Interesting. Suggested improvements:
(1) Calculate the deviation from the weekly or monthly trendline. If bitcoin prices are generally going up, you may end up with a false result that days early in the week are better because they are before days late in the week.
(2) Calculate the standard deviation for each day to see if the result is statistically significant (in principle you need to do a t-test, if I remember correctly, but calculating the standard deviation will be a good indicator)
Hmm interesting angle, will check this out later
@datageek, did you follow up on this idea yet?
Not necessarily so. You only get the effect that the last day of the week is best if you happen to calculate the results always at the end of the week and if you calculate over several years that effect will be small.
After all, the 'end of the week' is an arbitrary choice since days are cyclical.