Good day, Dear thinkers and developers, I post an Expert Advisor based on
the trend reversal and its pullbacks.
A scalper expert Advisor Limmon_v_10.2 timemy M1,M5
I would like to discuss the optimization of this expert Advisor because it has potential, I will provide the necessary information on its work. I don’t know myself
the programmer ordered from the developer for his vehicle.
I attach a screenshot of the test, indicators, SET, owl, in the attachment, The quality of modeling is very weak, as can be seen in the screenshot, I could not
to achieve better, you need to drive it away somehow.
limmon
How
the Expert Advisor works in Kratz:
Advisor
analyzes and trades only on the instrument and timeframe on which it is set.
sale
it is closed (if it is) and the purchase is opened (if it is not) at the time of opening the current candle if the last closed candle has
all four conditions listed below at the same time (on the same last closed candle) :
Example
for Purchases:
1.The
bar of the Elliot_oscillator_-_waves_1. 02 indicator histogram is yellow and below zero
2.
Crossing the lines of the KDJ indicator. The green line crosses the red line from top to bottom
3.
Fast line of the stochastic to be below a predetermined level in settings, for example setting is 30
4. the
line of the rsi indicator should be oversold and below the specified level in the settings, for example, with the parameter 25
all indicator
parameters are displayed in the expert advisor settings
If
these conditions are met, the EA should open a purchase from the next bar.
So
there is also an additional filter in the form of a mashka, where the owl looks so that the distance from the MA to the price is “X” points. The MA period and the number of points are
also customizable
Support of
the transaction controls a certain breakeven or its similarity (not trallit): if the price passes “X” points in our direction, then the stop moves to
breakeven by ” X ” from the order opening price set in the settings. For EXAMPLE: the Price went in our direction 13 points the breakeven moves
3 p. p. plus, if the price turns around and touches breakeven, go to the main profit for example 20 p. p.
and i
also tried to implement a kind of averaging: past the main order, there are still averaging 3 orders if the price did not go in our direction, where
you can configure through what number of pp to open them and with what profit.
Here
while all that i remembered on the principle of the owl, i probably missed something, ask questions.
mak1816w
good afternoon. is this the result only for this year or for others too?
igoreremenko
you need from 99% to 100%, so only MT5 and real ticks + arbitrary delays in execution
b
There is nothing to “fix”. The fact is that on M1, 25% modeling quality is the maximum limit for MT4 because ticks are modeled from minute
bars, from which the tester has data only on four points: OHLC. So roughly speaking 100% / 4 = 25%.
If you run testing on older (relative to M1) TF, then in addition to the native TF, points taken from younger TF are added and the accuracy of
modeling increases accordingly. For older tfs, the maximum simulation quality is 90%. This is all about MT4. For MT5
the accuracy is higher on any TF because there you can test on real ticks.
limmon
here’s a set on tf 15 min.
If you make the Eliot indicator so that the expert follows the appearance of the yellow column and analyzes the previous yellow column, then the
market entries would be much more accurate
igoreremenko
You need to rewrite it for MT5 and test it on real ticks with settings close to real execution
limmon
initially, i would like to see ilanovskaya averaging in the expert advisor, as well as closing
orders in one place, and remove some graphical settings of
indicators from the owl. In general, there are still a lot of ideas for optimization…