D&D 1E There were 10 saving throws

[Neonchameleon]

Grouped in 5 types,
Paralysation.
Poison
Death magic
Petrification
Polymorth
Rods
Staves
Wands
Breath weapon.
Spell

I have absolutely no hint of what r/s/w was talking about...

They made the saving throws functional then grouped them. You used whichever category you hit first.

Death/Paralysis/Poison a.k.a. Save or Die. The easiest group to save against and the most fatal. In earlier versions Paralysis was under Save or Suck - but in practice it was followed up by a coup de grace so was effectively a death spell
Rod/Staff/Wand a.k.a. Spell-in-a-can and easier to save against than default spells. The item does the work not the caster. It was always one point easier than save vs spell.
Petrification/Polymorph a.k.a. Save or Lose. Fail to save against one of these spells and you're out of the fight. But not dead and can at least scamper away or are protected by being stone
Breath Weapon a.k.a. Other Physical. If it wasn't poison and wasn't magic it came here.
Spell a.k.a. Other Spells. When you cast something and it wasn't one of the above it was a generic spell

One of the (many) things 3.0 did that overpowered wizards was threw this out of the window. In AD&D Stinking Cloud was a bad spell because it was a save vs poison, meaning it was pretty easy - while Fireball or even burning hands always did something and was hard to save against. Also spells got easier to save against at higher level in AD&D and harder in practice in 3.0.

 

[Lyxen]

One of the (many) things 3.0 did that overpowered wizards was threw this out of the window. In AD&D Stinking Cloud was a bad spell because it was a save vs poison, meaning it was pretty easy

Uh no, it was easy for some classes and harder for others just as in AD&D. In AD&D, poison was easy for fighters, just as fighers had proficiency on fortitude saves in 3e (and poison was harder for magic-users, just as wizards did not have fortitude proficiency in 3e). It does not translate easily as 3e was more standardised and proficiency on saves varied per class instead as per group of classes, but the general principle was the same.

while Fireball or even burning hands always did something and was hard to save against.

Again, it depended, but because there is no direct mapping between the save categories, the groupings were different even if the principles were the same.

Also spells got easier to save against at higher level in AD&D and harder in practice in 3.0.

Where 3e dropped the ball was in the fact that at high level, there was a huge difference between proficient and not proficient in a save, which meant that if you used the right spell, it was extremely hard to save, whereas it was almost too easy if you were proficient.

This was "cured" in 5e by bounded accuracy, but it's also because the power range of 5e is much more reduced than in 3e, a 20th level 5e character is barely the equivalent of a 12th level in 3e. Not that it's a bad thing, but it makes the games less epic.

 

[Helpful NPC Thom]

The old saving throw categories were bizarre and a product of organic design, but like 18/00 Strength, they have their charm. In a way, I miss the naming conventions because they're evocative in comparison to "Strength save" or "Dexterity save."

 

[Neonchameleon]

Uh no, it was easy for some classes and harder for others just as in AD&D. In AD&D, poison was easy for fighters, just as fighers had proficiency on fortitude saves in 3e (and poison was harder for magic-users, just as wizards did not have fortitude proficiency in 3e). It does not translate easily as 3e was more standardised and proficiency on saves varied per class instead as per group of classes, but the general principle was the same.

And monsters didn't have classes. Most of them would be using the warrior table.

Again, it depended, but because there is no direct mapping between the save categories, the groupings were different even if the principles were the same.

Which is missing the point given that fireball did half damage on a miss so always had an effect. Fireball always did something.

Where 3e dropped the ball was in the fact that at high level, there was a huge difference between proficient and not proficient in a save, which meant that if you used the right spell, it was extremely hard to save, whereas it was almost too easy if you were proficient.

That's only one of the many balls 3.X dropped. Yet another one was that any decently prepared caster would have a range of spells to target all three saves.

 

[billd91]

Where 3e dropped the ball was in the fact that at high level, there was a huge difference between proficient and not proficient in a save, which meant that if you used the right spell, it was extremely hard to save, whereas it was almost too easy if you were proficient.

I'm not sure I'd necessarily call it dropping the ball. There are factors in place that argue for the design.
1) The system was designed with a certain degree of symmetry in the end result. Strong saves ended up 3 points higher, weak saves 3 points lower than the highest contributing factor in a Save's base DC for a high level character. +12 or +6 vs the +9 to the DC for a 9th level spell.
2) Resistance bonuses were far cheaper than any enhancement bonus to a stat, so magic item pricing and design favored defense.
3) Each of the saves had a feat to boost its save that was a higher bonus than any corresponding feat to boost a spellcasting DC.

But that's all on paper. And with pre-gens that weren't pushing optimization, it might have worked OK.

Where I think this broke down is in integrating with the rest of 3e and player behavior in the wild where players are free to rev a game over it's red line.
Uncapped stat increase meant that stats people would focus on for increase, like a spellcaster's prime casting stat, could grow obnoxiously large compared to defensive bonuses - and it would always pay something as a side benefit so it was virtually always worth boosting it. And with the effect of an easy and player controlled magic item economy, people really went for it to the benefit of caster power/detriment of saving throw power.
This is also seen a bit with the save DCs of big monsters too, particularly any based on Constitution which jumped up with a monster's size due to monster design guidelines and the way HD usually needed to stack up to keep pace with PC power.

In a system where stats were capped, like in 5e, I think the 3e saving throws would work MUCH better than they did in 3e.

 

[Lyxen]

And monsters didn't have classes. Most of them would be using the warrior table.

Indeed.

Which is missing the point given that fireball did half damage on a miss so always had an effect. Fireball always did something.

That hasn't changed, but hit points were usually lower in AD&D especially at high level.

That's only one of the many balls 3.X dropped. Yet another one was that any decently prepared caster would have a range of spells to target all three saves.

That is true, after a certain level at least.