Hacking 2e: Saving Throws – Part 1

He's a big boy.

Judging by how often they’ve changed through the history of the game, the rules for saving throws might be the least sacrosanct of the D&D mechanics. For more details on the origins and changes throughout the editions, DM David and Grognardia talked about that years ago, but the rundown is this: TSR-era D&D uses saves against specific perils, with static values based on class and level. OD&D, AD&D, and Basic each have different groupings and values for these saves, with Paralyzation wandering about the table as the years passed. WotC threw it all out with 3rd Edition and created Fortitude, Reflex, and Will saves, now adjusted by your ability score. 4e turned these three into static defense numbers on par with AC, and used saving throws as a state-tracking mechanic. Finally, 5e reintroduced saving throws as a reactionary tool, but just made them modified ability checks. In short, the saving throw mechanic managed to save vs Petrification over the past 50 years.

The OSR approach to saves is anywhere between reverential and heretical. The faithful go no further than renaming the categories and maybe tweaking some numbers (though a notable exception is Swords and Wizardry, which cut saves down to a single column). The hacks, and other OSR content not wed to the sacred texts, just use 3e or 5e save mechanics or excise saves entirely. The dividing line is compatibility to older material; if you want your game to work with X1 or whatever, you’re going to make sure it has a ‘save vs Poison’ of some sort. Otherwise, keeping this system can only be a preference, either due to familiarity or in an attempt to evoke ‘the old school.’ Without such a need, a designer might discard them in favor of ‘better’ approaches.

Dolmenwood serves as an interesting case study in having your cake and eating it, too. Here, the Basic D&D saves are presented in an altered form: Doom, Ray, Hold, Blast, and Spell. The categories have been simplified in their presentation to the player, yet still retain their original assigned purpose – Doom still covers Poison and Blast still includes Dragon Breath. More than this, they’ve been expanded to include new meanings (Hold for falling rocks, Blast for explosions, etc.), moving beyond specific, proscribed events. In this way, the game maintains technical compatibility (well, at least with saves) while forging its own path.

I think that’s sufficient to say we have a bit of room to play around in this space. First, let’s look at some problems present in the 2e Saving Throw system. 

 

Problem 1. Missing Saves

There are Reflex and Will saves in AD&D, but they’re hidden and bundled into other saves. Per the 2e PHB, Dexterity provides the ‘Defensive Adjustment’ which adjusts “attacks that can be dodged—lightning bolts, boulders, etc.” and Wisdom provides the ‘Magical Defense Adjustment’ which adjusts “magical spells that attack the mind: beguiling, charm, fear, hypnosis, illusions, possession, suggestion, etc.” Instead of offering adjustments to distinct save categories, they impact any save as long as they meet these restricted set of conditions.

This exception-based saving throw mechanic requires something I find antithetical to good TTRPG design: expecting the player to remember stuff. If the player needs to remember to add that +2 every time it kind-of somewhat sounds like Dexterity could apply to the save, they’re just never going to do that.

If we’re giving numerical bonuses, then players need to write those down somewhere; that somewhere should be a single, easy-to-find location. AD&D provides no such easy notation for this conditional bonus, as it can affect any of the saves. Players just need to write it in the margins. Having an extra line halfway down the character sheet that says “also don’t forget this bonus when this situation applies” is a recipe for a forgotten modifier.

 

Problem 2. Under- and Overutilized Saves

Breath Weapon, Rod/Staff/Wand, and Petrification or Polymorph are woefully underused in AD&D, at least officially. Community response to how and when each save is actually used runs the gamut, from creative modifications of the save’s domain to defaulting to ability checks for anything not strictly matching the categories. To learn how TSR thought you should play it at the table, I ended up cataloguing 236 different instances of saving throw instructions.

I used seven sources in total:
AD&D 2e Player’s Handbook: 93 saves
AD&D 2e Monstrous Manual: 56 saves
S1 Tomb of Horrors (1e, 1978): 8 saves
N1 Against the Cult of the Reptile God (1e, 1982): 5 saves
I6 Ravenloft (1e, 1983): 2 saves
The Dancing Hut of Baba Yaga (2e, 1995): 32 saves
Night Below (2e, 1995): 40 saves

You’ll note that saves in 1e adventures are very low in number. Partly this is because I didn’t record any saves that were based on monsters already found in the Monstrous Manual, and partly because the 2e adventures have longer page counts, but that only accounts for so much. For the most part, 1e modules (at least the ones used here) simply didn’t use saves as often. It’s a small sample size, but it indicates an expansion of saving throws as a DM tool in later 2e.

The final totals of all used saves are as shown below:

Save	PHB	MM	1e Modules	2e Modules	All
Paralyzation	1	4	1	7	13
Poison	3	25	8	13	49
Death	3	5	0	5	13
RSW	0	0	0	4	4
Wand	1	3	0	1	5
Rods	1	0	0	0	1
Petrification	2	1	0	2	5
Polymorph	1	0	0	0	1
BW	0	5	0	0	5
Spell	81	13	6	40	140
Total	93	56	15	72	236

 

It shakes out that about 90% of all saves fall into either Paralyzation, Poison, or Death Magic (PPDM) or Spell.

Obviously, Spell is winning out in the PHB due to all the spells found there, and likewise for Poison in the Monstrous Manual, where all the poisonous things are. Still, this appears to extend to the adventures as well, with Spell comprising half of all saves and Poison another 25%. Most Poison saves are used for actual poison, but Spell is used for everything from avoiding insanity to avoiding spear traps, and a third of Spell saves in the adventures are used to withstand mental attacks. This tracks with the spells in the PHB, where a little under half affect the mind.

The remaining saves are either used for their exact, titled purpose, or used for odds and ends. Breath Weapon is exclusively used for literal breath weapons. Rod, Staff, and Wand is used to avoid falling rocks, to combat the illithid's mind blast, and to dodge exploding walls of acid. Even Paralyzation, taped to the side of Poison and Death Magic, has some odd uses: aside from actual paralyzing attacks, it is used for the bard's influence, to avoid dizziness, to not asphyxiate, and when combating Explictica Defilus's Permanent Cult Charm.

This means we’ve got three saves that are so hyper-specific that they almost never apply (RSW, P/P, and BW), one that smashes together three different saves that are not used equally (PPDM), and one that pulls double duty for both magical effects and mental effects (Spell).

 

Problem 3. Classes are Good at Weird Saves

Saves in AD&D are class and level based. As a result, their values tell a story about which classes are better against certain types of threats. That story is a little off from expectation.

RSW not listed, as it's just the Spell chart minus one.

Priests are exceedingly good at withstanding Paralyzation, Poison, and Death Magic, being better at it than any other class in any other save at level one. Rogues have their best defense against Petrification or Polymorph, and their worst against Breath Weapon. Warriors are the worst at everything to begin with, though their best of the worst is against PPDM, and they’re the best at everything compared to all other classes by mid level, where they begin to outpace the priest’s and rogue’s head start. Wizards are, sensibly, the best at Spell saves (and, by extension, RSW) early on, and they stay the best, but are okay to poor at everything else.

That’s an odd mix. The wizard is right where it should be, and I’m a fan of the warrior’s quick ascent to being awesome, but the priest and especially the rogue come off a bit sideways. I can see the suggestion that others have made for the priest being amazing at PPDM: they’re being saved by the intercession of their god or some other such divine intervention, though their Wisdom bonuses against mental effects suggest that they should be dominating our missing Will save. 

[Edit: This last point assumes high Wisdom scores for priests, which isn’t always true and shouldn’t directly impact any discussion of the class’s ‘correct’ strength or weaknesses in any save. What should have been said is: The priest’s theme of being an incorruptible and fearless servant of their god seems like it should extend the priest’s divine protection to mental saves.]

The rogue, however, makes no sense; how is the class renowned for their dexterous feats going to completely beef it on the save usable for avoiding the explosion? The wizard is better at it! Why are rogues better at Petrification and Polymorph instead?

 

Problem 4. Rolling Over

AD&D saves are a roll-high system; as saving throws improve, the values become lower (to the maximum value of 2 for priests of 19th level). I don’t like it. Low values work for AC because of its floor of zero, as noted in a previous post, but saving throws have no floor; they’re just another form of check, and we’ve already got a great system for checks as used with abilities. I suspect that this is why the 5e evolution of the saving throw found itself so tightly wound around the ability score.

 

Problem 5. The Names are Too Specific

Pretty much what it says on the tin. If the save is called Petrification or Polymorph, you’re locking yourself into having that save reasonably affect only those two baleful effects. When you expand the category to cover any physical alteration, you dilute the meaning of title.

If you’re then using that save to avoid damage from slipping down a sloping wall (because that’s the save that rogues are good at), it becomes somewhat ridiculous. Why are we shoehorning dodging into a save designed for avoiding dragon’s breath? The one use of save vs Rod in the PHB is for an animal’s save against the ranger’s overtures.

These are the issues I have with saves in AD&D. The next step is to construct a revised saving throw ruleset that preserves the original structure, feel, and gameplay function while addressing those faults.