Per Clucky’s CfJ, assume these facts are all true:

According to Josh, I had 24 power, Jumble had 23 power. This means, doing a little math, Kevan had 23 power and Pokes had 26 power.

This means, proportional to my heft (which, while reported as 17.6 was likely closer to 0.176470588%), I should have had a 24/96 chance of being selected, Kevan a 23/96, Jumble a 23/96, and Pokes a 26/96.

However, these are not the odds used. Pokes was instead given a 271/1000 chance, which while *close* to 26/96 is not quite 26/96. This means that an actually uniform probability distribution was not actually used, rather an approximation of one. And while it was a very good approximation, the rules do not permit approximations.

Make the resulting roll exactly match the odds as they should have been, in the following manner: Take the existing roll of 860, subtract one, and multiply it by 96. Add a DICE96 to this, so that the result has the same distribution as a DICE96000. Then, divide by 1000, and round up; this has the same distribution as a DICE96.

From this final result, on a 1-26 the winner will be Pokes, on a 27-50 the winner will be Clucky, on a 51-73 the winner will be Jumble, and on a 74-96 the winner will be Kevan.