:

[chodpaba]

:

[1713458754]

1713458754

[1713458754]

1713458754

[1713458754]

1713458754

[1713458754]

1713458754

[DrYe5]

Sounds plausible. Can you provide more technical detail about how you performed these correlations?

[marvinmartian]

The price needs to stabilize somewhere between $10 and $20 then double in value every two weeks (along with difficulty) for miners to keep mining.

That's a simplistic analysis, but time will tell how valid.

[Enky1974]

The noise is dominant compared to the signal strength, in this environment is very hard to have a good model unless you can successfully model that noise, unlikely you could do it.
I tried to model the price action with an ARIMA model but the confidence bands of the forecast are wide confirming  the fact that noise is high for now.

[Vandroiy]

I'd like to add something that's orthogonal to this kind of analysis, but really important.

Remember that external factors or changes in parameters are gigantic for Bitcoin. Take for example the recent stagnation in Google trends for "Bitcoin" -- was it because of Mt. Gox being down? Or are we facing a change in the "climate" of the market, possibly to slower growth? On what time frame might more early adopters start adding liquidity to the market due to personal or psychological reasons? Will we have trading adoption this year, will Bitcoin face legal troubles?

Every one of these questions answered in an extreme way can blow any behavioral analysis completely off the board, in either direction. So I think people should start doing more estimates on the not so likely scenarios; sum up all those, and they might actually dominate the forecast for speculation. This may be one of the rare cases where the median of expectations is significantly less useful than the average.

Not that I want to stop you from guessing about the most likely scenario. Just a reminder, because many people use such topics to apply directly in speculation.

[TraderTimm]

I agree with your analysis, lord-of-slack.

I need to update my chart, but my last analysis is here - http://farm6.static.flickr.com/5074/5867157443_5ed5d5b2d0_b.jpg

I think we'll need to bust out of that triangle before we really do anything interesting. I have some proprietary indicators down at the bottom, essentially, if they aren't above their respective zerolines, we're not doing much either way. Seems we are 'basing' for the moment, which is pretty healthy market behavior.

[Ellen Alemany]

The way it works right now is I look at 78 different time series patterns from pricing history, some of them single series, some of them double series, and rank them according to how well they correlate with recent patterns. I will then take the most promising of these to try and work out a projection. As you can guess this projection can have a lot of variance, but mostly I try an focus on the case for the median outcome.

So what you're doing is the same as looking at lots of roulette series and then when you find similarities try to predict the next spin?

I am not saying that this does not work but please explain how and why? To me it looks like superstition; time-series analysis seems very misleading.

[ascent]

I am not saying that this does not work but please explain how and why? To me it looks like superstition; time-series analysis seems very misleading.

Except for the fact that market prices push upon psychological boundaries due to herd mentality, then bounce due to the over reaction, and then repeat over and over, creating a pattern that looks suspiciously familiar, and traders see the patterns, and both consciously and subconsciously make an analysis of the pattern, partly due to intuition (I've seen this before somewhere) and and so forth...

[Ellen Alemany]

Ok ok, just reacted on the topic, because this is not anything that is mathematical at all  which you do agree with if I now understand you correctly?

I do not understand the LLN-part, I think your point here is (which I do agree with) that the returns are NOT gaussian?

On the topic TA (of this sort) working/not working I do not comment.

[stergium]

The price needs to stabilize somewhere between $10 and $20 then double in value every two weeks (along with difficulty) for miners to keep mining.

That's a simplistic analysis, but time will tell how valid.

excuse me for getting off topic.
we(miners) would like that  but as it is written many many times in this forum price is adjusted by supply/demant  and maybe a little by difficulty.
i would like a steady price(!)

[Ellen Alemany]

Regarding the LLN, please elaborate if you want to (you can be as technical as you like, I will be able to follow). I still don't understand how that would help you in this case.

[Ellen Alemany]

As I refer to the law of large numbers here it has to do with the number of traders, and the number of trades, if this is large enough then statistically you get fewer outliers from the actions of individual traders and individual trades. They don't disappear completely but they are less frequent. And if outliers are less frequent I think you can rely more on statistical measurements.

I think you're making a big mistake with this kind of reasoning, the returns are not normal (just download any course from yahoo and make a QQ plot). Assuming this is something that in my opinion will cost you money (or rather, you will severely underestimate the volatility). I believe outliers are important and provide a lot of information (and happens on the days which you read about in the newspapers....).

Btw, this rasied an interesting question: What is easier to predict, where 10 people are going or where 10 thousand people are going?

[TiagoTiago]

I'm looking forward to the graphs.

[Jonathan Ryan Owens]

As I refer to the law of large numbers here it has to do with the number of traders, and the number of trades, if this is large enough then statistically you get fewer outliers from the actions of individual traders and individual trades. They don't disappear completely but they are less frequent. And if outliers are less frequent I think you can rely more on statistical measurements.

I think you're making a big mistake with this kind of reasoning, the returns are not normal (just download any course from yahoo and make a QQ plot). Assuming this is something that in my opinion will cost you money (or rather, you will severely underestimate the volatility). I believe outliers are important and provide a lot of information (and happens on the days which you read about in the newspapers....).

Btw, this rasied an interesting question: What is easier to predict, where 10 people are going or where 10 thousand people are going?

You are absolutely right. Outliers provide important data. When I am talking about eliminating outliers I am referring to the results of the projection compared to actual events, not eliminating them from the source data. Because, there is a certain path dependence that comes from extreme events and this has to be included on the analysis.

It is absolutely easier to predict where 10 thousand people are going...

I probably did not phrase that quite correctly either. It is easier to spot true outliers from a large data population than a small one. If you only have ten, then does a 10% event really count as an outlier? It is hard to tell. But if you have 1000, then it is pretty easy to tell when a handful of events are true outliers...

The fact that this level of conversation exists for Bitcoin says everything. If it's not Bitcoin, it will be something else. I think it'll be bitcoin.

[Ellen Alemany]

I probably did not phrase that quite correctly either. It is easier to spot true outliers from a large data population than a small one. If you only have ten, then does a 10% event really count as an outlier? It is hard to tell. But if you have 1000, then it is pretty easy to tell when a handful of events are true outliers...

My conclusion was that I think it depends, if you are Harrison Ford in bodyguard then it is easier to keep track of the ten than of 10 thousand. If you're an urban planner, probably the opposite is true.

(viciously) Translated into finance: If you're selling options you're worried about Whitney, if you writing buy/sell recommendations then you're the urban planner.

On the outliers: I understand what you mean now and the skepticism is now reduced to something based mostly on personal beliefs

[peach]

I don't put too much weight in graph extrapolations. They are 100% technical, whereas the true economy is nearly 100% behavioral.

[Superform]

i think you would be better correlating #users with price

[TraderTimm]

chodpaba,
             Excellent work. I concur with your analysis of the refractory 'resting' period. Takes time for the market to build up an imbalance again, which will be expressed in the usual manner.

[YoYa]

That's real cool chodpaba.


Fuck tho, now that you've shared it, doesn't this lead to a danger of retrenching this as a pattern just by our awareness of it   

[marcus_of_augustus]

Self-similarity in USD/BTC price based on difficulty-period time scale ... ok, you've piqued my interest.