Impossible to say until I get dynamically scaled parameters. If you are asking for my opinion, it is somewhere around 1400. (I made the changes by hand and this number represents that. But they are still hardwired.)Quote from shortie:
what's your best guess for FV "normalized by hand" right now?
Quote from nitro:
All trading above high frequency is estimating with a relative-to-time-frame error-term. If you want a tight error, go to the high frequency. That is a mathematical limitation,similar to the Nyquist Shannon sampling theorem:
http://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem