I suggest you are mixing up your maths. The bell curve is a FREQUENCY DISTRIBUTION of something countable. It cannot be used as a tool to determine the nature of the countable things.

For example, take 3 people (Gawyn, Galad, Elayne) with IQs respectively of 80, 95, 140 (sd = 15, mean = 110). All we can say is that Gawyn has a higher IQ than 2% of the population, Galad has a higher IQ than 16% of the population and Elayne has a higher IQ than 84% of the population.

We cannot say what IQ means by contemplating the standardised curve. We cannot say that Gawyn and Galad are (together) smarter than Elayne. We cannot if IQ of 80 is sufficient to play stones. If we know these things, we know for other reasons.

To "solve" the power issue you simply need to construct formulae and solve. Given that we rely on terms such as "nearly", "twice" and so on, there is a big margin to play with. Using a scale is a good idea, however it should be linear. Standardising to 0-100 adds complexity, suggests log or exp values, and may be source of your misplaced use of the bell curve.

Once you had formula and data sorted out, then you could use normal distribution to say things. For example, you might say "gee, there is a real lack of strong women from Seanchan. What's the chance this occurs randomly? 0.001%! There is something odd afoot, I think."

Is there a list anywhere of every RJ quote on relative strengths? I'd love to look thru.