Check system update
Posted by Tehom on 11/17/21
I'm using the return of actions to start to give a trial run to an update to how checks are made. It's a little bit of a departure from how things are done currently, and it's to address a few shortcomings that I didn't like.
Our first iteration of checks used the idea of aggregating dice based upon stats/skills for a given check, keeping the highest. The main issue with this system is that the more dice were in play, the smaller the variations were in outcomes: higher dice totals became very deterministic, and made the difference in keep dice totals feel very stark, which created balance issues in that trying to challenge someone with a given dice total would prove to be an impossible challenge for someone with a relatively small difference between them. Outcomes were only numeric, which forced GMs to make off-the-cuff determinations for what they represented.
The next iteration, which we're mostly using right now, derived values from stats/skills that were combined to give you a base value in a single roll, which was then converted into a qualitative outcome. The benefit of a qualitatie success level is that it provided more guidance to GMs for how to interpret a result. The shortcomings of this system is that it supported a relatively narrow range of stats/skills vs set difficulties and that there wasn't an easy way to represent modifiers in a graceful way. The flat mapping of difficulties to different ratings was easy to understand, but left us unable to represent granularity in different types of threats in any meaningful way. Additionally, there was issues with the single roll with modifiers could make higher skills feel unsatisfying, and streaks could make skills/stats appear to have little impact.
This next iteration is intended to address the inability to factor in arbitrary modifiers and support a much higher scope of threats to support growth in character power as magic is introduced. The short version of how it works is this: You add any number of things together that can go into a roll, all of them converted into different point values. When you have that total, that's converted into a 'rank' for your roll. You do the same thing with what you're rolling against - a difficulty or opponent. You then compare these two ranks (roller vs target), and the difference between them becomes a table that you roll 1-100 on which has different outcomes. The higher your rank vs your target, the better your outcomes: if you're 8 ranks higher than your target difficulty, it all becomes different kinds of successes, for example. The inverse is then true if you're lower. You can already see some of this in the check command via 'check/consider' and 'check/view', which lets you see your odds of succeeding a few different kinds of checks that have been shifted over. I plan to add something similar to actions to let you see the exact difficulty table you'll get for the target rank GMs assign, so you can see exactly how likely you are to get a given outcome.
Our first iteration of checks used the idea of aggregating dice based upon stats/skills for a given check, keeping the highest. The main issue with this system is that the more dice were in play, the smaller the variations were in outcomes: higher dice totals became very deterministic, and made the difference in keep dice totals feel very stark, which created balance issues in that trying to challenge someone with a given dice total would prove to be an impossible challenge for someone with a relatively small difference between them. Outcomes were only numeric, which forced GMs to make off-the-cuff determinations for what they represented.
The next iteration, which we're mostly using right now, derived values from stats/skills that were combined to give you a base value in a single roll, which was then converted into a qualitative outcome. The benefit of a qualitatie success level is that it provided more guidance to GMs for how to interpret a result. The shortcomings of this system is that it supported a relatively narrow range of stats/skills vs set difficulties and that there wasn't an easy way to represent modifiers in a graceful way. The flat mapping of difficulties to different ratings was easy to understand, but left us unable to represent granularity in different types of threats in any meaningful way. Additionally, there was issues with the single roll with modifiers could make higher skills feel unsatisfying, and streaks could make skills/stats appear to have little impact.
This next iteration is intended to address the inability to factor in arbitrary modifiers and support a much higher scope of threats to support growth in character power as magic is introduced. The short version of how it works is this: You add any number of things together that can go into a roll, all of them converted into different point values. When you have that total, that's converted into a 'rank' for your roll. You do the same thing with what you're rolling against - a difficulty or opponent. You then compare these two ranks (roller vs target), and the difference between them becomes a table that you roll 1-100 on which has different outcomes. The higher your rank vs your target, the better your outcomes: if you're 8 ranks higher than your target difficulty, it all becomes different kinds of successes, for example. The inverse is then true if you're lower. You can already see some of this in the check command via 'check/consider' and 'check/view', which lets you see your odds of succeeding a few different kinds of checks that have been shifted over. I plan to add something similar to actions to let you see the exact difficulty table you'll get for the target rank GMs assign, so you can see exactly how likely you are to get a given outcome.