ok so the trend line in green of beti appears to be sloping down,
but the beti is quoted to against zero, but would you not have to move the zero the average of the beti or some such?
tl;dr where is your 0 and why?
I made 2 charts. "Historical" and "Current".
I think zero is still zero and relevant in both of the charts.
"Historical" chart computes its regression parameters (slope and intercept) based on that day's parameters and all days preceding it. However as new regression becomes available, it never uses any future data to recompute the past.
"Current" chart re-computes all past regression parameters (slope and intercept) based on all knowledge to date using current data regression parameters. Past BETI values will change.
The latest BETI (todays) will always match on each chart.
Thereafter, the values will diverge.
My naming may be backwards or confusing. Perhaps a better name for "Historical" would have been "Instantaneous", and maybe I should call "Current", "Historical".
I guess its all the way you look at it.
I am not sure of the significance of the green line. I am no maths nor stats major.
Perhaps jl2012 will be back and chime in.
https://i.imgur.com/VlU5czu.pngI kinda think a chart plotted to the beti slope line and recalculated from that may provide an interesting insight and is perhaps more correct. It would seem to allow for the "beti" to be calculated vs the expected eg. Sort of like a tare for weight.
I am not sure though how historical and current can be both correct.....from the data.....
I suppose what I am saying is the historical green line is actually your x axis.....
