Author

Topic: Wall Observer BTC/USD - Bitcoin price movement tracking & discussion - page 26707. (Read 26609914 times)

hero member
Activity: 565
Merit: 501
S> Cheap SocialMedia Hype's
legendary
Activity: 1568
Merit: 1001
full member
Activity: 196
Merit: 100
is the dumping convention still in town or what? Huh
legendary
Activity: 1596
Merit: 1030
Sine secretum non libertas
LOL. The CCMF indicator is very good, but usually only works after the fact  Smiley

Monkey recommends commodity channel money flow used with mixed moving correlation stochastics.
sr. member
Activity: 546
Merit: 250
On the topic of TA and indicators does any of you use the CMF indicator also found on Bitcoinwisdom and Bitcoincharts?

CMF indicator?

I was disappointed. I thought you may have made a CCMF indicator.

LOL. The CCMF indicator is very good, but usually only works after the fact  Smiley

I can't make much sense of the CMF indicator. Maybe it is useful for longer time frames and for divergences?!
sr. member
Activity: 448
Merit: 250
On the topic of TA and indicators does any of you use the CMF indicator also found on Bitcoinwisdom and Bitcoincharts?

CMF indicator?

I was disappointed. I thought you may have made a CCMF indicator.
hero member
Activity: 672
Merit: 500
It would be nice if we could make this place less about JorgeIdiot and more about Bitcoin.
Eventhough i have him on ignore his bullshit still fills most of the page.
I wish that guy would get a life. But i guess this is what narcissistic sociopaths do.
sr. member
Activity: 546
Merit: 250
On the topic of TA and indicators does any of you use the CMF indicator also found on Bitcoinwisdom and Bitcoincharts?
legendary
Activity: 2380
Merit: 1823
1CBuddyxy4FerT3hzMmi1Jz48ESzRw1ZzZ
legendary
Activity: 1078
Merit: 1441
hero member
Activity: 994
Merit: 1000
It looks to me like Huobi is forming a nice base at the 50% retrace level off the recent top. Bears can't get any momentum going.... launch soon? Well, I hope so. Smiley
sr. member
Activity: 546
Merit: 250
One of the things that make me skeptic of Technical Analysis (TA) is their general reliance on high and low points to determine trends and such.

Highs and lows are very noisy statistics, that may depend on a single decision by a single trader.  So much so, that even for exchanges that track each other very closely  (like Huobi and OKCoin), trend lines that connect highs or lows may be very different depending on which chart one uses -- even increasing/converging on one chart, while decreasing/diverging in the other.

Basic statistics says that the weighted mean a much more reliable parameter than high, low, open, or close; and a weighted least-squares-fitted line is much better than a line connecting any of those points.  To quantify the size of deviations from the mean or the fitted line, the standard deviation is much better than the high-low difference.


Fair point.

That is where experience and judgment come into play, also seeing TA for what it is, an indicator to aid judgement calls - it really would not be possible to have a market at all if there was an indicator that was right 100% of the time.


Maybe another good reason for you to start trading BTC Jorge Smiley I agree with what Empowering says. I don't know much about statistics, but it is useful to consider that the market is moved by two opposite forces (bulls and bears). When price fails to move below a certain point it is because bears have run out of steam. And when price fails to go higher is is because bulls have run out of steam. It doesn't matter if it is a single player who prevents further movement in either direction. All we can say is that price failed to go beyond that point and that is always significant. So maybe it is better to think about the market and TA as forces in physics instead of statistics. If price fails to go further in one direction it will bounce and move in the opposite direction. Like physical objects bouncing off each other when colliding.

It is not enough to consider one exchange alone, but since they all go together you can still use TA from one exchange as a general indicator for market direction.
legendary
Activity: 1596
Merit: 1030
Sine secretum non libertas

Highs and lows are very noisy statistics, that may depend on a single decision by a single trader.  So much so, that even for exchanges that track each other very closely  (like Huobi and OKCoin), trend lines that connect highs or lows may be very different depending on which chart one uses -- even increasing/converging on one chart, while decreasing/diverging in the other.

Basic statistics says that the weighted mean a much more reliable parameter than high, low, open, or close; and a weighted least-squares-fitted line is much better than a line connecting any of those points.  To quantify the size of deviations from the mean or the fitted line, the standard deviation is much better than the high-low difference.


Moving averages and z scores are t.a. indicators as well.  You are describing features.  Your job is to find a utile way to combine features to compose control decisions.  Some features are more naively utile.

legendary
Activity: 1470
Merit: 1007
One of the things that make me skeptic of Technical Analysis (TA) is their general reliance on high and low points to determine trends and such.

Highs and lows are very noisy statistics, that may depend on a single decision by a single trader.  So much so, that even for exchanges that track each other very closely  (like Huobi and OKCoin), trend lines that connect highs or lows may be very different depending on which chart one uses -- even increasing/converging on one chart, while decreasing/diverging in the other.

Basic statistics says that the weighted mean a much more reliable parameter than high, low, open, or close; and a weighted least-squares-fitted line is much better than a line connecting any of those points.  To quantify the size of deviations from the mean or the fitted line, the standard deviation is much better than the high-low difference.

Median (and averages in general) are used in TA, but you are right, a lot of trendlines are largely based on extrema. I disagree however with your claim that the latter are too noisy to be useful: you seem to apply the insight from other fields were a noisy, sparsely populated parameter is best avoided. Two arguments why this insight can't just be applied 1:1 to trading: a) price extrema are the result of trades "eating" through the order book first, i.e. they are the conclusion of a longer, motivated process, not a randomly generated "blip", b) the knowledge what the exact maximum/minimum price some human (or his algo) was willing to trade for is not irrelevant or meaningless at all. I understand that from a purely statistical point of view, your argument makes sense, but taking into account that we are trying to model human behavior, not, say, bacterial growth, not everything can be carried over as straightforward, I believe.
legendary
Activity: 1512
Merit: 1005
626 is 1 12 hr high. Is it the ignition?

legendary
Activity: 2380
Merit: 1823
1CBuddyxy4FerT3hzMmi1Jz48ESzRw1ZzZ
legendary
Activity: 1078
Merit: 1441
One of the things that make me skeptic of Technical Analysis (TA) is their general reliance on high and low points to determine trends and such.

Highs and lows are very noisy statistics, that may depend on a single decision by a single trader.  So much so, that even for exchanges that track each other very closely  (like Huobi and OKCoin), trend lines that connect highs or lows may be very different depending on which chart one uses -- even increasing/converging on one chart, while decreasing/diverging in the other.

Basic statistics says that the weighted mean a much more reliable parameter than high, low, open, or close; and a weighted least-squares-fitted line is much better than a line connecting any of those points.  To quantify the size of deviations from the mean or the fitted line, the standard deviation is much better than the high-low difference.


Fair point.

That is where experience and judgment come into play, also seeing TA for what it is, a set of indicators to aid judgement calls - it really would not be possible to have a market at all if there was an indicator that was right 100% of the time. Plus there are ways of helping reduce the noise MA's etc

hero member
Activity: 910
Merit: 1003
One of the things that make me skeptic of Technical Analysis (TA) is their general reliance on high and low points to determine trends and such.

Highs and lows are very noisy statistics, that may depend on a single decision by a single trader.  So much so, that even for exchanges that track each other very closely  (like Huobi and OKCoin), trend lines that connect highs or lows may be very different depending on which chart one uses -- even increasing/converging on one chart, while decreasing/diverging in the other.

Basic statistics says that the weighted mean a much more reliable parameter than high, low, open, or close; and a weighted least-squares-fitted line is much better than a line connecting any of those points.  To quantify the size of deviations from the mean or the fitted line, the standard deviation is much better than the high-low difference.
legendary
Activity: 1078
Merit: 1441
As far as I am concerned it doesn't matter whether the case for freedom and justice is made by pope or antipope, what matters is that it is made and pressed and pursued with unflagging vigour.  As long as Vorhees is making that case, he is representing me. I should prefer that Lawsky would be making it, but that's not to be, it seems.


The fight for freedom and justice is distributed and decentralised - any bottlenecks will be bypassed  Wink
legendary
Activity: 1596
Merit: 1030
Sine secretum non libertas
Old man Lawski has the Scooby Gang scared at the moment. Soon they'll discover he's just a Superintendent of a run down old school of thought. His scary monster act only works on the kids playing in the Status Quo club. The new investors are free to move their money anywhere that makes them happy and offers more freedom.

Nice.  Can you do league of justice? "Wonder twins powers...activate!"
Jump to: