game down. I think there is a differnce between complexity that happens in sessions and complexity that happens between sessions though, and this could lean to the latter. If you have a more complex formula for calculating say Charisma AC that can have odd little bonuses here and there then it doesn't bother me and as an AC can be pre-calculated and added to a character sheet it needn't slow down sessions.
This definitely points towards just adding the best relevant armor, instead of all applicable armor bonuses. Keep it simple and straightforward. 5e's thing of binary bonuses that you either add or don't add, and multiple bonuses don't stack, definitely helps deal with the issue of slowing the game down.


I think there are certainly a lot of blanace implications - there exist spells that (thematically at least) could target the AC of any stat and at the moment they are mostly saves ans somewhat managed by magic resistance and legendary resistance. Certainly someting worth playing around with I think.
In vanilla, AC was pretty much the general defense stat, making it easy to have a decent defense against most things, with saves being more niche. With this change, there isn't really a general defense stat anymore, though maybe DEX saves would mostly fill that role. This definitely changes things, and I'll have to see where this goes and what issues might arise.