Perhaps the graph on the chart is based on the historic BETI values, calculated on each day based on a and b values of that particular day? Just guessing. To get the correct answer, we'll have to wait for jl2012.
Hey, I think you might be right. I did check this by taking the parameters from the very first post and applied them to the 25.-NOV-2913:
Screenshoot:
https://i.imgur.com/nbQADHd.pngThis way I get similar results as in the graph in
https://bitcointalk.org/index.php?topic=470453.msg26855063#msg26855063. It would be great if jl2012 could confirm this.
The question to ask imo is whether we should reallyhold on to the old BETI values (BETI values in the past, calculated with parameters of the past) in the graph, or whether the whole BETI graph should always be based on the newest parameters only. In my opition the BETI graph should only be based on the newest parameters, because the old regression curve was caculated using "less information" (=btc price action after the calculation in that point of time). Thus we need to recalculate the past everytime we get new a/b parameters, thus we need to completely recalculate the BETI graph overtime.
