Well here's a chart of the first 200 rolls from my sample of "weighted dice"
Rolls 1 - 200Nice! Thanks for this addition.
This plot matches much more closely with Niara's NWN, NWN2, and S:CotM that it does with the BG3 unweighted sample.
I don't know how to do a quick statistical analysis that checks the steak-iness of a random number generator...the best I can think of is: for each number rolled, tally up the occurrences of each number rolled immediately after it. Then check that that distribution is randomly distributed. However, this would probably require even more than 500 rolls for statistical significance and seems like it'd be very tedious to do, so I'm just going to rely on the visual differences between the graphs for now.
I don't suppose you also have a similar plot for your unweighted die rolls?
Edit: A month ago, I did the something similar for Niara's dataset, where I grouped 1-5, 6-10, 11-15, and 16-20. I found something like:
-The average roll following a roll of 1-5 was 9.8
-the average roll following a roll of 6-10 was 10.3
-the average roll following a roll of 11-15 was 10.8
-the average roll following a roll of 16-20 was 11.3
-All the above had errors of ~0.4-0.8ish, so were not conclusively different from the expected 10.5
