Topic: How to explain google trends vs BTC/USD ratio? (Read 4445 times)
No!
Log won't work (as you'd need to log both the price and the trend) - it'd just make the increases look slower. Unless you're proposing an exponential relationship between the two? That is possible, but proving exponential correlation in my experience (social sciences) can be questionable as there are so many different ways to test if you go beyond linear correlation.
I've been puzzled by the fact that BTC price hasn't gone down as the Google trends might predict a price of closer to $45 (if you compare search trends with Jan/Feb 2013). Maybe it has crossed a threshold where it is self-sustaining (for instance there is consistent mainstream media stories on it)?
I'm not entirely convinced that MtGox prices would be the main driver for search activity. There could be another cause that is behind both.
I've been puzzled by the fact that BTC price hasn't gone down as the Google trends might predict a price of closer to $45 (if you compare search trends with Jan/Feb 2013). Maybe it has crossed a threshold where it is self-sustaining (for instance there is consistent mainstream media stories on it)?
I'm not entirely convinced that MtGox prices would be the main driver for search activity. There could be another cause that is behind both.
Try looking into the logged graph perhaps? 
How could we explain this spread between btc price and google trends.
The first time I learned about the word twerking I did what many others did and googled it.
I've heard the word twerking hundreds of times since and occasionally have seen it occur but I've not once done a search on it in ever since.
So even though there is (presumably) more twerking occuring today than there was yesterday, the frequency of Google searching for twerking is down today versus yesterday.
So similarly, for Bitcoin, Google trends will spike when there is an influx of those who are becoming aware of bitcoin for the first time, but Google trends can drop even though the actual subject of the term is seeing an increase in activity.
But because financial speculation discussion (due to a rising exchange rate) is the primary reason people are learning of the term "bitcoin", the google trends will lag, not lead, the exchange rate. Thus Google searching for the term bitcoin can drop significantly without the exchange rate falling as well.
Excellent explanation. Although I feel like we've covered this subject on this subforum many times previously... the dead horse picture was quite appropriate.
There is a strong correlation between number of people twerking and btc price. It is safe to assume causation at this point, imo.
Since the series contains a lot of information you can determine causation through finding a time-shifted correlation peak.
Gox leads Google Trend.
This has been long established.
some of your posts are perplexing. Are you a time-traveler? If so, what year are you from?
Yeah, I'm from 2011... We had all these conversations back then.
You are traveling the wrong way, fool.
I think it's pretty safe to assume there are not many new entrant to this market right now ... the only volume is 'whales' who are clearly HNW individuals painting the tape.
WHEN there is a BTC economy to underpin price a new cycle can start. Until then, it's just greed and manipulation.
WHEN there is a BTC economy to underpin price a new cycle can start. Until then, it's just greed and manipulation.
There is a strong correlation between number of people twerking and btc price. It is safe to assume causation at this point, imo.
Since the series contains a lot of information you can determine causation through finding a time-shifted correlation peak.
Gox leads Google Trend.
This has been long established.
some of your posts are perplexing. Are you a time-traveler? If so, what year are you from?
There is a strong correlation between number of people twerking and btc price. It is safe to assume causation at this point, imo.
How could we explain this spread between btc price and google trends.
The first time I learned about the word twerking I did what many others did and googled it.
I've heard the word twerking hundreds of times since and occasionally have seen it occur but I've not once done a search on it in ever since.
So even though there is (presumably) more twerking occuring today than there was yesterday, the frequency of Google searching for twerking is down today versus yesterday.
So similarly, for Bitcoin, Google trends will spike when there is an influx of those who are becoming aware of bitcoin for the first time, but Google trends can drop even though the actual subject of the term is seeing an increase in activity.
But because financial speculation discussion (due to a rising exchange rate) is the primary reason people are learning of the term "bitcoin", the google trends will lag, not lead, the exchange rate. Thus Google searching for the term bitcoin can drop significantly without the exchange rate falling as well.
Somebody, please tell me that I am not the only one who does cross correlation analysis...
Keep doing, I like yours analysis.
Somebody, please tell me that I am not the only one who does cross correlation analysis...
Sorry, I cant, could you link me whot it is explained like for 5y old kid?
Well, if I'd try to, I'd probably get some silent NaNs and spend a lot of time debugging. 
I've crossbred a penguin with a horse if it helps
It only makes the relevant equations more complicated.
I've crossbred a penguin with a horse if it helps
This has been talked to death. $/BTC leads Google Trends.
If you beat a horse long enough, it will eventually come back to life. Seen this many times.
Have you tested it—mathematically.
Cross correlation analysis is not that complicated.
I can divide by zero and don't need to test anything mathematically. Thanks.
I can multiply by zero and make everything equally meaningless.
That makes no sense. This is not an acknowledgement of your ability.
Leave the horse alone.
This has been talked to death. $/BTC leads Google Trends.
If you beat a horse long enough, it will eventually come back to life. Seen this many times.
Have you tested it—mathematically.
Cross correlation analysis is not that complicated.
I can divide by zero and don't need to test anything mathematically. Thanks.
This has been talked to death. $/BTC leads Google Trends.
If you beat a horse long enough, it will eventually come back to life. Seen this many times.
You might find a correlation if you look hard, but there is not an apperant reason. Think what happens if the hordes, having used the money for fun a few months, massively chooses to go all in.
Trade volume is in agreement with Google trends.
That's not trade volume. That's MtGox volume. I see this mistake over and over again.
Tru its oversimplify now, do you know some webpage where it is posible to check cumulative volume of bigest exchanges?
Because the Google trend is absolute whereas bitcoin price is cumulative.
The current value of bitcoin is an expression of all previous price activity - the google trend is just what happens each day.
The current value of bitcoin is an expression of all previous price activity - the google trend is just what happens each day.
I would belive if it was like that previously
Jump to: