Recent discussions on various Discord servers I belong to, including my own, have led me down the road of critically evaluating the Personae core mechanic, something I’m not sure has ever really been done since Personae’s inception but I know for a fact I haven’t done since I brought it into active development in the summer of 2022.
I wanted a concrete, qualitative method of determining what the probability of success looked like on a given challenge, and kept coming back to the realization that that wasn’t possible when the existing resolution method called for an opposed roll between player and GM: there was no baseline, not only to tell how easy or not it was to tell how you can succeed but also to see the change in a purely random roll versus one where a character’s attributes are factored in.
After consulting with colleagues it became clear that, given the nature of probability and independent vs dependent events, that the concept of a tie result, arranging the dice in descending order, and breaking ties with lesser results on dice beyond the highest, was moot: every event (in this particular instance, every individual die roll) has no impact on the sample space, or set of possible outcomes, of any other event, or any of the other dice rolled, whether you roll them at the same time or not.
On a roll of x dice, d10s in the case of Personae, if you determine success based on a fixed target, then that breaks down what success and failure looks like.
Sample space = set of all possible outcomes
Sample space of 1d10: {1, 2, 3, 4, 5, 6, 7, 8, 9, 10}
If the target is 6:
5 out of 10 possible outcomes fail, or 1/2, or 50%
5 out of 10 possible outcomes succeed, or 1/2, or 50%
You can turn the dial and make success less likely by adjusting the target:
If the target is 7, then
6 out of 10 possible outcomes fail {1, 2, 3, 4, 5, 6}, or 6/10, or 60%
4 out of 10 possible outcomes succeed, or 4/10, or 40%
The greater the target, the less likely success becomes, and more likely failure becomes.
This is, of course, all purely based on chance, before attributes are taken into consideration.
Going back to 1-5 fail, 6-10 success, an attribute value of 1 would turn a 5, a fail, into a 6, or a success. Now I succeed on 5-10: a 5 no longer fails, my set of possible outcomes that succeed are now 5-10, or 6/10, or 60%. The +1 increases my chances of success by 10%.
Someone with an attribute of value of 5, then, would succeed, given a target of 6, even if they rolled a 1.
I’m still fond of the space surrounding ties and breaking ties, but with a fixed target they now have to serve a different purpose. An attribute value of 5 would mean theoretically that I automatically succeed given a target of 6. In my ever-so-humble opinion, if an action is attempted there shouldn’t be an opportunity for automatic clear success, so I can effectively keep ties and say that they’re a marginal success (aka success with cost, consequence, partial success).
Outcome | |
|---|---|
1 | Failure (10%) |
2 | Failure (10%) |
3 | Failure (10%) |
4 | Failure (10%) |
5 | Failure (10%) |
6 | Marginal Success (10%) |
7 | Clear Success (10%) |
8 | Clear Success (10%) |
9 | Clear Success (10%) |
10 | Clear Success (10%) |
Given a target of 6, and an attribute value of 0, then I have a 50% chance of failure, 10% chance of marginal success, and 40% chance of clear success. An attribute value of 1 steers a clear failure at a result of 5 to a marginal success at a result of 6.
Additionally, since the rating of a skill means you roll more dice, this new core mechanic paradigm would mean you only need to succeed on one of your dice to have a clear success.
I’ve been toying with what unmodified, aka “natural” 1s and 10s would look like, especially given that up until now nat-10s have been the gateway to criticals, but I’m leaning away from nat-1s or nat-10s meaning anything special/separate anymore and instead criticals being when you succeed on two dice instead of just one.
