Prediction of random charts w/poll | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Is it just me or is cable acting random? 49 replies

Is random really random? 231 replies

Charts, Charts, Charts! 1 reply

Random charts 2 replies

EURUSD prediction poll for fun 5 replies

Attachments: Prediction of random charts w/poll
Exit Attachments

Prediction of random charts w/poll

  • #21
  • Aug 26, 2013 1:27am Aug 26, 2013 1:27am
Quoting erik1
{quote} This does not mean that they behave in the same way. They are statistically different. Run statistical tests for serial correlation and you'll see. Statistical fact.
that will require me to learn statistics.
  • #22
  • Aug 26, 2013 1:30am Aug 26, 2013 1:30am
Quoting EmeraldEyes
{quote} No, don't need to communicate w/ individual traders when dealers report theirs and larger player's orders & positions for me. Not at which second, but at what price level(s) and usually who they are trading for. Ex; Eur/Chf has a very large trader's buy order @ 1.2005. If the market moved randomly, then why did it not go down below 1.20 last year. Nope, don't care to waste time knowing 'strategies' Idk what a chartist is.
emerald nice explorer. i didn't found it in the yearly top 10 although it has the percentage. anyway you're not a chartist. you have a different method.
  • #23
  • Aug 26, 2013 1:43am Aug 26, 2013 1:43am
Thanks! Idk why it doesnt show up sometimes.
Oh, ok. Ya I dont look at the chart too much. Seems like I'm always late to the party when I trade with only the chart.
  • #24
  • Aug 26, 2013 1:48am Aug 26, 2013 1:48am
Quoting EmeraldEyes
Thanks! Idk why it doesnt show up sometimes. Oh, ok. Ya I dont look at the chart too much. Seems like I'm always late to the party when I trade with only the chart.
good thing that we have these 2 explorers in the top 10.
http://www.forexfactory.com/geo1683#57
http://www.forexfactory.com/barradas_fbs#52
  • #25
  • Aug 26, 2013 2:02am Aug 26, 2013 2:02am
Them some funky equity charts. lol
  • #26
  • Aug 26, 2013 4:17am Aug 26, 2013 4:17am
The way to ask the question, the answer is "not a single one". Because you want to prove some method can predict a random event which per definition is impossible. Can you profit from a random walk? yes. If the random walk is normally distributed, enter at random TP=N SL=2*sqrt(N), the bigger N the better R:R. Market aren't Normally distributed.

Can chartists be successful on random chandlestick charts?
Probably yes. After all it's not important you trade a fundamental trend or a spurious one, nor if you trade a price equilibrum vs a mean reverting stochastic process. After all a trend line is just a rough estimate of the rate of growth of the price: as long as price is above a rising TL the price goes in average at least as fast. Otherwise it would cross the line (Bolzano).
When price crosses the line it's no more the case and price may (or not) go in the other direction. The underlying reason of the direction persistance or change is not important.

@Custos, "But if I can show you one chart which doesn't act randomly, then your hypothesis is proven to be wrong."
Just you can't. If I flip a coin and get 20 HEADs you can't prove the coin is unfair, you can only have a legitimate suspicion. If a chart goes in perfect straight line you can't prove it's not just by chance.

@Kanzler, you are correct but in real life if the price nears zero the market will be stopped and your floating loss will be realized (zimbabwe dollar).

I have the feeling the question is equivalent to "how many trade do I need to prove I've an edge?"
No greed. No fear. Just maths.
  • #27
  • Aug 26, 2013 4:45am Aug 26, 2013 4:45am
Quoting PipMeUp
The way to ask the question, the answer is "not a single one". Because you want to prove some method can predict a random event which per definition is impossible."
who says that the event is random ? who decides that?
  • #28
  • Aug 26, 2013 5:15am Aug 26, 2013 5:15am
Yourself sir. First sentence of the first post "i've tested random charts with chandlesticks."
No greed. No fear. Just maths.
  • #29
  • Aug 26, 2013 5:19am Aug 26, 2013 5:19am
What I mean by random:
Quoting PipMeUp
Random has got 7 meanings in English :-/ To clarify, I mean the next bar is a stochastic variable. Let me flip a coin and ask you to guess the result. To your best knowledge, it's either HEAD or TAIL. No reason to privilege any. You don't even know whether the coin is fair or not! So your best prediction is 50%-50%. Here the uncertainty comes from the non-deterministic process to generate the outcome. I now toss the coin but hide the result. The result is now fully determined and won't change any more. It is a constant. Yet your prediction can still...
No greed. No fear. Just maths.
  • #30
  • Aug 26, 2013 5:20am Aug 26, 2013 5:20am
Quoting PipMeUp
Yourself sir. First sentence of the first post "i've tested random charts with chandlesticks."
nah. i said random because people know them as such. if i were to call them puppy charts nobody would know them.
  • #31
  • Aug 26, 2013 5:30am Aug 26, 2013 5:30am
Quoting grin
{quote} what is a random chart? how does it differ from a normal chart?
this is a chart generated with data from random.org.

i see 7 trades here only using price action. 6 win and 1 lose.
Attached Image (click to enlarge)
Click to Enlarge Name: testm2.png Size: 30 KB
  • #32
  • Aug 26, 2013 6:18am Aug 26, 2013 6:18am
You're not the only one to believe so http://www.forexfactory.com/showthread.php?t=410598
But I think his title is wrong. If TA can profit from a puppy chart it makes it quite powerful IMO.
No greed. No fear. Just maths.
  • #33
  • Aug 26, 2013 6:47am Aug 26, 2013 6:47am

that will require me to learn statistics.


Wow, you should definitely learn statistics then. How else are you going to know the degree of confidence with which you can assume a hypothesis is true? How are you going to know you have a statistical edge in the market? Learning statistics is fundamental. I would advice you to take a course in coursera : https://www.coursera.org/course/stats1. There are several other courses you can take, I would recommend this one in computational finance as well: https://www.coursera.org/gatech.

Increase your basic knowledge. Step one
  • #34
  • Aug 26, 2013 6:53am Aug 26, 2013 6:53am
Quoting PipMeUp
You're not the only one to believe so http://www.forexfactory.com/showthread.php?t=410598 But I think his title is wrong. If TA can profit from a puppy chart it makes it quite powerful IMO.
that guy doesn't have a winning aproach. he tries to learn from someone. that someone tries to explain the mythological support and resistance with some economic ideas . that mentor is also a gut instinct trader .
<hey mentor why did you put the support level here and not in other 20 locations?
because my son this is where my gut tells me to place this support. are you going to put more questions or are you going to feel what i am feeling?
yes mentor i'm trying my best. here is 100 dollars for this month. maybe in the next 10 months i will be like you.
>
little do they know that gut instinct cannot be taught.
  • #35
  • Aug 26, 2013 7:04am Aug 26, 2013 7:04am
Quoting erik1
How are you going to know you have a statistical edge in the market?
statistics comes in a texbook and you can learn it and it works.
technical analysis on the internet is garbage from an empirical point of view . it doesn't work in an objective manner.
the concept on the other hand that patterns repeat themselves works but you have to discover something yourself.
either you train your instinct either you objectively define what you see.

>How are you going to know you have a statistical edge in the market?<
what do you think a statistical edge is? let's talk numbers.
  • #36
  • Aug 26, 2013 7:13am Aug 26, 2013 7:13am
Quote
technical analysis on the internet is garbage from an empirical point of view . it doesn't work.
the concept on the other hand that patterns repeat themselves works but you have to discover something yourself.

either you train your instinct either you objectively define what you see.

You can define patterns using algorithms and you can search for pattern-based as well as technical indicator based systems and find methods based on both that have worked historically under market conditions not included in the generation process. I trade based on a genetic programming approach, I can tell you it works. In my opinion, your problem is that you lack a quantitative view. You're telling me that something "works" and something "does not work" but it seems that you can definitions for what these things mean. What does it mean for something to "work" ? How do you measure that ? How does that relate to profit under unknown market conditions?

Quote
what do you think a statistical edge is? let's talk numbers.

The definition of a statistical edge is not subjective. A statistical edge in trading is an ability to derive profits above random chance. In order to know if you have an edge you need to know statistics, how else are you going to determine if you are just being lucky or if you truly have something?

Do 200 trades with a 55% winning rate and a 1:1 risk to reward ratio mean you have an edge? What is the probability that you have an edge, what is the probability that you were lucky ? Trading is a statistical game my friend, if you don't know statistics someone will most likely take your profits away, because they have an advantage over you (a person who knows statistics can get a much clearer view).

Hey, I am just giving you friendly advice, fail to learn statistics, you have a HUGE disadvantage.
  • #37
  • Aug 26, 2013 7:17am Aug 26, 2013 7:17am
You can use Monte Carlo simulations to run TA approaches on 10 years of random data and 10 years of actual market data and I can assure you that you will find very significant differences (I just performed this experiment, 100K simulations). I now have this information - knowing the exact quantitative difference - while you lack it, because you don't know statistics. See what I'm saying?
  • #38
  • Aug 26, 2013 7:41am Aug 26, 2013 7:41am
Quoting erik1
{quote} A statistical edge in trading is an ability to derive profits above random chance.
i told you let's talk numbers and you tell me i don't know statistics.
70% and 1:1. how many trades?
  • #39
  • Aug 26, 2013 7:44am Aug 26, 2013 7:44am
Quoting erik1
You can use Monte Carlo simulations to run TA approaches on 10 years of random data and 10 years of actual market data and I can assure you that you will find very significant differences (I just performed this experiment, 100K simulations). I now have this information - knowing the exact quantitative difference - while you lack it, because you don't know statistics. See what I'm saying?
no. actually what you're saying is that your work in ta is useless. i would not invest a dime in your genetic algorithms. they will fail long term
  • #40
  • Aug 26, 2013 7:57am Aug 26, 2013 7:57am
Quote
actually what you're saying is that your work in ta is useless.

I never said it was useless. I said that I have a measurement of how useful it is! because I know how to carry out the simulations and measure the confidence with which I can assume a hypothesis to be true. I know that TA works differently in random charts and real charts. Do you? can you measure this? You're saying that my genetic programming is useless? Do you have any quantitative evidence about this?

Quote
they will fail long term

I have been trading them for 15 years Mr. They haven't failed Also, even if you wanted to invest in them, you wouldn't be able to. Again, if you don't want to hear my advice, that's your prerogative. But sincerely, there is nothing for you to lose in taking a few free courses about computational finance and statistics in coursera. What would you lose with gaining some knowledge?
0 traders viewing now