Hi everyone, newb here.I'm involved in research & backtesting of several asset classes, most notably stocks & futures, with emphasis on the changes in Volume (real-time) & Open Interest (EOD) in the options markets for the various underlying instruments. For modeling distributions, the Empirical CDF appears most appropriate given the 'clustered' nature of both.The primary objective is to find the weighted aggregate mean strike value. With this type of distribution, calculating skew and kurtosis seems to be of little use.What do you most commonly use for modeling options supply/demand?

Hello outrun, thank you for your explanation. On an immediate/near-term basis I certainly agree, but from the perspective of modeling the volume and OI across the entire strike and expiry chain, I am looking to see where 'clusters' of activity are found. Considering that most OI at expiration is not exercised and hence expires worthless, it seems apparent that any seller of an option must pick the strike(s) and expiration(s) of their various strategies carefully.With all of the various flavors of option pricing models in active use within the markets, the various multi-legged strategies, and the varying expectations of participants, it is certainly much too thick of a 'layer cake' to view in a succinct manner, but on a rough, 'satisfice' basis, trying to find a volume- or OI-weighted mean strike serves to identify where likely inflections may occur. Consider the CME Group whitepaper "Profiling Option Open Interest"Consider the bar graphs on pages 1-2...what might be the best method to model those on a 'smoothed'/spline basis?