Designing a core dice mechanic for a TTRPG: Some considerations

If you're an enthusiast designing your own core dice-rolling mechanic, we've composed a short list of things to consider which may not be immediately obvious to the hobbyist designer.

1. Dice accessibility. Games tailored to gamers who are already "in the hobby" can assume players already have access to polyhedral dice. Games meant for children or people new to TTRPGs can't, and so may need to either forego polyhedral dice or offer a physical product containing them. Reliance on highly accessible modern solutions (in the form of dice-rolling apps and sites) needs to be weighed against the tactile draw of physical dice-rolling. Which brings us to:

2. Dice feel. It's satisfying to roll "bigger" dice (in terms of number of sides). Few people like the d4. It's very satisfying to roll a big handful of dice. There are downsides, though:

3. Ease of arithmetic. Counting dice results which meet some criterion is easier than adding numbers, which is easier than subtracting numbers, which is easier than multiplying numbers. A dice-based table lookup can be very easy or very hard depending on the number of tables in the game and where and how they are presented. All of this stands in opposition to dice feel: the more dice you roll, the harder the arithmetic and the slower the gameplay will be. The same goes for rolling "bigger" dice: arithmetic is substantially quicker when it only uses single-digit numbers. And the point where you need to switch from mental arithmetic to using a calculator is another breakpoint for fun. This is related to:

4. Static and dynamic dice modifiers.

• Dynamic dice modifiers are those where different situations and contexts can affect the dice being thrown, the arithmetic operations being done to them, and/or the numbers they are being compared to. The advantage of dynamic dice modifiers is that they reflect how we naturally think of difficulties and probabilities and so increase, if not realism, then at least verisimilitude. The disadvantage is that they impose additional arithmetic, and so extra mental load and time requirements.
• Static dice modifiers are usually things intrinsic to a character or task and do not change. The advantage of static dice modifiers is that looking up a number is generally quicker than calculating one, and if there are a small number of character-specific static modifiers, then players will begin to just remember them, which is even better. The disadvantage is that there are few things about a character in the world that can't change under any circumstance, and so a game using static dice modifiers must either (a) have very few of them, or (b) lack verisimilitude, or (c) have supplementary rules for the edge cases when static modifiers should change.
• This is related to the whole question of...
  

5. How much precalculation will the game employ? More complex TTRPGs typically have character mechanics/statistics that affect other character mechanics/statistics. For example, equipment affects encumbrance which affects speed; or ability scores affect modifiers which affect defences; or training affects skills which affect subskills. There is a spectrum from "never precalculate anything and only work out what's what when it comes up" to "precalculate everything but make sure you remember to recalculate what you need when it comes up", and in practice almost all TTRPGs are somewhere in the middle. Ideally, precalculation increases time costs between games in exchange for reducing time costs during games. Of course, if a number is expected to change a lot based on circumstances, precalculating it is both a waste of time and risks diminishing the gameplay experience when someone accidentally uses a default value when it should have been recalculated. Precalculation is part of the problem of...

6. Dice roll management. The number of distinct dice rolls used to adjudicate a task (activity, problem, situation, etcetera) can vary hugely depending not only on the overall complexity of the dice mechanics, but also where the game chooses to "invest" more of that complexity. Using dice to adjudicate tasks (etc) should feel satisfying. Dice shouldn't, generally, be rolled for unimportant tasks (etc). A good (well-designed and well-explained) rule system minimises the amount of back-and-forth between player and game master to find out what to roll, what number is needed, what modifiers to apply, and so on. A good system makes it easier to acquire system mastery. Deciding when, how much, and in what way to roll is a matter of...

7. Overall dice mechanic design.

• This is really what this list is all about. Game designers often think their game needs a truly novel dice mechanic to succeed, and the first point to make is that they don't. 1d20+modifier vs target number is fine. 3d6 vs task difficulty is fine. Nd6 (where N is character skill) and count successes (where the task defines a success) is fine.
• ...But a truly novel dice mechanic can be cool, as long as it's not cumbersome and fits with the game tone.
  
• Rivers of digital ink have been spilled on the difference between various core dice rolling mechanics. In short, a single dice roll produces a flat probability curve (N outcomes, all equally likely); rolling and adding multiple dice approximates a Gaussian probability curve (N outcomes, with the central outcomes most likely and the extreme outcomes least likely); other mechanics like dice pools produce stranger (sometimes asymmetric) distributions. All of them "feel" different and are more appropriate to different game tones. Modifiers to the dice roll and the target number "matter more" for systems with a probability curve, at least when a task (or activity, etc) is very easy (or likely, etc) or very difficult (or unlikely, etc).
• Then there are the decisions about how to actually use the dice roll(s). Different audiences have different tolerances for arithmetic, table look-ups, and non-dice gimmicks (card houses, candles, blindfolds, knucklebones, dice swapping, etcetera). This too is something that ought to match the game tone.
• There are all sorts of design choices to be made here. Here's just a few possibilities spat out at random:
• You could make all modifiers the same magnitude, so nobody has to remember them or look them up ...but in some contexts that may be impractical or damage verisimilitude.
• You could design the system so that a high roll is always better for the players, or so that a high roll is always better for the "side" rolling the dice ...but certain dice systems might genuinely work better when aim-high and aim-low are used in different situations.
• You could design everything so that dice rolls are always in opposition to other dice rolls, whether that's another character or the environment, producing a pleasing symmetry and consistency ...but in games where almost all problems pitch characters against the environment, do you really gain that much?
  
• You could have modifiers apply to the dice target number, or to the dice roll itself, or both. A common approach is for modifiers related to the character get applied to the dice roll and modifiers related to the environment (or target) get applied to the target number ...but this is far from the only way of doing things!

Hope you found this interesting. Good luck and have fun. At Periapt Games we 💜 hobbyist game design and amateur game designers.