The NKR: My System for Building System

I'm of two minds regarding mechanical resolution systems.

On the one hand, people will find a way to have fun with anything. Have you ever seen people get excited and yell while playing Trouble? It's a terribly designed system, and people love it anyways. Likewise, there are a number of D&D 5e hacks that should really be using a different system, but people have a blast playing them anyways. Fun doesn't care about your facts. Do what you want!

On the other hand, systems can tremendously shape the way a game plays and feels, and I care a lot about them. After years of trying to accomplish dungeon crawls in D&D 5e, my first game of Into the Odd felt like heaven. The rules and the content are so crisp and they merge so harmoniously at the table. Turns out, all that OSR theory has an impact in play. Who knew?

So it's both the most important issue a TTRPG can resolve, and also literally doesn't matter. I have decided to treat it like it matters to satisfy the design freaks like myself, because the people who couldn't care less won't even notice (If you're my kind of nerd, you'll love this post. If not, feel free to skip to the end to see the solution I came to).

Thus, the core mechanics were the first thing I wrote for every TTRPG system I made. And that's fine. I like doing that kind of work! But it was a lot of effort to work through the same basic assumptions and come to slightly different conclusions every time. I wanted to come up with a common framework to build new systems off, some core mechanics that I could use as a starting point, knew intrinsically, and which would let me jump right to the interesting part of tailoring the system to the game. In pursuit of that, I overanalyzed every petty little detail of every game I know to come up with a big wish list.

I have tentatively named it the NKR (NathanKlas Ruleset, patent pending).

Big Number Good

This is basic as hell, but I feel very strongly in my lizard brain that bigger numbers are better. I understand the appeal of Roll Under system, and can admire the elegance of a 60% search skill showing your exact likelihood of success, but I simply would prefer to roll high, not low. Newbies intuitively agree.

So I'm looking for "higher skill number means better skill" AND "higher rolled number means better roll." But I also want to maintain the bounded accuracy you get from roll-under systems (e.g. you can't roll higher than 20 or lower than 1). No high level 3e/5e D&D shit where you're consistently rolling 30+ on a d20, and "easy," "medium," and "hard" have to scale up with you as you level like a fucking Bethesda game.

Ahem. Anyways.

You may wonder how I plan on making high rolls good, high stats good, and all rolls fit within a single boundary. Read on, gentle reader. This is only the beginning.

No Math, Only Dice

While gamers stereotypically love doing math, my players sure as hell don't. And each step of mental effort between rolling the dice and determining the result eats away at tension, engagement, and play time. More on that later.

You can do a lot with a coin flip, but you'll probably want some numbers to make things more or less likely, give characters strengths and weaknesses, and so on. Math is compulsory. But I want to keep it to basic addition and subtraction wherever possible. And I mean basic - even a +1 should make a meaningful difference to the roll, somewhere on the order of +5 or advantage in D&D 5e, or +20% in Call or Cthulhu. This also has the added benefit of focusing the design. If it won't make a big difference, it doesn't matter enough to simulate.

You know what's more fun than math? Dice! Envision the collective glee at the table when the Paladin smites on a crit, then starts going around collecting everyone's d8s for the big roll, and they feel great to shake and make an awesome noise when they roll. Fun! Then remember how you have to wait 20 seconds for them to count to 45, and how you probably doubled checked it yourself because no offense but you do not trust that player's counting skills. Less fun! I want the first part, but not the last part.

You may see where I'm going with this. But first, there are many more petty little things I want to talk about.

Maximum Impact

The best rolls have everyone leaning over the table to watch, then collectively exploding into cheers of groans the second the die lands. Not every roll can be this exciting, but we can work towards making more of them this exciting.

The first issue is post-roll processing. Rolling the die, adding a number, then asking the GM if it's high enough. Scanning your character sheet for inspiration, or luck, or a class feature, or a spell. Anything that's needed to determine or modify the results of the roll after it's already been rolled. It strains the tension and pulls the focus away from what happens next. I want as little of this as possible. Do it all before rolling, set the stakes, then pray to your dice god of choice. Whatever the die lands on, that's final.

The second issue is low stakes. When twelve goblins all attack the barbarian, seven hit, and they all do negligible damage, that's nineteen rolls (or thirty one if they all try to hide as a bonus action) that nobody cares about. That's when the phones come out. Here's the design objective, then: keep condensing until each roll is edge-of-your-seat important. The focus isn't on rolling as little as possible, not exactly. I just want the rolls to matter and to tell us everything we need to know all at once. So yes, condense the attack roll, damage roll, hit location roll, and spell mishap roll all into one. But don't condense it to zero.

One Universal System

How do we simulate that your fighter is better at swinging a sword than my wizard? You'd think that would be an easy question to answer, but it isn't necessarily. For example, in D&D 5e you could add 2 to your attack roll, decrease your opponent's AC by 2, add 1d6 to the roll, or give the roll advantage. It's not always clear which of these methods should be used in each situation, and choosing one over the other can have weird knock-on effects that result in different math in different situations, or which interact differently with class features.

And that's not even touching on all the special abilities that let you break the basic rules, like treating any roll of 7 or lower on an attack roll as an 8, letting you reroll natural 1s, letting you reroll three attack rolls per day, dealing damage even on a miss, and so on. It gets complicated fast, and leads to weird tactical decisions that have no bearing on the game world. Sometimes that's fine, because it's fun for your special little blorbo to break the rules in a way that no one else can. But when that is done, it should be done intentionally, with care. And it shouldn't be part of the core NKR.

Getting back to the original question, I want there to be only one way to simulate character skill, and for it to be the same answer every time, and for every other thing that would affect the odds of success (difficulty, help, enemy skill, buffs, situational factors , etc.) to all change it in the same way.

How do I do that? Drum roll please...

The Solution

Roll on a table, so that the number on the die tells you how well you did. For example, if the base roll is d20, then a 1 can be a crit fail, 2-9 a failure, 10-19 a success, and 20 a crit success. Before rolling, sum up all the modifiers to your roll, both positive and negative, then roll that many more dice and take the highest (or lowest if the modifier was negative).

It's just advantage! Instead of adding +4 to attack rolls, you add 1 bonus die. Instead of having to beat a DC 20 to hit your opponent, you subtract 2 penalty dice. Your friend's help adds 1, and the slime which reduces the opponent's speed adds 1 to your roll as well. That nets to +1 (1 - 2 + 1 + 1), so roll 1 additional die and take the highest.

It doesn't sound like much, but that's true of every elegant design. I've found this to be easy to design around, intuitive for players to pick up and play, and good at keeping the momentum flowing. It does what I want so well that I'm really not interested in designing any other way.

The only issue I've found is that people confuse the modifier for the number of dice they need to roll. So, in the previous example, with a net +1 modifier, they need to roll two dice. People pick this up after a few rolls, but I wish it were more intuitive. I've considered really stressing the difference between the main roll and the modifier dice, or making the modifier different from the main die (e.g. base 1d8, modifiers are d6s). More playtesting is required to smooth this out.

This works with any base die roll, by the way. 1d6, 2d6, 1d20, 1d4 + 1d8. The sky's the limit. You can pick the dice that work best for the game you're telling. Maybe d20s just feel right for your fantasy heartbreaker. I prefer d6s for silly simple games so they can crit on a 1 or 6, and for more tactical games that require adding up lots of modifiers for each roll. For a cowboy game, I made the base roll 2d6 almost exclusively so that I could call a crit fail snake eyes and a crit success a bullseye. I also worked it so that the attack roll also provides the damage (highest of 2d6) and hit location (lowest of 2d), culminating in two 6s being a bullseye that hits the heart for an instant kill.

It also leaves open exactly what the core table is. The big difference is usually A) how likely crits are, and B) if it's a binary pass/fail system or a nuanced pass/complication/fail system. Again, you can choose what works best for the game you're trying to make. A story game will have different requirements than an old-school dungeon crawler. But they can both use this system. The NKR isn't a game in itself, it's just a solid core that you can start with and build a game around.

Other Features of the NKR

The math is fun, in that it's intuitive but fuzzy and yields diminishing returns. Rolling with advantage feels great, and rolling with disadvantage feels scary. That communicates the odds of success easily, and without any number crunching. Diminishing returns mean that huge bonuses stop meaningfully increasing the odds of success after around +3 (though they continue to increase the odds of a crit for a lot longer), so players aren't as incentivized to get +100 to each roll.

This also leads to flexible action resolution, where you can choose to take penalties to get more bang for your buck on a success. For example, if I have +4 to hit, I might choose to reduce that to a +2 to deal extra damage (-1) and disarm my opponent (-1) in the process. In this way, the complexity is modular, both at the design level (add in lots of different ways to gain bonuses or penalties in a tactical sim) and at the personal level (I can get creative and push my luck, or be lazy and take the easy win. Neither one is necessarily the 'correct' choice).

Combining rolls has been a lot of fun, and I think could go a long way in OSR circles. Of course, I have experimented with combining the attack roll and the damage roll. But even further than that, you can combine two opposing rolls into one (two guys attacking each other make one roll to see who hits, and damage is automatically applied) or combine multiple rolls into one (twelve goblins ganging up on you is still just one roll, but their attack modifier adds like +4 or something). The specifics will be up to each game, but the potential is huge.

You can also avoid rolling at all, and just use the net modifier to determine what happens. Say positive is a success, negative is a failure, net zero is a mixed result, and +3/-3 is a crit. You could use this to eliminate rolling entirely if that's your style! I prefer to use it in place of some GM rolls (e.g. NPCs fighting NPCs, or to automatically decide if an NPC believes a PC's lie) and to simplify multi-roll procedures (e.g. if the attack roll hits, the damage roll is automatic). Speaking of...

While you can adapt this system for traditional hit points, I have avoided doing so in all my NKR games so far. HP is a relic and assumed part of many games where I think it simply doesn't belong. Instead, you can get rid of the damage roll entirely (e.g. the winner of a fight decides the fate of the loser), or you can resolve a damage roll using the typical system resolution (e.g. 1 is no harm, 2-9 they're injured, 10-19 they're downed, 20 they're dead). Which approach is best will depend heavily on the system being made, but having to think through it for each game has lead to much more interesting, genre-fulfilling systems than traditional HP.

That also gets a bit into my desire to remove meta-currencies like HP entirely, but this post has gone on long enough, so you will have to extrapolate for yourself from here.

Concluding Thoughts

I like this. Feel free to steal it if you want, just credit me if you do. And let me know so I can check it out!