Fixing the 5e saves

So there is an issue with the math at high level with regards to saves.

The first thing you have to decide is do you think high level characters should save more or high level effects should make them fail their save more. Once you have that figured out I think you can clearly decide the direction to go. Currently as set up the system allows two saves that give you a good chance at saving. Outside of that you have four other saves that at high level have not scaled at all and are going against an attacker who has scaled their DC up. This results in a very high fail rate.

Looking at 1e/2e the rates of success grew as you leveled up. Then in 3e/4e you had 3 saves and they scaled up and so did the DC. 5e splits and makes it very hard to defend against some effects because the DC is too high. So this thread is to look at several options.

Option 1
all stats have proficiency for saves.

Option 2
Add advantage to the save roll instead of proficiency. All saves work like a death save: '10' or greater is a success. Period. Don't mess with DCs and bonuses.

Option 3
Add yours here.

I don't have any suggestions yet because I'm still thinking what needs to be fixed.

In 3e we had:

- DC = 10 + stat mod + spell level
- 'good' ST = d20 + stat mod + 2 + 1/2 class level
- 'bad' ST = d20 + stat mod + 1/3 class level

Ignoring the anomaly of stacking ST in multiclassing.
Spellcasters have an easier time focusing on increasing one single stat mod, while defenders in theory have 3 stat mods to care about.
We also had unbound stats, which favor the spellcasters even more.
Furthermore, the spellcaster can often choose which ST to target, while the defender cannot choose.

So base DC goes up to 19 + stat mod.
Base 'good' ST goes up to average 22.5 + stat mod.
Base 'bad' ST goes up to average 16.5 + stat mod.

Difference between base DC and 'good' base ST goes up to -3.5 (defender needs a roll of 7+)*
Difference between base DC and 'bad' base ST goes up to +2.5 (defender needs a roll of 13+)*
Difference between 'good' and 'bad' base ST goes up to 6.
*assuming equal stat mod, which is not usually true, so the defender typically needs higher rolls

These are the limits, at each level the differences are not the same.
Something to notice is that these limit DC are for 9th level spells, but lower level spells have much lower DC.

In 5e we have:

- DC = 8 + stat mod + prof bonus
- 'good' ST = d20 + stat mod + prof bonus
- 'bad' ST = d20 + stat mod

In this case the DC is the same for all spells, and prof bonus increases equally for the caster and the defender.

Difference between base DC and 'good' base ST is always -2 (defender needs a roll of 8+)*
Difference between base DC and 'bad' base ST goes up to +4 (defender needs a roll of 14+)*
Difference between 'good' and 'bad' base ST goes up to 6.

So it does look like 5e is more advantageous towards spellcasters, making it generally harder to save vs spells, especially because this is for ALL spells, not just the highest-level ones.

The only mitigating factor is that the spellcaster won't normally enjoy a stat higher than 20, so the gap with the ST stat is not going to be as much as in 3e. This is hard to evaluate, but I think that having less than 10 in a stat is rare in 5e, so even when the spellcaster guesses right and targets the weakest save, she will have a further +5 bonus at best. I think this was often significantly higher in 3e.

Basically it's not that different between 3e and 5e, if you consider high level spells, but it is quite a lot different with regard to low-level spells cast by high-level spellcasters, because their spell DC is the same.

You could change the DC to be 10+spell level (cantrips being 0)+caster stat mod.
This takes the proficiency bonus out of the equation, all low level spells will be easy to save against for most people, all high level spells will be very hard unless you have proficiency in the save.

The major issue with this is like 3e, there will be different saves for every spell level this adds an extra layer of complexity that 5e game design is trying to avoid.

I still think that magic items make a good patch for the current save system, we will see rings and cloaks of resistance soon enough that will help with saves. Also don't forget inspiration, I would think at higher levels saving your inspiration for when you get hit with a save or suck/die spell is the best idea.

I like the half-proficiency bonus to non-proficient saves house rule. A +1 to +3 makes a bit of a difference without getting too crazy. For those who don't like it right away, it could kick in at 5th when proficiency bonus first goes up.

So at 1st : +2/+0, at 5th: +3/+1, at 9th: +4/+2, at 13th: +5/+2, and at 17th: +6/+3.

RhaezDaevan said:

I like the half-proficiency bonus to non-proficient saves house rule. A +1 to +3 makes a bit of a difference without getting too crazy. For those who don't like it right away, it could kick in at 5th when proficiency bonus first goes up.

So at 1st : +2/+0, at 5th: +3/+1, at 9th: +4/+2, at 13th: +5/+2, and at 17th: +6/+3.

Click to expand...

Another possible slight variation on that could be proficiency -2:

1st : +2/+0, at 5th: +3/+1, at 9th: +4/+2, at 13th: +5/+3, and at 17th: +6/+4

This one tackles the high level splitting off of the proficient vs. not proficient better than just 1/2 proficiency. The two numbers do split still because they are progressing at different rates. Monsters with CR higher than 20 would begin increasing the difference between good and bad saves more and more.

The better more intuitive way to showcase this options would be like this:
All saves are 8 + Proficiency + stat
Classes get +2 to the saves they are trained in.

Of course the flip could be done too:
All saves are 10 + Proficiency + stat
Classes get -2 to the saves they are not trained in.

This looks pretty solid to me.

Li Shenron said:

I don't have any suggestions yet because I'm still thinking what needs to be fixed.

In 3e we had:

- DC = 10 + stat mod + spell level
- 'good' ST = d20 + stat mod + 2 + 1/2 class level
- 'bad' ST = d20 + stat mod + 1/3 class level

Ignoring the anomaly of stacking ST in multiclassing.
Spellcasters have an easier time focusing on increasing one single stat mod, while defenders in theory have 3 stat mods to care about.
We also had unbound stats, which favor the spellcasters even more.
Furthermore, the spellcaster can often choose which ST to target, while the defender cannot choose.

Click to expand...

Excellent analysis, however I want to pull out one point. I am in a 3e high level game right now and I am a wizard with a 35 INT. It is all kinds of fits. 3e is not a good way to model saving throws in the game imho. I also feel like 1e/2e is not the right way too. The math has to be able to still make it a threat at higher levels, in 1e/2e you basically failed on a 1 if you had a ring and any class or race bonuses. It needs to be in the middle of these two extremes.

I had another idea:
Everyone has proficiency in every save d20 + stat + proficiency = avg roll at highest 22.5
The "good" saves you get double proficiency d20 + stat + 2*proficiency = avg roll at highest 28.5

Then DCs for spells 10 + stat + proficiency = 21 at highest

High level characters should save more often.