Changing 5e saves

[Moreb Benhk]

We've been discussing tweaks to make to 5e. One idea we've bandied around is shifting saves back towards the (semi-traditional) FORT REF WILL. One benefit is that it moves past the juggle of strong/weak saves, 6 saves to track feels weirdly fiddly for a DnD edition whose design aim was simplicity. It balances out classes whose main stat happens to fall on a weak save (-cough- Int or Str cough) vs those who don't (-cough- Dex and Wis). It also rewards characters who focus on one aspect - the STR CON babarian is REALLY good at Fort saves (but probably terrible at others) - helping people feel good at what they are good at. It also helps differentiate baddies and increase the importance of tailoring your spells/abilities towards the enemies weaknesses.

4e tweaked the 3e method by tying them to the highest attribute of each associated pair. This unfortunately lead to the unintended consequences of making classes/builds relying on two stats from the same defense weaker than those that didn't - as you end up with 2 weaker defenses vs 2 stronger defenses, for no added gains. We thought that each defense could be the SUM of the stats of the pair

Eg
STR 20 +5
CON 16 +3
DEX 8 -1
INT 18 +4
WIS 12 +1
CHA 8 -1

Would have attribute bonuses to saves being:
FORT = +8 (5+3)
REF = +3 (-1+4)
WILL = 0 (1+-1)



Obviously base save DCs would have to be adjusted to make it fair to ability users, but I wasn't sure how to calculate by how much. My gut says increasing by 2. And each class would only be proficient in one save (probably connected to the 'strong' one they were already proficient with). It does increase the power of the Resilience feat a little (though who was ever taking it to boost a weak save anyway), and possibly decrease the power of the Monk ability to be proficient in all saves, or anything similar, though I'm not certain. What other downstream impacts am I not accounting for? Has anyone else tried it?

[Arkhios]

We've been discussing tweaks to make to 5e. One idea we've bandied around is shifting saves back towards the (semi-traditional) FORT REF WILL. One benefit is that it moves past the juggle of strong/weak saves, 6 saves to track feels weirdly fiddly for a DnD edition whose design aim was simplicity. It balances out classes whose main stat happens to fall on a weak save (-cough- Int or Str cough) vs those who don't (-cough- Dex and Wis). It also rewards characters who focus on one aspect - the STR CON babarian is REALLY good at Fort saves (but probably terrible at others) - helping people feel good at what they are good at. It also helps differentiate baddies and increase the importance of tailoring your spells/abilities towards the enemies weaknesses.

4e tweaked the 3e method by tying them to the highest attribute of each associated pair. This unfortunately lead to the unintended consequences of making classes/builds relying on two stats from the same defense weaker than those that didn't - as you end up with 2 weaker defenses vs 2 stronger defenses, for no added gains. We thought that each defense could be the SUM of the stats of the pair

Eg
STR 20 +5
CON 16 +3
DEX 8 -1
INT 18 +4
WIS 12 +1
CHA 8 -1

Would have attribute bonuses to saves being:
FORT = +8 (5+3)
REF = +3 (-1+4)
WILL = 0 (1+-1)



Obviously base save DCs would have to be adjusted to make it fair to ability users, but I wasn't sure how to calculate by how much. My gut says increasing by 2. And each class would only be proficient in one save (probably connected to the 'strong' one they were already proficient with). It does increase the power of the Resilience feat a little (though who was ever taking it to boost a weak save anyway), and possibly decrease the power of the Monk ability to be proficient in all saves, or anything similar, though I'm not certain. What other downstream impacts am I not accounting for? Has anyone else tried it?

IF you were to add only half proficiency on saves, this wouldn't need that big an adjustment on save DC's.

Currently, even if you had maxed stat (excluding barbarian's cap stone and similar, or any external bonus such as magic items or spells), you get only up to +11 bonus on any save.

With this method, you may potentially reach a +13 (e.g. Str 20 and Con 20 plus half proficiency = 5+5+3), which is only 2 points higher than the standard.

Easy fix would be to adjust the calculation method for saves from '8+prof+ability' to '10+prof+ability', resulting in 2 points higher save DC's everywhere.

My alternative approach that I pondered on some years ago was to use averages (rounding up) of the two shared scored for each; Fort, Ref, and Will.

[(Str modifier)+(Con modifier)]/2+Proficiency=Fort
[(Dex modifier)+(Int modifier)]/2+Proficiency=Ref
[(Wis modifier)+(Cha modifier)]/2+Proficiency=Will

Eg
STR 20 +5
CON 16 +3
DEX 8 -1
INT 18 +4
WIS 12 +1
CHA 8 -1

Would have attribute bonuses to saves being:
FORT = +4 (5+3)/2
REF = +2 (-1+4)/2
WILL = 0 (1+-1)/2

[EdwardThereseJr]

My alternative approach that I pondered on some years ago was to use averages (rounding up) of the two shared scored for each; Fort, Ref, and Will.

The problem with averaging is that someone with a maxed Con, but average Str would have their Fort save brought down. While the math might not actually be that different from adding the scores together and upping the usual DC to compensate, it would feel worse to have 20 Con but only a +3 Fort because your Str is 13 or lower.


4e tweaked the 3e method by tying them to the highest attribute of each associated pair. This unfortunately lead to the unintended consequences of making classes/builds relying on two stats from the same defense weaker than those that didn't - as you end up with 2 weaker defenses vs 2 stronger defenses, for no added gains. We thought that each defense could be the SUM of the stats of the pair

I've never had an issue with the 4e model (or the 5e model), but your logic is sound here.

Ultimately, I think the main concern is how much does this affect the feel of niche abilities. Strength saves are pretty rare, but that makes the guy who has had to struggle against wisdom saves all campaign suddenly look more awesome when he has this one proficiency no one else has. Combine Strength saves into Fortitude saves, and now they feel a lot less special.

This might not be a concern in your campaign, but having a norm with outliers, while usually make the outlier in general less powerful, can help the outliers create more special moments. If that's less of a concern than balancing the saves, then your method might be preferable.

[Maat Mons]

You could just accept the weakness of the 4e model, that some some ability score combos aren't as good as others.

Even in unmodified 5e, there's one combo that already doesn't work well, Strength and Dexterity. And the only reason I can imagine anyone splitting their focus between two mental ability scores is if they're forced to by a multiclassing decision.



It actually would be possible to keep to the 1-to-1 correspondence between ability scores and saves and go back to the 3-save system... but you'd need to ditch the classic set of 6 ability scores.

[Kane0]

Instead of adding or averaging the two stat bonuses I would just pick the best of the two.

[Arkhios]

FWIW, I ditched the idea soon after, again some years ago. Just thought I'd bring it up since it was somewhat relevant to the topic.

[Ortho]

6 saves to track feels weirdly fiddly for a DnD edition whose design aim was simplicity.

Here's the thing: 5e did simplify saves. Conceptually, it's easier to grasp that each stat has a save than it is to mesh two stats into one save.

Practically, it's also simpler to have the saves just be equal to your stat, rather than doing math. Admittedly, you're not proposing anything complex, but I see people proposing taking the average, which is going to get very fiddly since you'd also be updating the save more often.

[Arkhios]

Here's the thing: 5e did simplify saves. Conceptually, it's easier to grasp that each stat has a save than it is to mesh two stats into one save.

Practically, it's also simpler to have the saves just be equal to your stat, rather than doing math. Admittedly, you're not proposing anything complex, but I see people proposing taking the average, which is going to get very fiddly since you'd also be updating the save more often.

It's true that 6 saves tied to equivalent abilities is very simple. But, my proposal had nothing to do with making it more simple than that. I merely brought up something that I had thought of before, as I've said multiple times earlier.

If anything, I see the attempt of going back to three saves as something that might appeal to "grognards" who may have trouble adjusting to not seeing the three saves they've grown used to over several editions. A little something to "ease their pain" so to speak. Also, having to do math more often is something someone used to, for example, 3.5 (and oddly enough: 4e) might expect to see in regards to D&D. I wouldn't say it's a bad thing, per sé. Besides, extra math is educational. Whether you like math or not is another question. :smallwink:

[Moreb Benhk]

If anything, I see the attempt of going back to three saves as something that might appeal to "grognards" who may have trouble adjusting to not seeing the three saves they've grown used to over several editions. A little something to "ease their pain" so to speak. Also, having to do math more often is something someone used to, for example, 3.5 (and oddly enough: 4e) might expect to see in regards to D&D. I wouldn't say it's a bad thing, per sé. Besides, extra math is educational. Whether you like math or not is another question. :smallwink:

For me its not so much so much about grognard appeal as it is simplifying conceptually (is a thing resisted by toughness, swiftness, or mental strength vs its resisted by strength toughness and not con toughness), plus it helps to even out the abilities a bit more - strength, charisma, int become less 'dumpstats' (depending on what your class actually needs) because so few abilities actually target them, and classes who max out these stats get better bang for buck. Anything that helps reduce dump-stattage is good in my books as it helps keep more diverse ability-arrangements a bit more viable, and reduce the cookie cutterness (eg - presently as a MONK for example, there is very little incentive to invest in INT or CHA at all, so if you want to be a knowledgeable-monk or a talky-monk you are taking a big hit for your character choice. If the INT and CHA at least boosted a common save type that would be a small payback for their investment.

For me the appeal of adding rather than averaging stats is the cool factor of getting to be REALLY good at one save, if you invest in it - a STR/CON barbarian is going to shrug off a lot of FORT targetting abilities. This helps characters feel more awesome.

[Onos]

The only reason I'd be hesitant to adopt this method is due to 5e's bounded accuracy. You run the risk of having a player be flawless at a given save if they really invest, though if you're tinkering extensively with your table it can be accounted for quite easily.

My 2 cents is to use the higher of the two stats and add half of the lower one. Simpler than averaging, but still gives the opportunity to "double down" on a save. Reckon I'll give this a whirl.

[Moreb Benhk]

The only reason I'd be hesitant to adopt this method is due to 5e's bounded accuracy. You run the risk of having a player be flawless at a given save if they really invest, though if you're tinkering extensively with your table it can be accounted for quite easily.

My 2 cents is to use the higher of the two stats and add half of the lower one. Simpler than averaging, but still gives the opportunity to "double down" on a save. Reckon I'll give this a whirl.

That's another interesting option to consider.

The changes I'm suggesting add only a usual maximum of +5 to saves as they currently stand. You'd have to invest highly to auto-succeed on lower DC saves - though this is already possible through paladin auras, bless, bardic inspiration... Then again I think that 5e already breaks its own rule of Bounded Accuracy in many ways, which I also think is a good thing. Bounded Accuracy just means that chance (the die roll) is always the biggest deciding factor.

EDIT - plus now that I think about it - flawlessness is actually impossible as iirc you auto-fail on a 1 and autosucceed on a 20. So no save-roll will ever be a fore-gone conclusion.

[JNAProductions]

That's another interesting option to consider.

The changes I'm suggesting add only a usual maximum of +5 to saves as they currently stand. You'd have to invest highly to auto-succeed on lower DC saves - though this is already possible through paladin auras, bless, bardic inspiration... Then again I think that 5e already breaks its own rule of Bounded Accuracy in many ways, which I also think is a good thing. Bounded Accuracy just means that chance (the die roll) is always the biggest deciding factor.

EDIT - plus now that I think about it - flawlessness is actually impossible as iirc you auto-fail on a 1 and autosucceed on a 20. So no save-roll will ever be a fore-gone conclusion.

That's for attack rolls only. It does not apply to saves or ability checks.

[Moreb Benhk]

That's for attack rolls only. It does not apply to saves or ability checks.

Oh wow. That's good to know. Wonder if we've been playing it wrong or not... tbh most of the time, given bonuses and DCs, its a bit of a moot rule - barring some diligent bonus stacking, you will fail a 1 and succeed a 20 regardless of how you build.

[Composer99]

Here's the thing: 5e did simplify saves. Conceptually, it's easier to grasp that each stat has a save than it is to mesh two stats into one save.

I don't know... the mental saving throws are a conceptual mess in 5e, when compared with the single, simple Will save. Also, the discrepancy between the "big 3" saving throws versus the "little 3", as implemented in 5e, is so great as to basically be a "Fort/Ref/Will" system with the serial numbers filed off.


Practically, it's also simpler to have the saves just be equal to your stat, rather than doing math. Admittedly, you're not proposing anything complex, but I see people proposing taking the average, which is going to get very fiddly since you'd also be updating the save more often.

I'm inclined to agree with the broader point here. I'm a big fan of the 5e design aesthetic, and if reintroducing the Fort/Ref/Will saves (or their 4e defences equivalent), I would also recommend that any such reintroduction be implemented with that aesthetic in mind.

[Lord Von Becker]

So my basic shot at this was "you add your Prof to all saves on account of being a hero, but saves you're actually proficient in get Advantage". Save DCs would go up by two.
But then I noticed that I wasn't really satisfied with how the stats work, so I probably want to work that out before I fix saves specifically.

[Moreb Benhk]

So my basic shot at this was "you add your Prof to all saves on account of being a hero, but saves you're actually proficient in get Advantage". Save DCs would go up by two.
But then I noticed that I wasn't really satisfied with how the stats work, so I probably want to work that out before I fix saves specifically.

An interesting take. Would lead to a net nerf on Save-based abilities, especially after the first few levels. Proficiency would initially be worth more but then not advance at all, so you'd get less class-differentiation. I think it would still leave all the issues I have with the current system too.

[Lord Von Becker]

An interesting take. Would lead to a net nerf on Save-based abilities, especially after the first few levels. Proficiency would initially be worth more but then not advance at all, so you'd get less class-differentiation. I think it would still leave all the issues I have with the current system too.
Would you specify those? Mine are roughly:
-Stats cost an enormous amount and are quite build-defining, but don't actually make you very good at things. The difference between a 10 and a 20 is a quarter of the d20's range, and that represents you being really exceptional. The range of stuff you can only attempt because you're amazing is very small; the range of stuff you can ignore by being amazing is very small.
-Training is a bit easier to get, except for saves, but it usually makes even less difference. Both together can make you semi-competent, but you have to stack them. The difference between a hyperspecialist and a generically-good-at-this is enormous; it is thus extremely difficult for generic specialists to use the skill minigame productively.
(--Relatedly, Jack of All Trades is rather unlikely to ever make the difference between success and failure.)
-Attacks assume roughly even victories for attacker and defender, assuming both sides have put in roughly the same effort. This feels nice and fair. Saves do not follow this pattern.
(--Spending a spell slot/other rest-based resource is, admittedly, putting in more effort. Have a limited resource to boost saves? Adapt Inspiration? Or just assume the effects are a bit more powerful than attacks tend to be?)

[Moreb Benhk]

Would you specify those? Mine are roughly:
-Stats cost an enormous amount and are quite build-defining, but don't actually make you very good at things. The difference between a 10 and a 20 is a quarter of the d20's range, and that represents you being really exceptional. The range of stuff you can only attempt because you're amazing is very small; the range of stuff you can ignore by being amazing is very small.
Agreed on this. Coupled with no differential in cost to acquiring higher scores - meaning rushing to 20 in your main stats is usually mechanically the best choice, while simulataneously not providing powerful differential for exceptionality.

The reason I started this thread is because Saves currently are caught up also by the problem of dump-stats - not all stats are created equal. Some are mechanically meaningful for most characters - DEX CON WIS, some are more binary - your class uses them lots so you should max, or they do not, and dumping costs you almost nothing (except for roleplaying purposes). Strong saves happen to map onto the already strong stats, doubling down on the disparity. So if your class happens to major on a weaker stat, not only does that limit your ability to raise the stronger stats, it also means your strongest saves are least likely to see use, and (in the case of STR) also only resist fairly weak effects. So when saves are in play, you get less opportunities to be awesome and more opportunities to fail.

I toyed with the idea of going to 4e style - and make it a 'choose the best of 2' and squeezing saves back into 3 - FORT (Str or Con) REF (Dex or Int) WILL (Wis Cha) - but that has its own issue of inadvertantly punishing classes who rely on stats from the same save - a Str + Con Barbarian is generally worse at saves then than... say a Strength + Wisdom Paladin or somesuch.

What I would love is something similar to what Pillars of Eternity did - which is providing mechanically meaningful benefits for all stats to all classes. But that'd take some pretty fundimental reworks.



-Training is a bit easier to get, except for saves, but it usually makes even less difference. Both together can make you semi-competent, but you have to stack them. The difference between a hyperspecialist and a generically-good-at-this is enormous; it is thus extremely difficult for generic specialists to use the skill minigame productively.
(--Relatedly, Jack of All Trades is rather unlikely to ever make the difference between success and failure.)
I'd say that I'm not sure I sit with you on this one. Plus it sounds a direct contradiction to your first point. I'd say that in general play I've mostly noticed that the skills checks are dominated by the d20 - roll low and fail roll high and succeed - until you start to get to mid-game. That said, higher levels + expertise + advantage + some of the other skill bonuses (eg - bard insp etc) can easily break through most DCs. I find in general the Bounded Accuracy schtick caps the things that could ever be achieved through the skill system (whereas Spells don't suffer such bounded limitations)



-Attacks assume roughly even victories for attacker and defender, assuming both sides have put in roughly the same effort. This feels nice and fair. Saves do not follow this pattern.
(--Spending a spell slot/other rest-based resource is, admittedly, putting in more effort. Have a limited resource to boost saves? Adapt Inspiration? Or just assume the effects are a bit more powerful than attacks tend to be?)
Yeah the investment/payoff sizes for abilities vs attacks is often rather a lot higher. I don't have a problem with this myself. Some limited usage ability to 'unfail' a save. I'd love (on both attacks and abilities) more of a teired success/fail system - so its not simply a suck-or-nothing affair.

Further - I've toyed with swapping in 3d6 for all d20 rolls, which effectively amplifies the impact of the smaller static bonuses, and opens up interesting tiered options for Advantage/Dis. But as I mentioned at the start- given that investing from 18 to 20 costs the same as 10 to 12, its only amplifies the urge to max out your key stat and further punishes jack-of-all-trades types. To my knowledge DnD has always struggled to encourage anything other than binary investments (min/max) except by making such choices relatively non-impactful. By contrast, a system like Shadowrun (though it has plenty of other issues) manages to get something akin to bounded accuracy (if you roll bad and baddies roll well even street thugs can be dangerous) while giving meaningful differences when investing to be expert, while also making the choice between being proficient and being expert a viably interesting one.

[Segev]

That's for attack rolls only. It does not apply to saves or ability checks.

:smalleek:

Well, I've been doing that wrong. I had seen the specific rule for attacks, and failed to notice it was lacking for saves, and, coming from 3.PF..... woops.

This has mattered, too: the wizard's familiar managed to nat-20 out of a froghemoth's tongue-lashing.

[Moreb Benhk]

:smalleek:

Well, I've been doing that wrong. I had seen the specific rule for attacks, and failed to notice it was lacking for saves, and, coming from 3.PF..... woops.

This has mattered, too: the wizard's familiar managed to nat-20 out of a froghemoth's tongue-lashing.

I imagine it would still have been close. Plus made for an epic and tense moment. I wouldnt' be surprised if its an accidental house rule in a lot of tables with 3.5 / 3.PF experience.

[Segev]

I imagine it would still have been close. Plus made for an epic and tense moment. I wouldnt' be surprised if its an accidental house rule in a lot of tables with 3.5 / 3.PF experience.

The DC was, IIRC, 23, and the familiar had a negative mod to the save. It wasn't possible to make it without nat 20s auto-succeeding. ^^; It did mean the little owl slipped right out and didn't get chomped as a bonus action.

[Moreb Benhk]

The DC was, IIRC, 23, and the familiar had a negative mod to the save. It wasn't possible to make it without nat 20s auto-succeeding. ^^; It did mean the little owl slipped right out and didn't get chomped as a bonus action.

Sounds like some high level sheenanigans. I don't think I've ever seen a published save DC over 20. But I'm sure some must exist.

[Segev]

Sounds like some high level sheenanigans. I don't think I've ever seen a published save DC over 20. But I'm sure some must exist.

Happen to have the book handy now. It was DC 18. Still, the owl had a -4 Str mod and no proficiency to the save, so even on a 20 couldn't get a DC 18.