Prediction of random charts w/poll | Page 3 | 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

  • #41
  • Aug 26, 2013 8:09am Aug 26, 2013 8:09am
Quoting deltatrade
{quote} 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...
I think he was more trying to disprove the guru/mentor.
No greed. No fear. Just maths.
  • #42
  • Aug 26, 2013 8:16am Aug 26, 2013 8:16am
Quoting erik1
{quote} 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} 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...
sir to simplify this. i have no interest to know about what you do.
ta is a very broad subject and very few ideas are successful . if what you know about ta doesn't work for random from my point of view is useless.
now i asked you about the number of trades for the system to have an edge.
  • #43
  • Aug 26, 2013 8:24am Aug 26, 2013 8:24am
Quote
now i asked you about the number of trades for the system to have an edge.

You can never be absolutely certain you have an edge, you can only say that you might have an edge within a given statistical confidence. This not only depends on the number of trades but on the actual characteristics of the strategy you're trading. If you want to say that you have an edge with a confidence > 99% you can run a variety of statistical tests to find this out (both numerical and analytic). For example you can run Monte Carlo simulations of a system that matches your systems R:R but has a random expectancy, you can then calculate the percentage of results that are above or equal to your profitability, which can give you the confidence with which you can say that you have an edge. Your confidence in having an edge would be 100-(Y/X)*100 where Y is the number of Monte Carlo iterations above or equal to your system's profits and X would be the total number of iterations.
  • #44
  • Aug 26, 2013 8:27am Aug 26, 2013 8:27am
You can perform this Monte Carlo simulations in R if you wish, it's simple as well. Here is an intro to Monte Carlo in R: http://www.stat.ufl.edu/~casella/Sho.../MCMC-UseR.pdf
  • #45
  • Aug 26, 2013 8:34am Aug 26, 2013 8:34am
i want a simple answer .
at 1:1 and 70% how many trades?

i'm using monte carlo simulations for determining max drawdown and consistency in equity chart. i found that consistent equity charts are produced with a profit factor of over 1.8.
  • #46
  • Aug 26, 2013 8:36am Aug 26, 2013 8:36am
Quote
at 1:1 and 70% how many trades?

I won't do your homework With the above information you can calculate this yourself.
  • #47
  • Aug 26, 2013 8:43am Aug 26, 2013 8:43am
Quoting erik1
{quote} I won't do your homework With the above information you can calculate this yourself.
haha. want me to learn r and statistics just for this answer?
anyway i think i have an answer for this. practical one.
  • #48
  • Aug 26, 2013 8:58am Aug 26, 2013 8:58am
Because you impose the R:R=1:1 you can calculate the likelyhood of the probability of your strategy given N trades. N is as big as you need the confidence interval to be tight. http://www.forexfactory.com/showthre...52#post6623952
No greed. No fear. Just maths.
  • #49
  • Aug 26, 2013 9:43am Aug 26, 2013 9:43am
I forgot to add that the edge may come from the R:R instead of the win/loss.
No greed. No fear. Just maths.
  • #50
  • Aug 26, 2013 9:47am Aug 26, 2013 9:47am
Quoting PipMeUp
I forgot to add that the edge may come from the R:R instead of the win/loss.
from testing they came in package.
  • #51
  • Aug 26, 2013 12:06pm Aug 26, 2013 12:06pm
Quoting deltatrade
{quote} from testing they came in package.
This package is called mathematical expectation or expectency for short ;-)
E = (1+R:R) x win/loss - 1
when R:R and win/loss are fixed.
No greed. No fear. Just maths.
  • #52
  • Aug 26, 2013 12:43pm Aug 26, 2013 12:43pm
technical analysis is useless without fundamentals...
  • #53
  • Aug 26, 2013 1:06pm Aug 26, 2013 1:06pm
Quoting openyoureyes
technical analysis is useless without fundamentals...
you did not respond to the original question. ta+fundamentals = astrology.
  • #54
  • Aug 26, 2013 1:20pm Aug 26, 2013 1:20pm
from the pool it seems that a lot of people saw what i saw. now we've seen the ufo . maybe we will make first contact soon. haha.
  • #55
  • Sep 9, 2013 12:19pm Sep 9, 2013 12:19pm
Quoting erik1
{quote} 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...
Erik1, what types of rules do you use in your GA approach?
  • #56
  • Last Post: Sep 9, 2013 1:31pm Sep 9, 2013 1:31pm
Quote
Erik1, what types of rules do you use in your GA approach?

Simple price action based rules but you can do more complex rules without any harm (I tried much more complex rules at some point with the same success). Be careful however to always perform many in sample Vs out of sample runs to ensure that the technique you are using works. It's not enough to develop a system using the last 5 years and then out of sample on the last, always make sure you perform a ton of in sample Vs out of sample analysis across randomly chosen intervals. A few people online are following similar approaches, if you google you should find some software to help you out if you wish to.
0 traders viewing now