Statistics theories in forex trading

Hi guys,

Anyway I came across chebyshev's theorem during my statistic lessons and it suggests that for all data regardless of distribution, (1-1/k^2)*100% would be within k standard deviations regardless of any distribution.

I was wondering if this law applies to time series data (such as price), and how it can be used. Because bollinger bands only sample from the latest x number of bars, the sampled data always changes with every tick/period. In that case, how can we utilise this law to gain an edge in terms of expected value and expected losses (regression of price within a defined number of standard deviations)?

P.S I have just started on my statistic classes so pls be patient haha!

My own personal view on statistical measures in the FOREX market is that it is a misguided effort as you are not dealing with a "price" but a RATE. So unlike stocks where you an measure a company in terms of historical performance and future expected performance combined with volatility of returns the FOREX market is a two sided coin where you are dealing with a pair which will likely counter such analysis.

A rate is different to a price I guess is my point. The Forex market is unique in this way. Warren Buffet himself lost millions or billions in this market. In my view fundamentally its a different game, although in terms of liquidity and market mechanics much is the same.