D&D General How Saving Throws broke in modern D&D

[ezo]

Regardless, 6th ed. does not seem to have diverging saving throws, so I think the problem, as I mentioned earlier, is ceasing to be a problem.

HUH?

 

[mellored]

adding half prof bonus to non-proficient saves is enough to even the playing field;
+1 at 1st level
+2 at 9th level
+3 at 17th level

I do this.

I like that PCs strength and weakness growing wider as they level.

But I also feel a level 20 barbarian should not be as vulnerable to a goblin casting sleep as a level 1 barbarian.

 

[ezo]

But I also feel a level 20 barbarian should not be as vulnerable to a goblin casting sleep as a level 1 barbarian.

Unless they are extremely low on HP, there is practically no way a 20th-level barbarian would be affected by sleep...

 

[GMMichael]

HUH?

 

[mellored]

Unless they are extremely low on HP, there is practically no way a 20th-level barbarian would be affected by sleep...

Well, 2024 sleep.
Or cause fear or whatever then.
Or a level 20 wizard making Str saves.

Point being that high level PCs should of picked up a few tricks even for their weakness.

 

[ezo]

Point being that high level PCs should of picked up a few tricks even for their weakness.

Agree completely. The fact most PCs have four saves out of the six that never get better is stupid IMO.

Our current house-rule (although our current game is more RAW) is everything is proficient in all saves. Anything you used to be proficient in for a save now has advantage. Monks, via Diamond Soul, end up with advantage on all saves.

I just didn't know if you meant sleep or something else like charm, etc. That's all.

 

[Clint_L]

On one hand, there's very little reason why level 20 in each class shouldn't end up with about the same hit points and the same saving throws - they're all superheroes anyway.

I disagree. I like that a high level barbarian can laugh off strength-based tests, but still pees their loincloth at the sight of an intellect devourer, and vice versa for a high level wizard.

 

[ezo]

adding half prof bonus to non-proficient saves is enough to even the playing field;
+1 at 1st level
+2 at 9th level
+3 at 17th level


or add proficiency bonus to all saves,

class bonus give +1 to two saves.

Another option would be non-proficient saves progress as proficiency bonus - 2, so it would max out at +4 instead of +3 and you don't get any bonus in tier 1, where (as you are just starting out) I don't think a bonus to non-proficient saves should be given.

Just an alternative way to do it.

 

[AlexofBarbaria]

Aren't save generally for half damage? A 10 point spell deals 10 or 5 (generally). There is likely some math in the save and take 5 or fail and take the 10.

Right but that doesn't matter for our purposes. My example with 10 damage holds whether it was save and take none/fail and take 10, or save and take 10/fail and take 20 (the 10 being the extra 10 damage in this case). It also holds for any bad status effect happening on a failed save because we can quantify them with arbitrary units (e.g. Dazed is worth 30 badness points, Confused is 50, etc.)

The basic idea is just this asymmetry between sliding towards auto-miss and auto-hit (whether for attack rolls or saves). Each step is worth proportionally more when sliding towards miss than hit. Hitting every time isn't game-changing; it's only slightly better than hitting almost all of the time. Missing every time is a game-changer because now your attack is worthless.

So at the level of a system's basic math, it's totally fine for offensive stats to diverge with level between characters/classes. But we should be very careful about allowing defensive stats to diverge. My theory is the 3e designers didn't properly recognize Save Bonus as a defensive stat.

You might take three times the damage, but the damage is still considerably less than it used to be. As the actual numbers grow ever smaller the relative percentage keeps growing, but that percentage is not really relevant, the actual damage is (unless you believe the damage dealt by monsters were scaling with the percentage of damage saved by the characters, but I very much doubt it does).

Fair point that the actual impact of this depends on how bad it is to fail a save. However this certainly does scale with level; in fact in 1e-3e I'd say it outpaces level. Fireball does another d6 damage each level while being cheaper to cast, the status effects on a failed save proceed from falling asleep to being held to being disintegrated, etc. The whole high level Rocket Tag phenomenon testifies to this.

And, on top of that, how do you analyze spells that don't do damage but cause loss of actions - the save or sit/die spells?

Looking at 3e saves, goods saves were generally 1/2 level, bad saves 1/3 level. Good saves started with a bonus +2, bad saves +0. The thing you're supposed to compare them with is the level-based piece of the save DC, specifically, the highest level spell a same-level caster can hit you with. The good save is generally ahead of the level-factor, the bad save is behind, culminating in the good save at a base +12 being 3 points higher than the 9th level spell's +9 to the save DC, the bad save being 3 points lower than the 9th level spell. And with the DC starting at 10+spell level, the target had a 50% chance of saving, 50+15% for the maximum good save, 50-15% for the maximum bad save. From a design perspective, that's not a terrible place to start.

I disagree, I think this was a poor place to start I think the math would be much improved with two changes:

1) the Save Bonus difference should start at +3/-3 and converge to 1/-1 at level 20. This in fact would be a close approximation of the 2e Save vs. Spells progressions. The Warrior and Wizard start out at 5 points difference (17 and 12) and close to 0 (6 and 6) at level 17.

2) The baseline should be more like 1.5 x PC level than at-level, taking into account the party tends to fight above their weight. Tying Save DCs to CR instead of level/HD certainly makes sense and helps, as long as CR is an accurate measure.

I think something really needed to be done about saving throws compared to 1e/2e because the way those worked, high level save or sit/die spells became increasingly useless against level-appropriate opponents - they were simply more likely to save and waste the caster's time because higher powered targets had (mostly) better saves across the board (and don't get me started on how shafted thieves are with the AD&D saving throw table). But 3e may have taken things too far in allowing the save or sit/die spell become too good from min-maxed casters. 5e is somewhere in the middle because bad saves really do stay bad, but save DCs tend not to get so out of hand, and concentration rules/repeated saves each round significantly blunt the power of save or sit spells by making them easier to eventually end even if the initial save fails.

Agree with all of this. 1e/2e saves do become a titch too high at high levels. 5e does do a lot to solve the Rocket Tag problem as you say, by nerfing high level spells/effects and/or making them more expensive. Would have been much easier (and IMO more flavorful) to simply bring the underlying math closer to 1e/2e.

I'm not sure I understand what the issue is. Is there a widespread perception that saving throws should converge more at high levels? To me it makes more sense that they would diverge. Though that might be because it's just what I'm used to. However, is the current system not working? Is saving throw progression widely perceived as a problem in games as they are actually played? It's not something that I can recall ever coming up.

Even in 3e? I haven't played high level 5e myself (I've gone back to 1e/OSR) but by all accounts the issue is not as bad, even though saves still diverge, because the effects of failed saves have been toned down considerably. I consider that a less flavorful and less elegant solution than just going back to the 2e math (kind of like crippling your engine to make a flat tire safer instead of just fixing the tire), but that leans into personal preference.

Insightful analysis, but I don't really want that result.

I know it used to be that as characters and monsters got stronger they just better at resisting magic, and the source (caster, spell, etc) of the magic rarely mattered at all (other than the occasional spell that said you saved with a -4 penalty or such). But...that isn't appealing to me. While it my be fun for a while that once I'm high level I rarely ever fail a save, I just think it would make opponents feel underwhelming. And playing a caster, well that just the longer the game goes the less likely my spells become to work against our opponents--even the higher level ones I'm getting.

The 3e+ style where there is an ascending defensive skill vs an ascending offensive as a parallel to ascending AC vs ascending to hit just feels more natural and satisfying to me. (Although to be fair, other 4e AC has never kept up with attack bonuses.)

But again, good analysis, and I learned from it.

Glad you liked it! I agree the 1e/2e saves scale just a bit too quickly. The context of OP is I'm working on my personal AD&D 3 that tries to bring just the d20 + bonus vs. DC core mechanic back to 2e while hewing much more closely to the existing math than 3e did. Will be posting more about that in the old D&D forum shortly.

On the other, isn't being a little broken the nature of D&D? Why do you think the masses rejected the perfect Matrix 4th edition? Regardless, 6th ed. does not seem to have diverging saving throws, so I think the problem, as I mentioned earlier, is ceasing to be a problem.

Does it not? I haven't looked into 6e at all. 5e has diverging saving throws (2 progress according to the proficiency bonus for each class). Does it work differently in 6e?

 

[UngeheuerLich]

adding half prof bonus to non-proficient saves is enough to even the playing field;
+1 at 1st level
+2 at 9th level
+3 at 17th level


or add proficiency bonus to all saves,

class bonus give +1 to two saves.

Make it +2 for class bouses and make DCs for saves 10+prof bonus+ability modifier.

So at low level everything stays the same.