21 replies to this topic

[arbman]

Have you worked with Elliott Wave analysis for any length of time?

By any means, you might be gifted about this, but I spent hours, and I mean I am an artist as much as I am an engineer and I did try! All I can say is, I can give you a completely random chart with only 3 or 4 cycles superposed and you will be able to identify a ton of EWP patterns --and worst of all you will convince yourself about a certain outcome or expectation at the end!!!

IMHO, most of the patterns are mere coincidence and they simply do not exist. I put together serious pattern searching algorithms using the discrete wavelet decompositions and the accuracy of any specific pattern or group of conditional patterns is much worse than the cycles, or the results are not any better than a simple momentum system, so why bother to decipher the patterns? Most of the general rules about the deep retracements, flats etc are naturally explained by the cycles and in a sense more scienfitically anyway...

I can not show to my partners any EWP chart, but the momentum, rate of change, correlation, cycles...

[jjc]

Have you worked with Elliott Wave analysis for any length of time?

By any means, you might be gifted about this, but I spent hours, and I mean I am an artist as much as I am an engineer and I did try! All I can say is, I can give you a completely random chart with only 3 or 4 cycles superposed and you will be able to identify a ton of EWP patterns --and worst of all you will convince yourself about a certain outcome or expectation at the end!!!

IMHO, most of the patterns are mere coincidence and they simply do not exist. I put together serious pattern searching algorithms using the discrete wavelet decompositions and the accuracy of any specific pattern or group of conditional patterns is much worse than the cycles, or the results are not any better than a simple momentum system, so why bother to decipher the patterns? Most of the general rules about the deep retracements, flats etc are naturally explained by the cycles and in a sense more scienfitically anyway...

I can not show to my partners any EWP chart, but the momentum, rate of change, correlation, cycles...

An experiment might be interesting. Lift a few charts from history and compose a few randomly; publish the
time series and let people forecast and let's see what we get for forecasts. I'd be willing to do some leg work.

I for one, have to know what chart I'm looking at (ie what market, what time frame; Time of Day or Year...) so any forecast I make from such a chart I predict to be worthless.

[da_cheif]

Have you worked with Elliott Wave analysis for any length of time?

By any means, you might be gifted about this, but I spent hours, and I mean I am an artist as much as I am an engineer and I did try! All I can say is, I can give you a completely random chart with only 3 or 4 cycles superposed and you will be able to identify a ton of EWP patterns --and worst of all you will convince yourself about a certain outcome or expectation at the end!!!

IMHO, most of the patterns are mere coincidence and they simply do not exist. I put together serious pattern searching algorithms using the discrete wavelet decompositions and the accuracy of any specific pattern or group of conditional patterns is much worse than the cycles, or the results are not any better than a simple momentum system, so why bother to decipher the patterns? Most of the general rules about the deep retracements, flats etc are naturally explained by the cycles and in a sense more scienfitically anyway...

I can not show to my partners any EWP chart, but the momentum, rate of change, correlation, cycles...

ew is simple....all u gotta do is look at the chart and look for something that hasnt occured yet in terms of elliott..........remember .....something that exists but hasnt happened yet is coming..........thats why im a screaming bull......the epicenter of primary wave 3 up hasnt happened yet....so everything that is happening or has happened is justa series of ones n 2.s till it arrives......SJNORJT

>I can not show to my partners any EWP chart, but the momentum, rate of change, correlation, cycles...
<.......u forgot fibonacci......and clx analysis.

Edited by da_cheif, 18 June 2007 - 05:59 PM.

[Tor]

I missed your multi decade Dow post so if you could direct me to it.

Here you go....and for others who have asked about this in the past.

Fib

My count is that the 2002 dip is a massive wave 4 correction. Wave 5 to come in my view not a three.

Edited by Tor, 18 June 2007 - 06:08 PM.

[arbman]

An experiment might be interesting. Lift a few charts from history and compose a few randomly; publish the time series and let people forecast and let's see what we get for forecasts. I'd be willing to do some leg work.

Enjoy...

[jjc]

An experiment might be interesting. Lift a few charts from history and compose a few randomly; publish the time series and let people forecast and let's see what we get for forecasts. I'd be willing to do some leg work.

Enjoy...

If you publish the series; comma, serperated value It might be useful to run some charting tools against;

[da_cheif]

I missed your multi decade Dow post so if you could direct me to it.

Here you go....and for others who have asked about this in the past.

Fib

My count is that the 2002 dip is a massive wave 4 correction. Wave 5 to come in my view not a three.

big 3.s are not called recognition waves for nothing.....simply because few see it coming......like i sed......if it hasnt happened yet and it hasnt.....its coming.... ....and besides 4.s dont bring out end of the world prognostications like 2002 did.....waves 2.s do that.....

Edited by da_cheif, 18 June 2007 - 06:55 PM.

[jjc]

An experiment might be interesting. Lift a few charts from history and compose a few randomly; publish the time series and let people forecast and let's see what we get for forecasts. I'd be willing to do some leg work.

Enjoy...

If you publish the series; comma, serperated value It might be useful to run some charting tools against;

Oh. I see it is already given as being generated.

[jjc]

An experiment might be interesting. Lift a few charts from history and compose a few randomly; publish the time series and let people forecast and let's see what we get for forecasts. I'd be willing to do some leg work.

Enjoy...

If you publish the series; comma, serperated value It might be useful to run some charting tools against;

Oh. I see it is already given as being generated.

An experiment might be interesting. Lift a few charts from history and compose a few randomly; publish the time series and let people forecast and let's see what we get for forecasts. I'd be willing to do some leg work.

Enjoy...

If you publish the series; comma, serperated value It might be useful to run some charting tools against;

Oh. I see it is already given as being generated.

It's ok. I think we both know what the outcome of such an experiment would be.

[eminimee]

http://stockcharts.com/c-sc/sc?s=$SPX&p=M&st=1987-01-01&en=2007-12-31&i=p36430880995&a=68755843&r=1742.png