Oracular D20: Survey Results

We got 100 responses! This makes it very convenient for us, so let's work with it. Bitch.

Results & Analysis

Below are the five basic tiers or grades of results:

Grade	50C	67C	80C	90C
Terrible	1–3	1–3	1	1
Poor	4–8	5–8	5–7	—
Okay	9–13	10–12	10–12	11–12
Good	14–17	15–16	15–16	—
Excellent	18–20	19–20	20	20

On the table above, each 'xxC' column represents confidence of xx% over a certain interval. This means the 50C column indicates ranges with which at least 50% of people are confident; since the union of all ranges is from 1 to 20, this column covers all possible results of D20 and could be used as a table if you were nasty.

The 67C column has gaps at the inflection points. The following values are missing, meaning that they have less than 67% agreement: 4, 9, 13, 14, 17, 18. Some of these values just barely fall short of the cut-off; however, 4 and 9 are very relatively contentious with less than 60% agreement.

The 80C column is what I would refer to as the "exemplary" ranges: these numbers are agreed upon by at least 80% if not more survey-takers. This means that 1 is most agreeably terrible, 5–7 is most agreeably poor, 10–12 is most agreeably okay, 15–16 is most agreeably good, and 20 is most agreeably excellent. Everything in-between has some wiggle room, although they tend to gravitate towards the nearest exemplar.

The 90C column are super-exemplary, and not super helpful. Most everyone agrees that 1 is terrible, 20 is excellent, and 10–12 is just okay; I am surprised that the 'okay' range is not 10–11, and is instead 1 pip over the average for D20.

There were 2 survey-takers who interpreted the question as lower-is-better; did not count these, although their answers were basically the average in reverse.

Implications

Like I mentioned above, you could use the above table directly for D20, but it's not easy to remember and use on the fly. Since D&D 5e expects an average ability modifier of +1 or +2, and even greater with proficiency, I would grade each result by how many increments of 5 it surpasses:

Grade	Result
Terrible	1–4
Poor	5–9
Okay	10–14
Good	15–19
Excellent	20+

When you modify the roll by +2, the results of the incremental table mostly align with the 50C results while superficially resembling the 80C or 90C results. For example, although only results of 20+ are excellent, you have a 15% chance to roll 18–20 and add +2 to get a final result from 20–22.

I also think a table like this bridges the gap between binary and granular results on the D20, as well as between formal and improvisational interpretations of the D20. In the former respect, rather than ask what minimum result is necessary to succeed at a task, ask what is the minimum performance grade to succeed. Can someone just do "okay", or even "poorly", in order to accomplish what they want to accomplish, or must they aspire to do "well" (i.e., a final result of 'good')?

In the latter respect, we can read the grades as oracular results to improvise the outcome of an action. Maybe I don't have a specific number in mind, but I would narrate "okay" (10–14) differently than "good" (15–19)—the lesser result could also introduce complications or compromises. Not just that, a result of "okay" is definitely not good enough to (e.g.) perform open heart surgery.