Quote from: rpietila on April 01, 2014, 11:06:18 AM

The beauty of the 5.25-year trendline with exponential fit (R^2=0.93 which is pretty darn good) is that it takes into account every worry of every person who has ever owned or not owned bitcoins. I put more weight on that than the individual worries of a single person.

You put a lot of weight in your humongous pride, as if someone presenting analytical discussion is worried. I am not worried about the BTC price. I could careless because I don't own any BTC nor am I itching to buy any. And you put a lot of weight in arbitrary curve fits.

And people who think they know every thing for sure (and you don't even enumerate the arbitrary assumptions in your model), eventually get a lesson in respecting chaos. Maybe not this time. No one knows. But eventually yes.

I hope you realize the following chart is arbitrary BS.

Quote from: rpietila on April 01, 2014, 11:12:19 PM

Quote from: smoothie on April 01, 2014, 10:45:52 PM

Which graph can you point me to for the 5.25-year exponential trend line?

Code:

http://i.imgur.com/ycT9ulP.png

And why not drawn like this? A least squares fit is an arbitrary choice of slope, because the curve you are fitting is not very linear. The 2011 outlier is your big problem with choosing this arbitrary fit. And we can't really trust that early data back in 2010.

The green line looks much more accurate to me. It removes that bubble from 2013 and follows a trendline from before the bubble started. Or we stay with the purple trendline, in which case the price is almost down to the purple line.

You need to update your chart. The price is now lower than shown.

Also refer to my upthread post quoted below pointing out that adoption (and thus BTC ≅ adoption^2) is likely log-logistic, not logistic. Thus we see the first slope to 2011 was higher, then the slope from 2012 to 2014 was the purple line, and now we may be ready to transition into an even lower slope, such as the green line.

Quote from: AnonyMint on March 31, 2014, 12:01:11 PM

Quote from: rpietila on March 31, 2014, 10:18:27 AM

Quote from: AnonyMint on March 31, 2014, 09:00:36 AM

We should now shift into yet again a lower adoption slope than from July 2011 to December 2013.

What a heck makes you think so at a face of universal awareness that is just achieved?

Math Risto. You can't deny math. Here is the shocking revelation...

Because if you put a ruler on the chart along the bottoms of unique addresses on the log 10 chart, you see the slope was higher before the July 2011 crash, than it has been since 2012.

Also it is very likely that Bitcoin adoption is not logistic where the maximum rate of adoption is at 50% of the adoption as follows:

http://en.wikipedia.org/wiki/Diffusion_of_innovations

Because Bitcoin is not adopted for utility, rather the probability distribution of the adopters is power-law because that is the distribution of money[1] as follows.

http://en.wikipedia.org/wiki/Power_law

Thus Bitcoin adoption is log-logistic as follows. Note that for B=1/2 which is the power-law distribution, that the slope of the log-logistic function gradually declines for the life of the curve. And that is exactly what we are seeing happen thus far as I stated above.

http://en.wikipedia.org/wiki/Log-logistic_distribution

[1] A. Dragulescu and V. Yakovenko. Exponential and power-law probability distributions of wealth and income in the United Kingdom and the United States