RPG Design: Of dice…

I love dice… have been loving them since I started roleplaying. I love different sizes and shapes and I love the thrill of excitement that random die rolls bring with them. While I’m also enchanted by diceless systems, none have ever worked for me because the lack of randomness makes me feel like once more reading a story… it’s no longer a game. And I want to game. So dice it must be. But dice mechanics and choices shape a game more than any other aspect I can think of – for many reasons. Here is the dice system I will be using for my latest RPG design outlined in my recent blog post.

I strongly believe that dice mechanics are one of the most important decisions to make for a game design. Dice must represent the feel aspired by the game. They must reflect the reality of the game in some way. Therefor I also believe that there is no one universally perfect dice system covering all genres… some games demand for minute details so I will need a rather wide spread of dice results… other games paint a story in bright and vivid colors, so a small result scale is perfectly fine. The most important design goals for my current project include the following requirements:

• supports all styles of play from competetive to storytelling modes
• uses a simple and streamlined rules engine
• is fast in play

The most important decisions IMHO usually include:

• the amount of dice to use
• the type(s) of dice to use
• the way to roll (open ended, closed)
• the way of reading the die results (e.g. adding, multiplying, comparing to thresholds, counting specific numbers)
• the way the dice to be rolled are determined

I will try to explain the choices for my latest design step by step.

One of my requirements was to have a system that is fast in play. This immediately implies that few dice should be rolled because rolling larger number of dice always means that you need more time to determine the results, no matter how simple actual determination might be. This heavily reflects on the game design.

Take the example of Shadowrun, a game I love dearly and have played extensively in its first and second editions. Shadowrun is a fast and furious cyberpunk/fantasy game – and it uses one of the most inappropriate dice systems I would be able to think up. It requires you to roll buttloads of six sided dice in order to do anything. A typical combat action results in a number of dice being rolled based upon your skill + modifiers. These are compared to a difficulty score (with an open-ended roll system), than the successes are checked against your weapon damage code to determine actual damage potential. Next the target rolls a number of dice to determine damage resistance taking away from your successes and reducing damage (again open-ended). This easily results in 10-30d6 being rolled for one gunshot. Terrible. No sense of speed or drama. Just dice flying all over the table. While this is fun once or twice it gets annoying real quick when you try to stage a dramatic fight. And don’t get me started on Matrix combat, where the number of dice explodes… for a simulation of actions that occurr in milliseconds.

So die pool systems almost never work for me – not even in Storyteller. Come on, how great is the idea of rolling xd10 (with a system somewhat reminiscient of Shadowrun) in a game focussing on a story. They should have gone with 1d10 or 2d10 and ignored dice probabilities (especially since probabilities never really worked out in the early editions).

After having eliminated die pools the next question is how to determine results. Adding is faster than subtracting. Subtracting is faster than multiplying. Multiplying is faster than dividing. Just looking at a number is even faster. Therefor I will go with a system that will require you to roll a certain amount of dice, add your modifiers and compare that result to a threshold (not much mathemagics here). Higher will always be better since this is very intuitive.

And one of the biggest mistakes many games make… take e.g. Kenzer & Company and their Aces & Eights game, a game I had been looking forward to for a long time (and a company I love for their KODT comic strips and Hackmaster). The die size is reduced for higher skills, so 1d4 is better than 1d10. Mmmhhh… I hate it. Counter-intuitive and really strange. 

Since I want to be fast I will roll but one dice. Since I want to have  a wide range of mechanical character evolution options (another requirement for my design) I want to have a wider range of numbers. The best option thus is to roll a d20 for a tests of skill and luck since it’s one simple and easy roll with a decent spread of numbers.

Wow… I just invented the d20 system. Ain’t I a great designer 🙂 ? But there’s more to do.

I also decided to go for a very limited range of dice types – I don’t need all the polyhedral dice in my game (although I love them). All too often I had to watch newbies mistaking a d10 for a d12 and vice versa. Or a d20 and a d12. Even after three or four game sessions. So away with that nonsense – I want my game to be newbie friendly and thus decide to use two die types at most. I decide to use a different die type for combat (damage specifically, and also for some other secondary attributes – more of that in a latter blog post). Inspired by Spellcraft & Swordplay all weapons in the game simply will cause 1d6 damage – the reasoning being that skill is more important than weapon type and a dagger can be as deadly as a twohanded sword. Additionally it takes some of the optimization pains from players by allowing them to select the weapon(s) most appropriate for their character design. There will be some bells and whistles to this – more on that in later posts.

To accomodate my taste for drama and high fantasy I will add one minor detail: All die rolls are open ended – whenever you roll the highest possible result on a die you may reroll that die and keep adding the results in order to increase the final total. This adds both the potential to succeed despite unlikely odds (for tests) and adds drama to combat (because in theory damage can rise to incredible levels allowing for instant kills of anything).

So there we are:

• All die rolls in my game are open-ended (even during character generation). Keep rerolling and adding whenver you roll the maximum result for the die type at hand (20 on a d20, 6 on a d6).
• All tests use 1d20 + modifiers, comparing the result to some threshold. Higher is better and you must pass the threshold.
• All damage rolls use 1d6.

Clean, fast, simple and hopefully furious. Not very innovative but that’s not a goal I find particularly compelling.