A pattern. - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
There is a cool pattern:
If you take a large sliding volume of tick samples (for example 10,000), the average variance of such a sliding data set practically = const.
There is a cool pattern:
If you take a large sliding volume of tick samples (for example 10,000), the average variance of such a sliding data set practically = const.
10000? - Why torture yourself and the computer like that? And not just the tick sample. On sample minutes even 1000 is enough to be roughly constant).
There is a cool pattern:
If you take a large sliding volume of tick samples (for example, 10,000), the average variance of such a sliding data set practically = const.
i've learned it in 2011, that's what inspired me to start writing in mql.
i have experimented with this method. The most effective was MA obtained by the product of the euro ticks and gold ticks - i suspect that it just caught the correlation with the filter for smoothing quotes of the trading server of your brokerage company
the method is not bad, the only thing is the % of deviation from the tick MA to open the order is changing, not often, but when it will change is not clear
HH; the formula is simple: select the length of the array experimentally, I had about double tick[3000], every new tick put in place tick[0], the array pre-shift element by element to the right, then get the tick MA by summing the array elements and dividing by the total number (3000) and the resulting average compared to the obtained tick in % relation. The model works, the profit is on the spread
------------------------
I also remembered a pattern, a bit unclear, but the working model: EUR always makes a certain number of points per day, earlier it was 300 points, now it is about 200 points (on minute TF, consider it M5-M30), regardless of volatility, that is, the sum of intraday zigzag knees is always about the same, of course, it is not precise, but miracles do not happen - like yesterday the ZZ knee total was 700 points, and tomorrow it will be 100 points, and the next day will be 800 points
10000? - Why torture yourself and your computer like that? And it's not just the tick one. On minutes sampling and 1000 is enough to be roughly constant).
Did I understand correctly that we are talking about the approximate constancy of the variance of the sliding series of closing minutes rates over a period of 16 hours and 40 minutes or more? Exactly the rates, not the increments?
Regularity is a fundamentally important thing in predicting processes for a trader...
There isregularity in the markets, but it will not help you make money. Markets are driven by conditions, not events.
There is a pattern in the markets, but it will not help you make money. Markets are driven by conditions, not events.
Events also move the markets and how. Remember the twin towers, the tsunami. Maybe you mean the others?
And the pattern can appear in different forms and by analyzing their combinations, you can find the ideal conditions for entering and exiting the market.
Perfect!!!
There is a cool pattern:
If you take a large sliding volume of tick samples (for example 10,000), the average variance of such a sliding data set practically = const.
Did I correctly understand that we are talking about the approximate constancy of the sliding range of closing rates for the period of 16 hours and 40 minutes and more? Exactly the rates, not the increments?
It's not about courses or increments. It is the variance of the deviations from the regression line. From one day to the next, the variance is virtually unchanged.
If taken over short intervals, the variance will naturally jump a lot.
It's not about courses or increments. It is the variance of the deviations from the regression line. From one day to the next, the variance is virtually unchanged.
If taken over short intervals, the variance will naturally jump a lot.
It's just on this chip fly on the fora (lose our way).
Because it doesn't.
For example all sorts of garches, armas, etc.
It's enough to fly out of the channel once and not come back and confidence is over.
But that's the phrase I've been waiting for instead of the whole "from theory to practice" thread
And yet the chick flew out......
Thanks mate!
It's not about courses or increments. It is the variance of the deviations from the regression line. From one day to the next, the variance is virtually unchanged.
If you take short intervals, the variance will naturally jump a lot.
So you don't change it within a day, then on other frames more than 12H for example, it doesn't work anymore?