This is sort of like asking how physics and chemistry are used in manufacturing widgets... an impossibly broad question unless you provide some guidance about what exactly you're trying to do.
If you already have a profitable discretionary trading system, the math of it all is minimally relevant. You need to have a basic, broad-brush, practical (as opposed to theoretical) understanding of concepts like variance, probability and kurtosis/skewness. You need to understand how the various properties of your trading results, such as win rate and risk:reward ratio, interact to produce an equity curve over time. You need to have an awareness of 'tail risks' (unlikely but catastrophic events) and long-term theoretical vs. short-term realized outcomes.
Beyond that, unless you are already well-versed in the relevant mathematics and skilled at statistical analysis in other applications, I wouldn't bother with it. Successful discretionary traders are as a rule incorporating far more information at a far more nuanced level, as a function of training and experience, than any simple statistical study will reveal. You're far more likely to enhance your edge by working to further improve your existing strengths and skill set.
