The core of a role-playing gameâ€™s mechanics is the decision mechanic, what you do to decide whether an attempted action succeeds or fails. Itâ€™s certainly not the whole of the mechanics, but itâ€™s a very important part, as the rest of the mechanics have to be designed to work with it. That means that itâ€™s a good place to start the mechanical part of game design.
There are a lot of options for the decision mechanic, the most classic of which is â€œroll one or more dice, and try to get more or less than a certain number on the totalâ€. There are lots of variants on dice rolling, and mechanical systems that use cards, coins, counters, even piles of sticks. Iâ€™ve been back and forth on this a lot of times while thinking about this game, but for the moment Iâ€™ve decided to go with a random resolution method, and to use dice.
I decided I wanted a random mechanic because some of the best moments in my role-playing experience have come from unexpected die rolls. Unexpected success and failure add a lot to the story. In addition, given the general concept of this game, unexpected failure is unlikely to mean character death and the end of the story. Using dice as randomisers is largely a matter of convenience, but convenience is important in a central mechanic. I donâ€™t want the players to have to think about the mechanic; they should be thinking about the decisions their characters are making.
The mechanic Iâ€™ve chosen, at least for now, is the following:
Roll a certain number of d6s. Keep a certain number of them, and add up the scores of the dice you keep. Compare this total to a target difficulty. If your total equals or exceeds the difficulty, you succeed. If you donâ€™t, you fail.
The mechanic is based on six-sided dice because those are the easiest to find. I donâ€™t think that the advantages of dice with more sides, essentially a larger range of results, and so the possibility of finer-grained distinctions, outweigh the penalties to accessibility.
The main reason I like this mechanic is that it gives me three ways to change the probability of succeeding in a task, and the different interventions work differently.
First, you can change the difficulty. A higher number is harder, and although the complex probability curves this method produces make it difficult to say exactly how much harder a +3 makes a task, the direction is clear. I think this is a good place to make changes to reflect the objective difficulty of the task, rather than changes that depend on what the character is doing to approach it.
Second, you can change the number of dice kept. This changes the average result, but also changes the maximum and minimum. If you are keeping four dice, your result will be between four and twenty four. You canâ€™t get twenty five or higher, and you canâ€™t fail against a difficulty of 4 or less. This is a good way to reflect skill, a characterâ€™s learned abilities. As a character gets better, their best possible results improve sharply. Their worst performances also improve, but less sharply, as anyone can lose concentration for a moment. The average also improves, and performance gets more reliable; the more dice you add, the tighter the distribution around the average gets.
Finally, you can change the number of dice rolled. This has no effect on the maximum or minimum, but moves the average. It can move the average a long way towards the maximum, if the number of dice rolled is significantly higher than the number kept. This is a good way to reflect natural talent, and things the character does to help with the task. A character still canâ€™t do better than his best, but if he sets the situation up right, he is likely to perform quite close to his peak.
This means that training beats talent and preparation for one task. A character rolling and keeping two dice has an average result of seven, so can expect to beat a character keeping one die, no matter how many dice the latter has to roll. I like this result; talent should make a difference, but someone with basic training should not suddenly become able to beat masters just because of natural ability.
There is an obvious complication. To get the full benefit of these different ways of changing the numbers, they all need to be independent. Thatâ€™s not a problem for the difficulty, but there is a potential problem for the number of dice rolled. What if itâ€™s less than the number you are supposed to keep? If you just keep all of them, some of the benefit of skill is lost. If you roll at least as many as you are to keep, some of the benefit of talent is lost.
So, bearing in mind that the intent of the rules should be to make this a rare situation, Iâ€™m thinking about the following. When you have fewer dice to roll than keep, add enough dice to your rolling number (probably to be called your â€œpoolâ€) so that itâ€™s as many over the number you will keep as it used to be under. Then keep the worst dice you roll. So, if you have two dice to roll, and would keep three, you should roll four dice and keep the lowest three. If you have one die to roll and keep four, you should roll seven dice, and keep the lowest four.
This leaves the range as it should be for your level of ability, while biasing the average down. This feels like the right result to me, but the rule is a bit complicated. I think it might be harder to explain than it is to use in practice, but itâ€™s something to watch in development.
The other thing to watch is the number of dice being rolled and kept. Adding numbers between one and six is not difficult, but if youâ€™re rolling thirty dice and keeping fifteen, itâ€™s going to take a while. Iâ€™m going to aim for no more than ten dice rolled or kept in most situations, but I donâ€™t want an absolute limit. Again, this will be something to watch in development.
Finally, there is the possibility of exceptionally good or exceptionally bad results. There are obvious mechanics for this; if you keep all 1s, you get an exceptionally bad result, while you get an exceptionally good result if you have multiple 6s in your kept dice. It shouldnâ€™t be â€œall 6sâ€, because that makes a exceptionally good result harder for characters who keep more dice. On the other hand, if you keep four dice, you have a better change of having two 6s. However, Iâ€™m not sure whether I want such a mechanic yet. Itâ€™s easy to add during development if I decide it would enhance the game, so for now Iâ€™ll just leave it as a possibility.