Logarithmic (non-linear) regression - Bitcoin estimated value

[smoothie]

Requesting updated chart since last one did in May 2015.

two weeks went by and no response?

Is this thread still alive?

[mmortal03]

Requesting updated chart since last one did in May 2015.

two weeks went by and no response?

Is this thread still alive?

Maybe just PM Trolololo and ask him directly. He seems to still be active here.

[mskryxz]

bump interested in an update as well

thanks for your work it is much appreciated

[manfred]

Forecasting is totally useless. Based on currently available date you get this result by 2020
 
The past will not tell you the future

[Pente]

Forecasting is totally useless. Based on currently available date you get this result by 2020
 
The past will not tell you the future

I wouldn't say that chart is worthless. The power regression formula makes the most sense to me and I expect (90% confidence) BTC to break $1000 again before 2020.

[macsga]

^^
Those predictions have only got a meaning when you actually have witnessed the periodic phenomenon. Log fit curves tend to produce enormous error factors when you attempt to go for a long term prediction. I too believe though we're en route for greater and bolder prices.

[manfred]

To have different starting dates on one forecast chart does not work. So  here is a fix. Not that is really meaningful. The other issue is the missing early data for bitcoin, it is needed so i started a one cent.


and

[manfred]

Please, Sirs, let's keep this thread clean.

There are other threads to talk about Risto or/and Monero or/and any altcoin vs bitcoin.

Thank you.

 
  
Trolololo, I don't know if you know me or not, but I'm a big fan.  How much of a donation would it require to get you to create a chart like this for Monero? (even though I know we only have a year of history to go off of)

As i sad before the past does not tell the future.

[franckuestein]

Thanks for the updates @manfred, it's always interesting to see new graphics. Glad to see this thread updated again
ping @Trolololo

[Dufkin]

Please, Sirs, let's keep this thread clean.

There are other threads to talk about Risto or/and Monero or/and any altcoin vs bitcoin.

Thank you.

 
  
Trolololo, I don't know if you know me or not, but I'm a big fan.  How much of a donation would it require to get you to create a chart like this for Monero? (even though I know we only have a year of history to go off of)

https://i.imgur.com/5RJcLZ3.png
As i sad before the past does not tell the future.

Interesting graphic, thank you manfred. However, as you already said it is not very reliable due to the fact, that this graphic shows only a very short period of time compared to the graphics that show the regression models of btc and ltc. From my point of view a model which compares the early stages of btc, ltc and xmr would be more informative, although one always have to keep in mind, that btc was the first cryptocurrency which had probably more influence and fans in the beginning and nowadays there are a lot of altcoins available on the market. Furthermore, the price level of btc on the market has influence on the price level of all other coins since they are mutually dependent (kind of, if you ignore trades from real money to cryptocurrencies). Creating a regression model that consideres all these problems is kind of impossible I think.

[mmortal03]

To have different starting dates on one forecast chart does not work. So  here is a fix. Not that is really meaningful. The other issue is the missing early data for bitcoin, it is needed so i started a one cent.

Who's to say that starting the bitcoin price at one cent is meaningful?

[zimmah]

To have different starting dates on one forecast chart does not work. So  here is a fix. Not that is really meaningful. The other issue is the missing early data for bitcoin, it is needed so i started a one cent.

Who's to say that starting the bitcoin price at one cent is meaningful?

although I think the idea is good, the numbers seem off by a lot.

I don't think bitcoin will stay at or below $1000 for that long.

based on this chart, it would seem bitcoin would stay around $1000 far into 2020, which is completely ridiculous.

[mmortal03]

To have different starting dates on one forecast chart does not work. So  here is a fix. Not that is really meaningful. The other issue is the missing early data for bitcoin, it is needed so i started a one cent.

Who's to say that starting the bitcoin price at one cent is meaningful?

although I think the idea is good, the numbers seem off by a lot.

I don't think bitcoin will stay at or below $1000 for that long.

based on this chart, it would seem bitcoin would stay around $1000 far into 2020, which is completely ridiculous.

Yeah, we're above both of those trend lines right now. Maybe the exponential line is closer to the truth.

[americanpegasus]

Well the answer is obviously not quite exponential, but a (n * log(n)) trendline, or a Modified Metcalfe Line.  

Also, how can I generate those gorgeous Monero price charts like you did?  The price per USD has moved up and I think it would be interesting to show.

[Trolololo]

14 April 2016 update:

R²= 0.900263
a = 2.84737734739566
b = -19.2400124152881

The price today (424 USD/BTC) is 74% undervalued from its Log regression expected value (1607 USD/BCT, calculated HERE)

I will upload chart asap.

[Swisk]

I got so excited when I saw regression - I was thinking the whole statistical deal: Adjusted R^2, parameter selection, maybe even a bit of time series modelling. 

I mean, best fit curves are useful, but I wouldn't call it regression per se. I'm not a professional though.

[mmortal03]

I will upload chart asap.

Welcome back!

[Trolololo]

Donations: bitcoin:1Em6yNYWG9FXSLWQki4aDDis1ETQ3WwKpB

[mskryxz]

Nice update!

[mmortal03]

Donated. Thanks!