The reason why the Bitcoin power law estimate in 2013 (although not well-established statistically at that point) could end up close to correct as seen now is that it is Log Age that matters. That was age 4 plus a little. With a power law it is the doubling of age or the increase of age by some ratio that matters. For example with a β = 5.69 measured exponent (OLS no weighting) a 50% increase in age (1.5 multiple) projects a 10.0 x increase in trend price (coincidentally). So think in terms of ages as geometric series: e.g. 1, 2, 4, 8, 16 years and now we are closing in on 18. The next proportional age would be 32, the year 2041. Patience is in order to evaluate power law endurance and reap the benefits. Age 4 was already ~ halfway to Now in log terms. It turns out that even now due to high auto-correlation only 23 data pairs distributed evenly in log age are sufficient to accurately depict the power law nature. https://t.co/twQiH6VT1n