trend following delusion shattered

[nononsense]

Quote from Thunderdog:

Nononsense makes an interesting point, but I fail to see how you would arrive at that conclusion based on his post.

In my opinion, if trading the markets only came down to a matter of signal processing, then it would be a relatively straightforward engineering problem. However, that does not appear to be the case.

No Thunderdog, this is not what I am saying. I only noticed that a lot of the terminology is borrowed from signal theory and is often applied in an absolutely nonsensical manner.

Random? I indicated in simple terms how this is defined in signal theory. You tell me how you define random for the ES or DAX. That's the problem. If you can't, it is wiser not to talk too much about it.

After you did, we will be ready to tackle 'The Trend'.

 

[nononsense]

Quote from NickelScalper:

If a series is random within known parameters, then it is predictable to some extent. It is also nonrandom to that same extent.

That's your own definition.

Mathematically you can even define something like a 'random sine wave' which is wholly non-deterministic, ie random. Colored noise, wholly undeterministic is random. Knowing a random signal by its statistical parameters doesn't make it less random.

However, suitable 'random' signals can be predicted using suitable methods, prediction in the sense of well defined mathematical criteria. This does not make such a signal 'less random'.

 

[hank rollins]

Quote from NickelScalper:

If a series is random within known parameters, then it is predictable to some extent. It is also nonrandom to that same extent.

to simplify things, isn't this the concept behind channeling stocks.com and proflogics ideas? old worn out, non testable concepts of little relevance to the modern trader.

 

[Mr Subliminal]

Quote from NickelScalper:

In addition, to say that something is, to some extent, predictable is not inconsistent with it having a random component according to my definition.

Then why did you answer my question with the following :

Quote from NickelScalper:

Tradability derives from extrapolate-ability, a property randomness does not have.

 

[Thunderdog]

Quote from nononsense:

No Thunderdog, this is not what I am saying. I only noticed that a lot of the terminology is borrowed from signal theory and is often applied in an absolutely nonsensical manner.

Random? I indicated in simple terms how this is defined in signal theory. You tell me how you define random for the ES or DAX. That's the problem. If you can't, it is wiser not to talk too much about it.

After you did, we will be ready to tackle 'The Trend'.

Whoa. We seem to be at cross purposes here. My post was to hanksurfer, who drew a false conclusion based on your post. I happen to like many of your posts, what little I understand of them once you go technical. Really. You will not see me using terminology that is over my head. Please go back and confirm this. Further, I do not really need to tackle "The Trend" for reasons I mentioned in my post way back on page 53 of this thread:
http://www.elitetrader.com/vb/showthread.php?s=&threadid=46620&perpage=6&pagenumber=53

 

[NickelScalper]

Quote from nononsense:

That's your own definition.

Mathematically you can even define something like a 'random sine wave' which is wholly non-deterministic, ie random. Colored noise, wholly undeterministic is random. Knowing a random signal by its statistical parameters doesn't make it less random.

However, suitable 'random' signals can be predicted using suitable methods, prediction in the sense of well defined mathematical criteria. This does not make such a signal 'less random'.

As is common practice, you say it's only my opinion but offer no valid counterexample.

If a random series is to some extent predictable based on its know parameters, then it's these determinable aspects that are the basis of the prediction, not the random ones.

 

[NickelScalper]

Quote from hank rollins:

to simplify things, isn't this the concept behind channeling stocks.com and proflogics ideas? old worn out, non testable concepts of little relevance to the modern trader.

Oh, you're looking for the "old worn out, non testable concepts of little relevance."

Back of the store, middle shelf. By the used eyeglasses.

 

[nononsense]

Quote from NickelScalper:

As is common practice, you say it's only my opinion but offer no valid counterexample.
...

This is no matter of opinion or frivolous 'counterexamples'.

If you didn't like what I wrote, read further for yourself:
'Probability, Random Variables and Stochastic Processes', Papoulis, McGrawHill

Or better, write your own.

 

[Thunderdog]

Quote from hank rollins:

to simplify things, isn't this the concept behind channeling stocks.com and proflogics ideas? old worn out, non testable concepts of little relevance to the modern trader.

Thus spoke NeoGann, king of the modern traders.

 

[NickelScalper]

Quote from Mr Subliminal:

Then why did you answer my question with the following :

Better traders trade the nonrandom component.