Author

Topic: rpietila Wall Observer - the Quality TA Thread ;) - page 230. (Read 907248 times)

hero member
Activity: 763
Merit: 500
Those shorters may have a very hard time if Houbi breaks 2800CNY.
for the record: broken at around 10:00 UTC on 10k volume.
There's nothing significant about 2800. Breaking 3200 would be significant.
for the record: 3200 has been broken yesterday at 2300 UTC (on low volume). volume increased until 3450 and ended with a sharp 10k decline to retest 3200.

this could be the first wave of pullbacks, but basically i had to reply to your earlier posting ;-)
full member
Activity: 170
Merit: 100
Eureka! It is this simple:

- Every predictor gives two prices in log scale eg. "In 2014-5-16 the price is between 2.7 and 2.85 (roughly 500 and 700)"

- When the actual price is known, you take min [ abs ( actual - upper_limit); abs ( actual - lower_limit) ]

- Whoever has the lowest average error after a reasonable number of predictions (predictions can be renewed as often as you wish regardless of their maturity) is the best!  Grin

- Proof omitted  Wink

I would be very grateful if you could explain this to a simpleton like myself.
donator
Activity: 1722
Merit: 1036
I have not thought it through, wanted to have feedback! Smiley

From my background the intuitive thing is to use F1 measure as a score.  Or area under ROC curve.  There are good wikipedia articles about these.


These seemed to be concerned with binary outcomes. I want a metric that is continuous. I am developing one currently Smiley

Eureka! It is this simple:

- Every predictor gives two prices in log scale eg. "In 2014-5-16 the price is between 2.7 and 2.85 (roughly 500 and 700)"

- When the actual price is known, you take min [ abs ( actual - upper_limit); abs ( actual - lower_limit) ]

- Whoever has the lowest average error after a reasonable number of predictions (predictions can be renewed as often as you wish regardless of their maturity) is the best!  Grin

- Proof omitted  Wink
donator
Activity: 1722
Merit: 1036
I have not thought it through, wanted to have feedback! Smiley

From my background the intuitive thing is to use F1 measure as a score.  Or area under ROC curve.  There are good wikipedia articles about these.


These seemed to be concerned with binary outcomes. I want a metric that is continuous. I am developing one currently Smiley
hero member
Activity: 686
Merit: 501
Stephen Reed
At the moment, China is just waking up and volume is smoothly climbing on Huobi's 30 minute chart as displayed at Bitcoin Wisdom.

On the daily Bitstamp chart, $550 appears to be the trend line of resistance.


legendary
Activity: 1148
Merit: 1001

The name Panda Express says it all. That's the express train the panda bears are taking in order to catch upt to our beloved choo choo train so that they don't miss it.


Ha Ha Ha.  The irony that it was "Panda Express" was lost on me until you said that.

That makes it an even greater sign that we are on the "Panda Express" train to the moon now. Wink
full member
Activity: 235
Merit: 100
I was promised da moon
Quick TA update:
- 6H candle color/volume: BREAKING - huge green candle beats the previous reds AND action afterwards is very promising, conclusion: probability of recent reversal further increased
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $84, slippage to buy: $143, conclusion: potential for extreme volatility
- Trendline comparison: we are now at -0.341 log units. The trendline is at $1,015 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: short covering/panic buys are starting
- Prognosis: getting better; probability for going <400 gets smaller by the day, long-term buy zone

And only 1 day after...
Quick TA update (at $500):
- 6H candle color/volume: after a row of tall reds we have turned exclusively tall green, conclusion: reversal seems confirmed
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $77, slippage to buy: $83, conclusion: potential for high volatility
- Trendline comparison: we are now at -0.310 log units. The trendline is at $1,022 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: awaiting a breakout or a pullback
- Prognosis: reversal seems confirmed and <400 fades, long-term buy zone


 Smiley  Good times ahead it appears! 

Hey, my fortune cookie from today at Panda Express said, "YOU WILL BE COMING INTO A FORTUNE"

Does that count as confirmation!?   Grin
Missing Panda express, maybe the best fast food joint there is. I guess it U.S. only. One week macd has also turned to declining red, that should be a nice sign.

The name Panda Express says it all. That's the express train the panda bears are taking in order to catch upt to our beloved choo choo train so that they don't miss it.

I was expecting a strong dent today after hitting 500 before we go further up again, but that hasn't come true. This short term recovery has quite a bite to it, leaving quite nice prospects for the long term. I'm tired of the bears saying we have to break thousands of downward trendlines before we can go up again. Once every single one of those lines is broken, we ARE already up again.
hero member
Activity: 728
Merit: 500
Here is the Blockchain adjusted transaction quantity chart, for 180 days using the 7-day moving average. The rightmost 4 days appear to confirm the reversal of the price trend.



Thanks. I noticed we also had something similar hapoening between May and June 2013, and confirmed a reversal in price trend, albeit it was not the bottom.
full member
Activity: 155
Merit: 100
Quick TA update:
- 6H candle color/volume: BREAKING - huge green candle beats the previous reds AND action afterwards is very promising, conclusion: probability of recent reversal further increased
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $84, slippage to buy: $143, conclusion: potential for extreme volatility
- Trendline comparison: we are now at -0.341 log units. The trendline is at $1,015 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: short covering/panic buys are starting
- Prognosis: getting better; probability for going <400 gets smaller by the day, long-term buy zone

And only 1 day after...
Quick TA update (at $500):
- 6H candle color/volume: after a row of tall reds we have turned exclusively tall green, conclusion: reversal seems confirmed
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $77, slippage to buy: $83, conclusion: potential for high volatility
- Trendline comparison: we are now at -0.310 log units. The trendline is at $1,022 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: awaiting a breakout or a pullback
- Prognosis: reversal seems confirmed and <400 fades, long-term buy zone


 Smiley  Good times ahead it appears! 

Hey, my fortune cookie from today at Panda Express said, "YOU WILL BE COMING INTO A FORTUNE"

Does that count as confirmation!?   Grin
Missing Panda express, maybe the best fast food joint there is. I guess it U.S. only. One week macd has also turned to declining red, that should be a nice sign.
legendary
Activity: 1148
Merit: 1001
Quick TA update:
- 6H candle color/volume: BREAKING - huge green candle beats the previous reds AND action afterwards is very promising, conclusion: probability of recent reversal further increased
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $84, slippage to buy: $143, conclusion: potential for extreme volatility
- Trendline comparison: we are now at -0.341 log units. The trendline is at $1,015 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: short covering/panic buys are starting
- Prognosis: getting better; probability for going <400 gets smaller by the day, long-term buy zone

And only 1 day after...
Quick TA update (at $500):
- 6H candle color/volume: after a row of tall reds we have turned exclusively tall green, conclusion: reversal seems confirmed
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $77, slippage to buy: $83, conclusion: potential for high volatility
- Trendline comparison: we are now at -0.310 log units. The trendline is at $1,022 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: awaiting a breakout or a pullback
- Prognosis: reversal seems confirmed and <400 fades, long-term buy zone


 Smiley  Good times ahead it appears! 

Hey, my fortune cookie from today at Panda Express said, "YOU WILL BE COMING INTO A FORTUNE"

Does that count as confirmation!?   Grin
hero member
Activity: 686
Merit: 501
Stephen Reed
Here is the Blockchain adjusted transaction quantity chart, for 180 days using the 7-day moving average. The rightmost 4 days appear to confirm the reversal of the price trend.

donator
Activity: 1722
Merit: 1036
Quick TA update:
- 6H candle color/volume: BREAKING - huge green candle beats the previous reds AND action afterwards is very promising, conclusion: probability of recent reversal further increased
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $84, slippage to buy: $143, conclusion: potential for extreme volatility
- Trendline comparison: we are now at -0.341 log units. The trendline is at $1,015 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: short covering/panic buys are starting
- Prognosis: getting better; probability for going <400 gets smaller by the day, long-term buy zone

And only 1 day after...
Quick TA update (at $500):
- 6H candle color/volume: after a row of tall reds we have turned exclusively tall green, conclusion: reversal seems confirmed
- Bid/ask strengh at market (Bitstamp): slippage to sell 5k: $77, slippage to buy: $83, conclusion: potential for high volatility
- Trendline comparison: we are now at -0.310 log units. The trendline is at $1,022 and rising $7 per day, conclusion: rock bottom (note: it is not necessary that 'rock bottom' will change until the parabolic uptrend starts, because the trendline is itself rising)
- Sentiment: awaiting a breakout or a pullback
- Prognosis: reversal seems confirmed and <400 fades, long-term buy zone
legendary
Activity: 1162
Merit: 1007
The market today does not give more than 12% probability of that happening; probably much less than that.

That's why I am buying and telling others to buy.

Bitcoin has a long and glorious history that its price has had nothing to do with long-term expected value. My estimate of scenario-probability-weighted discounted 2020 value of 1 bitcoin is about $500,000 and the fact that the majority of players in the market think differently, is my advantage.

And I am quite sure I know the reason for that. It is the extreme information asymmetry between Bitcoin owners and non-owners. (To simplify,) Bitcoin owners know that bitcoins are very valuable, and they buy and hold as much as they can. But as a result of their price going up, it forces people to sell, not because they don't trust Bitcoin, or expect the EV to be anything less than $100k or whatever in a few years, but because their entire net worth consists of bitcoins and it is rational to diversify (cf. Kelly betting). Even if the price goes down, these true believers cannot buy much more because the same optimal allocation strategy holds true.

The ones who don't own bitcoins, don't know it is good Smiley They are converting to bitcoiners at an exponential rate, though (see my previous post, and this thread for the ownership distribution). This increasing adoption gives the steam to the rise of price, in a self-sustaining virtuous cycle.

First only very few people knew about Bitcoin, and they were technologically astute and early adopter type. Many could start using Bitcoin immediately upon hearing. In the second phase more people heard, but they were, on average, slower. In the third phase or so, in which I belong, I heard and it took me 12 months to buy. In the next one, people who heard in 2011, propelled the boom in 2013 because their average decision making time was even longer. In the whole of 2013, especially the time around the peaks, 100s of millions of people heard about Bitcoin. On average, these people are even slower than previous ones, perhaps taking on average 24 months to enter (and also a smaller percentage of them enters). But they are so many! There is nothing you can really do to prevent the fraction of these people adopting Bitcoin, and by doing it, their action increases the price again by a decade, letting the rest of the planet to hear about it and putting hundreds of millions of people in the pipeline.

Bitcoiners know it, the rest don't. The price does not go to $1 million in an instant because only the ones who realize the above thing, can push the price up, but it will nevertheless go there eventually, because the above thing is a self-sustaining loop. Therefore buying bitcoins when you read and realize it, is the most lopsided bet you can ever make. You can actually pay $460 for something whose EV (expected value = average payback of all scenarios weighted by their probabilities) is in between $100k and $1M+. ONLY the fact that people realize this one at a time, slowly but exponentially, keeps the bitcoin price "at check", growing at an average of 1000% per year, for 5+ years now.

My wealth is dependent on the above being true. Nobody has explained to me, why it is not so. The stage is yours! Smiley

Kudos for the person who explained this to me in 2010, after which it took me 12 months to buy, after which it took an additional 18 months to realize it, which realization threatened my sanity, which was only saved by starting to smoke cigars. After 12 more months I tell it to you, please don't repeat my mistakes. Consider if it is valid, and invest accordingly. I am quite sure that next month you will not see this low prices any more, because the information is freely available.


+1000

This is such a concise articulation of what I also believe to be exactly correct that I am considering printing it out and hanging it on my wall Smiley

Agree.  This was a fantastic post Risto, and it further clarified my understanding of the dynamics at play.
hero member
Activity: 686
Merit: 501
Stephen Reed
There is a book which seems pertinent,  "Evidence-based Technical Analysis" although I have yet to read it -- it's next to my bed, but in another state.
http://www.amazon.com/Evidence-Based-Technical-Analysis-Scientific-Statistical/dp/0470008741/ref=sr_1_1?ie=UTF8&qid=1397582403&sr=8-1&keywords=evidence+based+technical+analysis


Thanks. I ordered it. It may be helpful in my work observing the bitcoin price data series.

My own trading, if any, is limited to selling at the bubble peak and buying back later. I am very happy making small almost daily purchases from my nearby Robocoin ATM, when the price is below the trendline.
hero member
Activity: 686
Merit: 501
Stephen Reed
I have not thought it through, wanted to have feedback! Smiley

Wisdom of Crowds theory says that the average of our guesses should be more accurate than any single guess. This should be a motivation for us to contribute our predictions regardless of our self-supposed expertise.

What I would add to your thinking is to distinguish bubbles as the chief bitcoin price phenomenon.

Furthermore, I suggest a survey in addition to particular predictions. Survey data to be gathered would include . . .

  • whether the participant agrees that there are bubbles
  • what is the current bubble phase, i.e. pre-peak, post-peak, pre-bottom, post-bottom
  • what is the predicted time to the next bubble phase
  • peak price of the next peak
  • month of the next peak
  • price of the next post-peak bottom
  • month of the next post-peak bottom
  • year during which 50% adoption occurs

If there is no better place, I would conduct such a survey each month on my logistic price model thread.
legendary
Activity: 1596
Merit: 1030
Sine secretum non libertas
I have not thought it through, wanted to have feedback! Smiley

From my background the intuitive thing is to use F1 measure as a score.  Or area under ROC curve.  There are good wikipedia articles about these.

Precision is not always a good thing.  It leads to over-fitting.

There is a book which seems pertinent,  "Evidence-based Technical Analysis" although I have yet to read it -- it's next to my bed, but in another state.
http://www.amazon.com/Evidence-Based-Technical-Analysis-Scientific-Statistical/dp/0470008741/ref=sr_1_1?ie=UTF8&qid=1397582403&sr=8-1&keywords=evidence+based+technical+analysis
donator
Activity: 1722
Merit: 1036
My congratulations to Risto for his predictions.

Ten days ago or two weeks ago he predicted that we would probably be back above 500 by the 15 April.

Thanks. 500 was just flashing when I first read that Smiley

I am thinking to show the example of accountable forecasting, as follows (figures example only):


In 30 days, 2014-5-15, USD/BTC (Bitstamp daily vwa) is:

75% = 700
50% = 600
25% = 500

In 90 days, 2014-7-15, USD/BTC (Bitstamp daily vwa) is:

75% = 1800
50% = 1000
25% = 600

After the prediction expires, let's say the first one closes at 550, then we calculate the difference of the price in log scale to the forecasted prices:

75% = 0.105
50% = 0.038
25% = -0.041

Now our common goal is to find a formula that lets us compare different people's price forecasts equitably.

With the midprice, I propose that we use the least-average-error method, ie. take the absolute value of each error over time and average them. (It should be penalized imo if someone is consistently wrong to either side, though, but how to do it without making the system gamable?)

With 25% ja 75% bands it is more tricky. My intention is that it should be a target to give as tight bands as possible, and allow 25% of the forecasts to be in error to the tail side. Perhaps calculate the difference of the 75% and 50% forecasts (in log) and make an average of these, up to the limit of (1-75%) of the forecasts to be erroneous. If the forecast is too conservative, it is reflected in wide bands and thus a worse score. If it is too tight, and more than 25% of the results end up on the tail side, there would be a significant penalty based on the difference of actual prices and the forecasts multiplied by the number of excessive erroneus results.

I have not thought it through, wanted to have feedback! Smiley
sr. member
Activity: 338
Merit: 250
My congratulations to Risto for his predictions.

Ten days ago or two weeks ago he predicted that we would probably be back above 500 by the 15 April.
legendary
Activity: 1106
Merit: 1007
Hide your women
Hmmm. This is interesting. From wikipedia: http://en.wikipedia.org/wiki/Sunk_costs

Features characterizing the sunk cost heuristic

Two specific features characterizing the sunk cost heuristic worth mentioning are:

    An overly optimistic probability bias, whereby after an investment the evaluation of one's investment-reaping dividends is increased.
    The requisite of personal responsibility. Sunk cost appears to operate chiefly in those who feel personal responsibility for the investments that are to be viewed as sunk cost.

Just though I'd put this in here instead of doing the probability estimation thing.
legendary
Activity: 3430
Merit: 3080
The only coins mined are the Bitcoins mined by Bitcoin miners, and any bitcoin can be purchased into a sidechain where they are stored until such time as they may be sold back into the Bitcoin blockchain to again become available for ordinary bitcoin transactions.

I don't believe that "purchased" and "sold" are the best way to describe what is happening. I think it is better to describe the operations of moving into and out of sidechains as "conversions", with a conversion rate that is determined by the way the rules that a given sidechain is coded.

How much expression is permitted to define the convertibility rules will be a significant point of interest. The full implications may be over the heads of those that will make the decision to accept or reject any proposed protocol alterations (I'm thinking primarily of pool operators here). One worth watching IMO, in any event.
Jump to: