[GRAPH] This is not a bubble | Bitcointalksearch.org

thefiniteidea

hero member

Activity: 509

Merit: 564

"In Us We Trust"

Quote from: cloon on February 13, 2013, 04:29:48 PM

this also means that it is not linear in logarythmic scale, why i dont believe in thefiniteidea's chart.

Compared to your's, I don't either Cheesy

It'll work in the short term, that's about it.

Polynomial expansion is way more accurate, especially when you line it up properly- why I threw ya a +1

phelix

legendary

Activity: 1708

Merit: 1020

"The fit curve is an exponential percentual to the chart least squares fit - in other words: I played around until it looked good."

As fiat money is inflating exponentially I guess bitcoin value measured in fiat money should increase exponentially, too. Also stock prices have increased exponentially over a long time.

cloon

sr. member

Activity: 387

Merit: 250

here you have some fancy graphs!
(for mor significance ;-))

http://blockchain.info/en/charts
http://blockchain.info/charts/my-wallet-n-users?showDataPoints=false&show_header=true&daysAverageString=1×pan=all&scale=1&address=
http://blockchain.info/charts/n-transactions-excluding-popular?showDataPoints=false&show_header=true&daysAverageString=1×pan=all&scale=1&address=

about difficulty:
linear
http://www.bitcoinx.com/charts1/chart_large_lin.png
logarythmic
http://www.bitcoinx.com/charts1/chart_large_log.png

the one sure thing out of these graphs is that the community does grow slower than an exponential function.
(its a polynominal form and not of the form a^bx)

this also means that it is not linear in logarythmic scale, why i dont believe in thefiniteidea's chart.

piramida

legendary

Activity: 1176

Merit: 1010

Borsche

Quote from: notme on February 13, 2013, 04:09:10 PM

3 samples can not possibly give you a very high significance level.

I see 6 bottoms on one straight line. We'll see. This looks about right to me, the other charts (like the total transaction fees which I also think is significant) follow the same exact pattern.

luffy

hero member

Activity: 607

Merit: 500

its good to have also a difficulty/btc creation/usd value diagram. i think the last year their relation is linear?!

notme

legendary

Activity: 1904

Merit: 1002

Quote from: piramida on February 13, 2013, 03:42:17 PM

the price chart alone makes no counterpoint to anything. it is the same chart, just harder to see the relative growth rate when you look at absolute values, that's all - not everybody can make the mental exercise of taking log of a graph.

What we are interested here is not an absolute value, but a percentage growth - value % added in a period of time - which is a log chart, meaning the speed at which bitcoin adopters / holders / speculators / users crowd grows.

And about statistics - that bottom line, 4x minimum yearly growth, is being confirmed for three years in a row now, and has yet to be broken. That is somewhat statistically significant, and unless we break below that line ($16 now, $32 in august) it will only be strengthened.

3 samples can not possibly give you a very high significance level.

piramida

legendary

Activity: 1176

Merit: 1010

Borsche

the price chart alone makes no counterpoint to anything. it is the same chart, just harder to see the relative growth rate when you look at absolute values, that's all - not everybody can make the mental exercise of taking log of a graph.

What we are interested here is not an absolute value, but a percentage growth - value % added in a period of time - which is a log chart, meaning the speed at which bitcoin adopters / holders / speculators / users crowd grows.

And about statistics - that bottom line, 4x minimum yearly growth, is being confirmed for three years in a row now, and has yet to be broken. That is somewhat statistically significant, and unless we break below that line ($16 now, $32 in august) it will only be strengthened.

evolve

hero member

Activity: 700

Merit: 500

daytrader/superhero

Quote from: cloon on February 13, 2013, 03:34:49 PM

cmon! i can look this up on bitcoincharts.com myself! and why quoting it?
at least comment it please!

Quoting it as a +1. The chart alone makes a perfect counterpoint to OPs log graph....no explanation is really needed.

thefiniteidea

hero member

Activity: 509

Merit: 564

"In Us We Trust"

FINALLY! A THREAD LIKE THIS! Grin

I AM SO HAPPY! $50+ by October IMO

+1 for cloons, dacoinminster charts

cloon

sr. member

Activity: 387

Merit: 250

Quote from: evolve on February 13, 2013, 03:31:30 PM

Quote from: fcmatt on February 13, 2013, 02:17:59 PM

cmon! i can look this up on bitcoincharts.com myself! and why quoting it?
at least comment it please!

evolve

hero member

Activity: 700

Merit: 500

daytrader/superhero

Quote from: fcmatt on February 13, 2013, 02:17:59 PM

Melbustus

legendary

Activity: 1722

Merit: 1004

Quote from: Vandroiy on February 13, 2013, 02:32:14 PM

And no, taking the logarithm of one of them doesn't make this relation any more reasonable. Roll Eyes

Despite being a long-term bull on bitcoin, I agree with this...

You need an underlying thesis to support the reason why you think it's appropriate to graph this logarithmically. Should you forecast the S&P on a log scale? Oil? Gold? Why or why not?

Seems to me that if you think log scales are appropriate, then you're making an implicit assumption that exponentially more people (or other money) are coming into bitcoin as time goes by. Perhaps that's true (I actually believe it *is* true long-term, but that's a different conversation). Perhaps not... Regardless, it makes no sense to say something is undervalued or overvalued relative to some curve without saying why you think that curve makes sense.

And no, generating that curve purely from a very short history of price data that's operated against a market with wildly changing dynamics does not justify the curve in and of itself. Statistical significance is a bitch.

cloon

sr. member

Activity: 387

Merit: 250

Quote from: TTBit on February 13, 2013, 02:14:00 PM

Well, the purple line fit pretty well in mid 2011 too:

i think this courve here is more expressive, cause it connnects the bottoms which I see as the real value of bitcoin while everything above is just speculation (you call it bubble?). very likely it'll touch this line in future again.
unfortunately i couldn't make the courve as a function in sierra chart (someone knows how to?)

Vandroiy

legendary

Activity: 1036

Merit: 1002

You are aware you are fitting price to absolute supply... which should have an inverse relationship, right?

And no, taking the logarithm of one of them doesn't make this relation any more reasonable. Roll Eyes

dacoinminster

legendary

Activity: 1260

Merit: 1031

Rational Exuberance

Quote from: TTBit on February 13, 2013, 02:14:00 PM

Well, the purple line fit pretty well in mid 2011 too:

Yes, hindsight is 20/20 with things like this. People in 2011 had a decent argument that they weren't in a bubble too. I'm sure in a couple years there will be a much better approximation function.

Imagine the insanity if we are in a correction back to the purple line!

piramida

legendary

Activity: 1176

Merit: 1010

Borsche

Thanks! I agree with TTbit - we can say that in 2011, bitcoin was ascending in the same fashion only much faster - turned out, that was too fast, so this time it grows at a much more comfortable rate.

I also don't think there will be a third line, because if you make a straight line through *minimums* at the log graph, you would see a very strong bottom that was confirmed multiple times over three years. This bottom, which approximately correlates to minimum price doubling every half a year, says that the currently possible minimum price is $16, and would be $32 by autumn.

Guess we'll know soon

fcmatt

legendary

Activity: 2072

Merit: 1001

TTBit

legendary

Activity: 1137

Merit: 1001

Well, the purple line fit pretty well in mid 2011 too:

dacoinminster

legendary

Activity: 1260

Merit: 1031

Rational Exuberance

Quote from: piramida on February 13, 2013, 01:38:44 PM

it also is nicely approximated by a linear graph on log scale; can you calc average deviation for this and linear graphs? Maybe starting at July 2010, since there was no real exchange before then.

Sure, why not? There are only a couple data points before July 2010, so they don't really affect Excel's fitting function very much.

Excel says R-Squared for the Exp function (= 0.010044939004803*EXP(0.005412898222163*X)) is 0.6990; for the power function, R-Squared is 0.7728

Either way, the 2011 bubble stands out like a sore thumb, and today's prices seem quite reasonable. If we see $200 bitcoins before this summer, that would be a bubble of roughly the same magnitude as 2011, not that I would sell my bitcoins

nobbynobbynoob

hero member

Activity: 784

Merit: 1000

Annuit cœptis humanae libertas

It mightn't be a bubble... yet.

There's never a bad time to exchange fiat for BTC, but I'm holding off buying BTC for now. I'm still reeling from missing out on bASIC, when that company folded.

Topic: [GRAPH] This is not a bubble (Read 2358 times)