The Evolution of the Saving Throw
A relatively short post this time. I have been away for a while! I haven’t been inactive, but I am a little unsure of what I want the blog to be. Do I post my unfinished notes, my musings, my designs-in-progress? If I limit myself to what’s complete and tested, it seems oddly piecemeal when my original goal was to talk theory more openly. My historical practice has been to workshop in private with my most interested player(s) who bounce ideas back and forth or else act as the metaphorical rubber duck.1 I have developed a lot over the last month, from structure rules to a mild overhaul of how I handle light, as well as a rework of my magical item system. I will certainly post about that last one soonish.
Anyways, the crisis of purpose aside. Today it is about saves. A brief history, first. Saving throws should have some differentiation between them. This, the editions of D&D have agreed upon.
In the earliest titles, and most of the retroclones spawned by them, saving throws were a line of stats unrelated to the character’s other ability scores. Classes progressed in the scores at rates determined by their class; the Fighter had nearly twice as good a Breath Weapons saving throw at level 13 than the Magic-User did.
In 3rd Edition, this was crunched into being indirectly based on a character’s ability scores. You now had only three scores: Fortitude, Reflex, Will. Each was modified by a single ability modifier. Differentiated class progression was retained in the basic scores.
In 5th Edition, it was crunched down further. Saving throws now closely resemble ordinary ability checks. They can be made against any of a character’s abilities, although certain types tend to be favored, I believe. If classes are differentiated, it is only by way of discrete special features or by having chosen different abilities to prioritize. A wizard will have naturally high Intelligence, in most cases, and therefore will have better odds on such saving throws.
I, being an iconoclast, borrowed bits from the GLOG upon my discovery of it but refused to adopt the saving throw as presented. I was dimly aware of the history, but having come primarily from a system without the concept (Dead Simple RPG2 — a very simple free d20 clone made for homebrewing), I wanted to replicate the differentiated saving throws of old while retaining my familiar and preferred system.
So, like a fool, I did two things, as an experiment:
I kept the saving-throws-based-on-ability-score/aspects that DSRPG had, which was likely derived from 5e, rather than keep the saving throw separate.
I split the GLOG’s proposed Save score into two (Health & Luck), differentiated them by domain, and then added it on top of the saving throw.
Taiao’s save system was therefore set up to be like this, which it has largely not changed from since then: all saving throws are TN 15. With a d20, that’s about a 30% chance of success. You get a modifier from the aspect the check calls for — say, Intelligence — and you may add your Health or Luck depending on what the check is for. Critically, Health/Luck are not normally specified until the GM calls out the check, as I found it annoying to have to specify two differentiating features; I probably should have noticed what this meant sooner.
It has been approaching three years since I started this experiment. Here’s what I’ve discovered:
It is, in fact, relatively annoying to have to specify whether to use Health or Luck. The conditions are relatively clear (Health for temperature, disease, bodily function, etc and Luck for everything else) but if there’s the slightest room for doubt and someone has an imbalanced set of scores players have the incentive to ask every time anyways.
The scores are differentiated enough just by relying on a character’s aspect modifiers. Someone with +2 to +3 Intelligence is going to have an edge over someone with –1 Intelligence. That’s a 15% to 20% difference. Isn’t that fine enough?
The two changes I initially made essentially achieve the same goals. This wasn’t clear to me at the time, because I didn’t have a firm grasp on the purpose of a saving throw, but splitting your save bonuses and tying them to aspects is silly redundancy.
You might be saying “well, the outcomes should have been predictable!” And you’re probably right! I’m not the smartest, and I spent a few years just feeling out the design space, learning as I went. I’ve played a lot more games now, and I’m a bit better at predicting outcomes.
I’m considering this experiment closed: it is not great to have two separate saving throw modifiers if you aren’t going to specify which one to use when. I thought it’d be cute to be able to improve a character’s Health score through unique means i.e. magical sources of health, or similarly with Luck, but frankly, that can be simulated with off-hand abilities granting numerical modifiers or advantage.
For now, the saves will likely be crunched into one. Perhaps I’ll consider further methods of differentiation, or maybe this will be its final evolution. Personally, I anticipate that I’ll stick with discrete abilities where I feel it’s necessary. From history, players seem to appreciate a free +4 on a certain category of saves more than they appreciate a gradual progressive bonus.
The only question is now: what do I call it? I still like the theme that the term Luck provides, but I believe those who argue for the old saving throws have a valid point in that the ambiguity of the terms and their disconnect from ability provides some more flexibility to the imagination. Calling it Save retains a bit of that, even if it is still attached to an aspect score; calling it Luck sounds nice it but may reduce that flexibility.
Slowly, but surely, I will re-invent the wheel.
Of the rubber duck debugging method.
Dead Simple RPG’s homepage can be found at deadsimplerpg.wordpress.com.