Who wants a strategy? Lots and for free) - page 22

 
zfs писал(а) >>

I have unfortunately been rushing into the real since 2002. I also trade on the MICEX and Forts. In forex I can say that I am back. I have noticed that the strategies are simple and may not be worth your attention, but if you insist - I will give you all my xml-ki.

I can send you all my xmls if you insist. By the way, in my opinion, the program easily generates them as it is. But checking in the demo, in real time is a real hassle, especially for 4-hour charts. A good strategy is easily distinguished by its parameters. The different periods of testing, the presence of profits, profitable trades>losing trades, the average transaction more than 30 pips, etc.

I don't understand why "unfortunately" :) but if you have a serious real experience, it's another matter. I don't know why, but if you have real experience it's a different story. My e-mail will send you in a personal message.

 

zfs писал(а) >>

Reshetov >>:

Results of strategies by a competing analyzer: Results of SSB strategies


In contrast to the

examples given by you, the results given by me contain attached files with strategies, uploading which to SSB, any dummie in programming can get the codes of EAs in MQL4 in a fraction of second, as well as the results of tests on these very EAs on historical data.

I understand this is publicity. Sorry, Yuri, what's wrong... where do you get your Stoke, by the way? And how much?

1. Take the above. This is not just advertising, but vile spam and vicious off-topic with the aim to make a quick buck and escape to the Canary Islands.

2. Nowhere to be taken, as all copies have been destroyed, the screws unformatted, crushed by a hydraulic press and sunk in the Pacific Ocean above the Mariana Trench itself. All witnesses have been removed.

3. No money is enough. When the price was called, W. Buffett, J. Sorros and B. Ghaith all fell off their stools and said in one voice that they could not afford it, even together.

 
Reshetov писал(а) >>

It's not just advertising, it's vile spam and vicious off-topic to make a quick buck and flee to the Canaries. ...

... У. Buffett, J. Sorros and B. Gate... said they couldn't afford it, even in a joint venture.

:) :) :) Cool!

zfs, I'm richer than the aforementioned tycoons! I mean, I took it and use it. (The thing!!!)

Yuri, will you take into account your wishes about ssb looping? (See personal info).

 
Reshetov >> :

1. take it higher. It's not just advertising, it's vile spam and vicious off-topic to make a quick buck and flee to the Canaries.

2. Nowhere to be taken, as all copies have been destroyed, the screws unformatted, crushed by a hydraulic press and sunk in the Pacific Ocean above the Mariana Trench itself. All witnesses have been removed.

3. no amount of money would be enough. When the price was called, W. Buffett, J. Sorros and B. Ghaith at the same time knocked their heads on the floor and said in one voice that they could not afford it, even together.

Well let's say I downloaded it, where's the instruction manual, because Buffett and I can't figure it out.

 
zfs >> :

Well let's say I downloaded it, where's the instruction manual, because Buffett and I can't figure it out.

Sorry, I found it... how do I get out of Metatrayder now... need a 2nd computer...

 
zfs >> :

Sorry, all found... how do I get out of Metatrader now... need 2nd computer...

There is no way you can get away with two computers to log out of MetaTrader, don't even dream. The only way out is with 583 computers connected to each other via fibre optics.


On one computer only Reset can work, and not always, but only if the button hasn't been eaten by a moth.

 
Reshetov >> :

There is no way you can get away with two computers to log out of MetaTrader, don't even dream. The only way out is with 583 computers connected to each other via fibre optics.


The only way to exit is via Reset, not always, unless a moth eats the button.

Yuri, you're sarcastic, not everyone is as smart as you are... a month ago I wished I had an analyzer... now I have two :) Tried your brainchild... I can of course pay you all sorts of compliments...! applause...! but I think some constructive criticism won't hurt. The program gave me the code for the hour hand, using only 2 tools - mcdee and ssiay without any stops, profits... working 100% in the market, so to speak. And it gave me a staggering result... The problem is that I used a limited set of (unsmoothed too) tools and found the periods of trend reversal on history, so it can be an excellent filter for entering the market, but alas, it can't be a strategy... Maybe I missed something... I drew my conclusions from a single code. Maybe it's more relevant to a 15 minute chart.

 
zfs >> :


...But I think constructive criticism won't hurt. The program gave me a code for hourly, using only 2 tools - makdi and sisii, without stops, profits... working 100% in the market, so to speak. And it gave me a staggering result... The problem is that I used a limited set of (unsmoothed too) tools and found the periods of trend reversal on history, so it can be an excellent filter for entering the market, but alas, it can't be a strategy... Maybe I missed something... I drew my conclusions from a single code. Maybe it is more relevant to a 15 min chart.

It's a matter of habit, because many people assume the strategy:


1. It should not be constantly in the market, and therefore it is necessary to install all sorts of filters of false trades

2. stopLosses, takeprofits, or trawls (according to taste) should be present in the strategy

3. The number of indicators and oscilloscopes used should go over the top of all imaginable and even unimaginable limits. I.e. the more complex and sophisticated, the better.

And so on and so forth.


I personally have a different point of view, at least regarding the process of strategy selection. Namely, when a strategy is selected on historical data, it is advisable to:


1. That the strategy is constantly in the market and the exit signal is an entry signal in the opposite direction, i.e., only reversals. The use of filters that supposedly "filter out" false trades in the strategy selection process is self-defeating. Filters, in most cases, reduce the reliability of the trading system, as the market is not stationary and their selection on historical data in most cases is a fitting.

2. The strategy should not have a hedge in the form of Takeovers, Losses and trawls. All of this can be added later according to taste and colour, and even then, only in case of extreme necessity.

3. the simpler the strategy, the more effective according to reliability theory.

4. When selecting a strategy, trading should be a constant lot, no MM etc. Martins, otherwise it is self-defeating. Capital and risk management can be added later, if needed. Also, do not open multiple positions on each confirmed signal, as this is a type of MM.

5. In strategy selection, no optimisation of the input parameters can be done, not even limited. Optimisation + fitting = fitting squared. After fitting, optimization can be done, and that only if it has a positive effect on the forward tests.


It is clear that if a strategy has gone through fire, water and brass pipes without any insurance, protection, limitations, tips and other customary supports, and still shows positive results, then it should at least be worth paying attention to. But the so-called "strategies" consisting of only crutches, supports and guards with trading signals barely noticeable in the background are not even worth looking at.
 
Reshetov >> :

This is a matter of habit, because many people assume that the strategy:


1. It should not be constantly in the market, and therefore it must be equipped with all sorts of false trade filters

2. stopLosses, takeprofits, or trails (according to taste) should be present in the strategy

3. The number of indicators and oscilloscopes used should go over the top of all imaginable and even unimaginable limits. I.e. the more complex and sophisticated, the better.

And so on and so forth.


I personally have a different point of view, at least regarding the process of strategy selection. Namely, when a strategy is selected on historical data, it is advisable to:


1. That the strategy is constantly in the market and the exit signal is an entry signal in the opposite direction, i.e., only reversals. The use of filters that supposedly "filter out" false trades in the strategy selection process is self-defeating. Filters, in most cases, reduce the reliability of the trading system, as the market is not stationary and their selection on historical data in most cases is a fitting.

2. The strategy should not have a hedge in the form of Takeovers, Losses and trawls. All of this can be added later according to taste and colour, and even then, only in case of extreme necessity.

3. the simpler the strategy, the more effective according to reliability theory.

4. When selecting a strategy, trading should be a constant lot, no MM etc. Martins, otherwise it is self-defeating. Capital and risk management can be added later, if needed. Also, you should not open multiple positions on each confirmed signal, because this is a type of MM.

5. When selecting a strategy, no optimization of the input parameters can be done, not even a limited one. Optimisation + fitting = fitting squared. Optimisation can be carried out after fitting, and then only if this has a positive effect on the forward test.


It is clear that if a strategy has gone through fire, water and brass pipes without any insurance, protection, limitations, tips and other customary supports, and still shows positive results, then it should at least be worth paying attention to. But the so-called "strategies" consisting only of crutches, supports and guards, and the trading signals barely visible in the background, do not even make me want to look at them.

And I agree with you, Yura. Probably 100%.

The strategy should be simple - that's for sure.

We can build pyramids fixing previous profit with a stop, please note, but this is a subtlety.

I don't understand the 5th point, but what has your program done? Was it not optimizing?

There are only 6 tools. Or maybe I didn't understand something.

Well, your analyzer is interesting, of course, and it's even more interesting to develop it further.

After all, we don't want 200% per year - we want 20000% per year :)

Besides such strategies are unlikely to raise my small deposit.

After all, the main thing is not to lose.

!---- applause !---!

 
zfs >> :

And I agree with you Yura. I agree with you, Yura. Probably 100%.

The strategy should be simple - that's for sure.

You may build a pyramid by fixing previous profit with stop loss, please note, but this is a subtlety.

I don't understand point 5, what was your programme doing? Was it not optimising?


There is no optimisation in the programme. It is strategy selection, i.e. for each oscillator, the algorithm tries only three discrete states: direct signal, reverse signal or no signal at all. Optimization is selection of input parameters and, in this case, what has been used as input parameters for selection is absent in the source code of the ready Expert Advisor. And what has not been used for the selection of a strategy is left as input parameters for the Expert Advisor. Therefore, the Expert Advisor can be further optimized if necessary.


zfs >> :

After all, we don't want 200% per year - we want 20000% interest per year :)

Dream on.