Last time, I talked about the divergence between strong saves, weak saves, and saving throw DCs, as observed in 3.x and theorized in 5e. Today, I will discuss two save systems that avoid the problems associated with save divergence, namely the high susceptibility of high-level characters to the nasty save-or-suck effects that get thrown around at that level.
Turns out, TSR got one thing right in B/X. Well, at least one thing. The fact that their saving throws grow in a very peculiar way is often obscured by the ridiculous category names that they chose for their saves, but upon some examination, it is notable that save DCs don't exist. There is no question of your stats vs the stats of the guy who cast Finger of Death on you; your class and level give you a target number that you must roll to resist, and barring a handful of spells and effects that specify a modifier to your save, that's all that goes into it. The save DC was inside you all along. And that's actually how it started, too; saving throws were "You're dead to rights, but throw the dice and maybe luck will save you." You weren't any deader-to-rights if the guy casting Finger of Death on you had 18 Int vs 16 Int, and high level characters were assumed to be tougher and luckier, so their saves were substantially better than at low levels.
This has some interesting effects on high-level [TO]SR play. At low levels, spells that give saves are pretty great, because the enemies have few HD and consequently really bad saves. We added a save vs death to Sleep, and it still wipes rooms with no problem. But, at this level, spellcasters have few slots and little versatility. At high levels, they have tons (well, more) of slots and more spells to choose from, too! ... but everyone's saves are much better than they used to be, so your spells are much more likely to fail or have limited effects. There are a couple variations in the uniformly good saves, like fighters being relatively weak in saves vs spells, and clerics being particularly strong in saves vs death, but most saves tend to be pretty clustered together somewhere in the 50-70% success range (at ACKS' level cap, at least). And high-HD monsters have really good saves, like -1+. So at high levels, wizards are much more useful for taking out massed weak opponents or doing battlefield control or buffing or summonging than at save-or-sucking single targets. This is the balance of the quadratic wizard. They can have effective spells but few slots, or ineffective spells and many slots. The power of the spellcaster is tied inextricably with the structure of saving throws.
So, summary: in TSR/Basic-derived games (caveats because I don't feel like digging around in my backups for OSRIC...), saving throw success rates get strictly better over levels. This is exactly the opposite from the way saves work in WotC games. What's more, the TSR save system is dead-simple outside of its save categories, requiring no math. It's very straightforward to analyze, and its consequences fall out naturally. Convenient!
The downside, of course, is that since saves are "you're dead to rights", at low levels, your saves really, really suck. About as badly as save suck at high levels in 3.x, really. When the entire party gets poisoned in an OSR game at low levels, you probably have one or two survivors, much like a banshee in high-level 3.5. Further, all of your saves suck at low levels.
I think there's a happy middle ground to be had between the TSR and the WotC Ways of Saving, though. It plays nicely with the Hero's Journey, too.
Consider the archetypical farmboy with a sword. He might be inured to bodily distress from years of hard labor, and honest, pious, and pure of mind as befits his status as 'salt of the earth', but he's not quick on his feet or courtly-mannered (I can only assume that Cha saves are for extricating yourself from social situations), nor will he best the Sphinx in riddles or perform feats of great strength. But over the course of his journeys, he will become able of these things, and by the time he achieves full hero status, he is resistant to many things which would've spelled his end as a callow youth.
And so, convergent saving throws. You get a set of good saves; maybe they're from your class, but even better would be from your background. These saves start out high and grow slowly (say +6 +1 per three levels, on a 3.x scale) while the rest of your saves start out low but grow more quickly (+2+1 per two levels, again on a 3.x scale), meeting your good saves at or near the level cap.
Consequences: low-level characters are flawed and vulnerable, but not absolutely-godawful-dies-to-anything vulnerable like they were in TSR. High level characters are invulnerablish like they were in TSR D&D. They are also more resistant to effects whose DCs grow at the same rate than they were at low levels; they fail to degenerate, except possibly in their strong saves, but that's acceptable, because those start really strong. Spellcasters are forced to diversify their role in combat at high levels and rely on finger of death and similar single-target save spells for dramatic effect, taking out soft targets quietly, or as desperate gambles.
To return to Trailblazer's problems with the save structure: action points are a dirty post-hoc patch rather than a systemic solution. If saves aren't growing at rates sufficient for them to remain effective at high levels, change the base save growth rate. If saves are diverging and causing people to keel over dead whenever they get hit in a soft save, fix the cause of the divergence. Hell, give bad saves such a high growth rate that they pass good saves, on the assumption that good saves will tend to be backed by good stats. I dunno man. Write new laws on new tablets.
... quoting Nietzsche is probably a good sign that I should sleep.