Devised an amusing dice mechanic. You roll a pool of dice (d6s currently) based on your competency (3, 4, 5 or 6), and the objective is to get a pair of dice with the same face-up value, and that value to be above the target number.
For example, Bob tries to pick the lock, which has a target number of 3. He's fairly competent, so gets 5 dice. His odds of succeeding? About 67%. He rolls 1, 2, 5, 5 with his dice, giving him a pair of 5s, a success (one beyond the target number).
I had to write a calculator in Python for the unusual success calculation method in order to get an idea of the probabilities of things happening. I've set it up to have changeable values for dice pool size and dice faces, and to be able to give you probabilities for one or all target numbers given those values. It gives percentages, giant fraction and a bar of hyphens to show probabilities, handy for getting an idea of any curve properties across different possible TNs.
Sample output:
For rolling 5 6-sided dice:
Target Chance
1+ 90% |---------------------------------------------.....| 7056 / 7776
2+ 80% |----------------------------------------..........| 6230 / 7776
3+ 67% |---------------------------------.................| 5264 / 7776
4+ 53% |--------------------------........................| 4158 / 7776
5+ 37% |------------------................................| 2912 / 7776
6+ 19% |---------.........................................| 1526 / 7776
I really like this. I feel I've done something productive and a bit creative, even though it's simple and common stuff.
One of the cool things I envisage doing with this system is giving certain opposed rolls an element of betting by players based on their rolls. For example, players representing two combatants fighting with swords each roll their dice pool, hiding it from the other combatant's player. They then decide what action they want to declare with the result they have. Certain actions might favour results that far exceed those of opponents, like powerful strikes; others might be more suitable for a result you have little confidence in, like blocking; some characters may even be able to exploit a terrible roll and make it a success by feinting, if the opponent rolls much much better and tries to strike.
Additionally, players can increase how often and well they can succeed with mechanical traits (or decrease how often with mechanical flaws). For example, a composed character might apply a trait to their lockpicking skills, so that if they roll and get no pairs, they can treat their result as a pair with value of the lowest die they rolled. A perfectionist might apply a flaw to their katana skill, so that they cannot declare any sort of attack with a result of less than a pair of 5s.
Characters supercharged with chance might be able to get traits allowing them to reroll some dice after they have been revealed to an opponent in an opposed check, to always have a chance of shock success. Some traits would allow you to do special things post-reveal whenever you roll a 6. Some flaws would force you to do certain things post-reveal whenever you roll a 1. The possibilities are extensive and exciting.
tl;dr: cool potential with this new dice mechanic, especially for combat/similar things, and I programmed something which makes me happy.
