Note: My previous blogpost and the events that inspired it motivated me to put the final touches on this post which has been in the works for months. Enjoy.
The model for the modern core mechanic is
d20 + Bonuses >= Difficulty Class
That DC is static - like a law of physics. If a human wearing plate armor has AC 18, that's an objective truth whether they're being attacked by a level 1 fighter or a level 20 wizard. What differentiates the fighter from the wizard is the bonuses - adjustments calculated from their natural abilities and experience. This is also static and a pretty good model for a character as well as a world. You can calculate the mathematical probability of certain actions consistently. The other benefit is the intuition that higher numbers are better - both in terms of stats and rolling the dice.
So what's left to quibble about? Well, should plate mail actually be AC 18? What's the difference between an average human and a level 1 fighter trying to hit plate mail? According to D&D 5E it's 10% plus 5% for every 2 points of Strength above 10. But who decided that? You can see how it quickly gets tedious for the sake of "realism" in a fantastic world also filled with magic and dragons.
It quickly gets tedious for the sake of "realism" in a fantastic world also filled with magic and dragons.
You'll also notice ability bonuses compress the influence of stats. This was more true in Basic D&D where only the truly remarkable stats got bonuses of up to 15%. The difference between 10 and 11 Strength is negligible when it comes to combat, and wouldn't matter at all in 5E if not for carrying capacity. And yet stats can't improve too quickly or risk inflation.
The core insight of XDM and Dungeoncraft is all d20 rulesets are actually complicated ways to calculate an appropriate target number to roll on a d20. If an orc has AC 13, you need to roll a 13 to strike it with a weapon. But if you have +5 to-hit with your sword, you actually only need to roll an 8. Can we cut out the middle man and get straight to "roll an 8"? I think so.
Can we cut out the middle man and get straight to "roll an 8"?
The d20 produces nice 5% increments of probability and beautifully covers a 3-18 range of stats while maintaining possible, if unlikely outcomes at the extremes. It also suggests a natural system of progressive improvement. While there are myriad systems of progression that offer unique ways to dole out bonuses, I believe this ratio (as well as high-level magic) is why D&D breaks down beyond level 10.
So how could we preserve the differentiation of characters by their stats, reduce the math needed in the game, and get the visceral excitement of instantaneous tension and resolution? One option is fixed target numbers for the PCs, rather than the world. Let's look at 2 systems that do this: The Black Hack and XD20.
The Black Hack
The Black Hack uses the usual 6 stats generated by rolling 3d6 for a range of [3-18]. The core mechanic is to roll a d20 under the relevant stat to succeed. It may take some adjustment to prefer rolling 1s to 20s, but in this way higher stats are still better and there's no math. Simply roll the d20 and compare to your stat.
This roll-under mechanic also has the benefit of finer-grained resolution. Instead of a range of stats all mapping to the same modifier, each point of the stat matters and your stats will have a more pronounced effect over the course of the game. If we didn't want such diverse outcomes then why are we rolling 3d6? Unfortunately this also means an unlucky PC with a 3 will nearly always fail certain actions and will quickly stop attempting them. An extraordinarily fortunate PC with an 18 will almost never fail in that category of actions and is dis-incentivized to try anything else.
Thankfully these instances should be rare and a PC with an 18 maybe ought to feel near invincible. For the cursed player that rolled 3 there's hope. Stats trend upward over time as rolling above the stat on a d20 improves it by 1 when the PC gains a level. Thus poor stats should improve more quickly than high ones.
My main criticism is there doesn't seem to be any indication that the GM has license to adjust a stat, which is where we can examine the brilliance of XD20.
XD20
XD20 proposes 3 stats that roughly map to a character's physical, mental, and magical capabilities. These stats are generated by subtracting 1d8 from 15, producing a range of [7-14] each. These are your default target numbers and note that lower stats are better. To lift a boulder, break down a door, swing a sword, or wrestle an orc, roll a d20 vs your physical stat. The warrior may need to roll an 8. The fire mage will likely have a higher target - perhaps 14. That's it.
You might think the whole world seems flat if the warrior always has to roll an 8 to do physical things. Is lifting this larger boulder, swinging this magic sword, or wrestling an ogre also an 8? No. 8 is the baseline, the default for the warrior's average physical task. Yes, "average" is arbitrary but no more than any other codified d20 system out there. If the specific task attempted should be easier or more difficult, the GM should simply adjust the target number down or up.
If the specific task attempted should be easier or more difficult, simply adjust the target number down or up.
I'm tempted to recommend a rule of thumb for bonuses here. Due to the [7-14] range of stats/target numbers I'd recommend bonus adjustments of 2-4 and only 6+ in extreme cases or you might end up with target numbers outside of 1-20. But that's not the right way to think about this system. You don't want to go back to adding numbers to die rolls. You definitely don't want to be applying bonuses or penalties to stats as you'll get all sorts of confused (bonuses are subtracted and penalties are added? what???). What you want to do is pick a reasonable target number between 1-20 for this character attempting this task. You can start with the PC's relevant stat as a suggestion and modify up or down from there, but the goal is to pick a reasonable number that reflects the difficulty of the task on a d20.
Skip the math of bonuses and penalties. There are only 20 sides on the die. Pick a reasonable outcome and roll it.
Now for some, myself included, this feels a bit distant from D&D. Only 3 stats leaves little statistical nuance between PCs - and while there's something to be said for relying on the fiction to make them unique, I think we can make traditional fantasy role-players feel a bit more at home while still freeing them from the shackles of optimizing crunchy bonuses.
I'm not proposing a new system so much as a blended mental model. Let's use something closer to the traditional 6 stats in the range [3-18] by rolling 3d6 or 4d6 keep 3 depending how heroic you want to start out. But, we eliminate all the bonuses. The stats merely suggest baseline target numbers - something like the difference between the stat and 20. Strength 15? A typical strength-based task might be an unmodified DC 5 on the D20, adjust to situation and taste. Easy.
Situational Modifiers
Where's my Armor Class? How does cover work? Relax. Again, the above is merely the baseline. If you feel the DM hasn't accounted for something, point it out. Respect their decision, but negotiate a little. Your armor should make you tougher to hit, at least against certain types of attacks. The same is true for cover, shield spells, blessings, or any other fictional nuance that might affect the game. Trust your DM to make a fair call.
How do PCs improve relative to themselves? Relative to the world?
When they level up, perhaps PCs roll a d20 for each stat. If they roll above it, that stat improves by 1. If that's too much, perhaps they only roll 3d20, or maybe only 1 and have to choose which stat(s) they're attempting to improve.
Another option is for the PCs to simply write their level on their sheet, but gain no mechanical benefits. One of the core principles of XDM is that leveling up actually ruins the game. By allowing the player this lever to negotiate with the DM that, due to their level, perhaps the given target number ought to be lower, they feel a sense of progress without breaking the power curve of the game.
Conclusion
This is my new experimental ruleset of choice. Still feels like D&D due to the core 6 stats generated by 3d6. Rely on the DM to set reasonable target numbers based on that stat and what the PC is attempting. No laundry list of skills, crunchy bonuses, or maths. Easy to roll up PCs. Easy to run. Easy to play.
Next Up
Classes? Spells? Equipment? Some players need a list of suggestions to choose from. Even if it's not comprehensive, even if I don't write any rules for them. Being a ranger won't get you any mechanical benefits, but the archetype helps inform the gameplay. We may decide that you're good at hunting with a bow, tracking in the woods, and identifying plants. You'll get easier target numbers for those activities. We may steal spell names from OD&D, B/X, or even 5E. But what they do depends on what you attempt and how the dice say it goes. Also we need standard equipment kits to speed the shopping.
