Where is the line between fitting and actual patterns? - page 34
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
Profitable or what?
I hope we will make it profitable together, and we will consider the task of this branch accomplished). The Expert Advisor should rather just be a bit more intelligent and fit our goals (there should be something to optimize/train).
So that's the question... If there is a clear system, something should be done about it. I've plotted statistics for many symbols (about 50) and different timeframes. They are quite even. Accordingly it is possible to investigate the regularity of "good" times in correlations between instruments. My guess is that if some mechanism does it, it occurs according to a certain algorithm.) Analysis is dead! Long live analysis!
I meant exactly the following: if one combination of two candles causes a reversal and the same combination of two candles causes a continuation, what combinations preceded the reversal and continuation? What happened to other pairs? I was conducting an experiment - I was pulling pips (adversaries, don't laugh) at every, even the smallest, pullback and obtained a cluster of pips, which 99% of the time are reached by the price. I wonder what situations precede the above candlestick combinations. It is quite possible that the 50/50 ratio may shift to some extent by analyzing the nearest past that precedes these combinations.
I have something in my experience - in more than 90% of cases the Expert Advisor enters and exits correctly, but what we still have not solved is the disappointing deviation from its system which leads to the drawdown when the price misses the calculated target.
The variant I've written about (measuring each pullback and setting targets by calculated levels) is too resource-intensive and is unlikely to work in an EA. But it would be interesting to analyze some combinations of candlesticks and identify the overshoot.
I hope that together we will make it profitable, and we will consider the task of this branch accomplished). Rather, the EA should just be a bit more intelligent and fit our goals (there should be something to optimise/train).
Ok. A simple pattern.
5 candlesticks. If the last one is larger than the previous 4, then above its high is a buy, below its low - a sell.
All right.
5 candles. If the last one is bigger than the previous 4, then above its high is buy, below its low is sell.
-first type.
And for the second one?
-the first type.
And for the second one?
Is that the one on the mashcats?
Take a mashka say 100. and a mashka say 20.
And above a mash 100 buy a rebound from a mash 20.
I meant specifically: if one combination of two candlesticks results in a reversal and the same combination of two candlesticks results in a continuation, what combinations preceded the reversal and continuation?
Are the 2 sequels so equally probable? We calculate the probability of one variant and the other and either sit on the fence or open for the more probable one, if this delta is significant to the value we need. We can also increase the depth of description of the situation, the main thing is to gather statistics that are significant for it.
I have something in my experience - in more than 90% of cases the Expert Advisor enters and exits correctly, but what we still have not solved is the annoying deviation from its system that leads to the drawdown, when the price falls short of the calculated target.
This is the worst thing) For a long time I was trying to solve this problem. When the Expert Advisor hits - the market has a "logical" (natural, predictable) course, when it does not - "illogical" (unpredictable, atypical). The secondary problem is that the 10% of illogical market behavior is associated with periods of increased volatility, which can eat up profits of 90% of successful actions of the Expert Advisor. True, my numbers were a bit different (70-80% by 20-30%)
But I think this is a little off topic.
I meant exactly the following: if one combination of two candles causes a reversal and the same combination of two candles causes a continuation, what combinations preceded the reversal and continuation? What happened to other pairs? I was conducting an experiment - I was pulling pips (adversaries, don't laugh) at every, even the smallest, pullback and obtained a cluster of pips, which 99% of the time are reached by the price. I wonder what situations precede the above candlestick combinations. It is quite possible that the 50/50 ratio may shift to some extent by analyzing the nearest past that precedes these combinations.
I have something in my experience - in more than 90% of cases the Expert Advisor enters and exits correctly, but what we still have not solved is the unfortunate deviation from its system which leads to the drawdown when the price misses the calculated target.
The variant I wrote about (measuring each pullback and setting targets according to calculated levels) is too resource intensive and is unlikely to work in an EA. But it would be interesting to analyze some combinations of candlesticks and identify the overshoot.
I would consider this question from a completely different plane: fitting is just fitting. Namely, to find out the system according to which the skewness occurs in the negative and positive sides, and you can automatically forget about the time-consuming process of analysis of expansions (about fibs) and preceding patterns.
By system I mean the pairing of arrays of data obtained from multiple instruments in one time frame.Most likely this is neuronics...
Is that the one on the mashcans?
Take a mashka say 100. and a mashka say 20.
And above a mash 100, buy a rebound from a mash 20.
-the first type.
Are the two extensions so equally likely? We calculate the probability of one option and the other and either sit on the fence or open towards the more probable one, if this delta is significant by the value we need.
This is the worst thing) I've been struggling with the solution of this problem for a long time. When the Expert Advisor hits - the market has a "logical" (regular, predictable) course, when it doesn't - an "illogical" (unpredictable, atypical) one. The secondary problem is that the 10% of illogical market behavior is associated with periods of increased volatility, which can eat up profits of 90% of successful actions of the Expert Advisor. My numbers are a bit different though (70-80%20-30%)
But I think this is a little off topic.
Yes, I agree, the topic is not about that. However, the Expert Advisor earns up to 100% a month, but there is nothing much to pay - it works based on calculated values. And those 8-10% of erroneous actions of the Expert Advisor are caused by sudden and unexpected market movements. It is good that at this moment most of the positions have been partially closed and there is a small amount of drawdown. It is possible to close them bluntly, but I would like to find ways to react to "abnormal" situations.
As for equal probability - if we take the conducted study of these candlestick patterns as an axiom, it is 50/50 with a small deviation, which can be ignored. But if we make one or two or three more analyses of the combinations preceding the studied ones, we will probably find an edge which can be taken into account.