Looking for patterns - page 201
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
I am actually more interested in whether there are those who invented and calculated the statistics? Who took events and calculated over 10 years for example where does the probability on a long time frame "look"?
There may not be here, but there is enough work, not public. it's the first thing that comes to mind for a speculative investor. There is enough capacity, technology allows, and it does not cost that much, in terms of prospects) Williams has a deadline of just 10 years in the works. And he was doing this in 2011.
I am actually more interested in whether there are any people who have figured out how and calculated statistics? Who took events and calculated for 10 years for example where "looks" probability over a long period of time?
You can see a seasonal pattern from the statistics, for example, but I have not been able to predict
You can see seasonal patterns in statistics, for example, but I have never been able to predict
What is the seasonal pattern based on?
What is the seasonal pattern based on?
Surely there are other ways to get statistics of the analyzed instrument, I practice a run on the history of the algorithm, then I export it to Excel and lay it out, summarize by order type and order time, thus getting statistics on activity and purchases or sales, visualize it as a backtest chart where you can visually see the pattern. For example, it is better to buy on Monday and sell on Friday, or it is better to buy in spring and sell in autumn... But these are just statistics that show the probability of entering in one direction or another, we still enter according to the algorithm we believe in :)
There are no patterns, you probably realise that yourself.
Just pure motion statistics, just the formula P(x) = (n/N)*100, where n is the number of favourable x events out of all N events.
That's it.
favourable events - how do you understand that?
To put it simply, an indicator (anything) has shown up (down) - price has gone up (down).
To put it simply, an indicator (anything) has shown up (down) - price has gone up (down).
Yes, you can just export the indicator readings for example
To put it simply, an indicator (anything) has shown up (down) - price has gone up (down).
not just an indicator tested with the EA + tester will be obtained as a result, but analyzed using probabilistic statistics
the approach is clear
not just an indicator tested with an EA + tester, but in terms of probabilistic statistics
approach is clear.
Surely there are a lot of methods to get statistics, I am practicing this and analyzing the algorithm at the same time
No. Everything about the phrases testing levels, gaining strength, etc. is essentially a play on the imagination, there is nothing like that. There is just when price fluctuates and when it moves impulsively. When the price hesitates, which can be anything, you cannot predict the price movement, all you can say is that after the impulse there will be consolidation, after the consolidation there will be impulse and that's it.
This is true, but only in part. We have, incidentally, already discussed this above.
Consolidation is the oscillation of price around a support/resistance level with decreasing amplitude. If the fluctuations are symmetrical (both consolidation levels are close to each other) I personally cannot predict the direction of the further movement, unless fundamental analysis methods are involved.
It is quite another matter if the consolidation is asymmetrical: one triangle boundary is close to the level, while the opposite one is horizontal. This almost always indicates that this horizontal boundary will be broken by an impulse and the trend will move in that direction. The classic "Three Indians" pattern. The first one is standing, the second one is sitting, the third one is lying down. If it is mirrored relative to the horizontal axis - then upwards. It works in almost every first case.