Counsellors to whom. Lots of them and for free! - page 10
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 understand the reason for the oscillator not showing the actual picture. It turns out that some of the signals are floating. After all, a number of parameters change during the whole period, the signal disappears and disappears with every tick. So it turns out that what the tester displays differs significantly from the real one. Consequently, profitable Expert Advisors can become unprofitable. The solution is not to analyze the current bar, but the previous one. Otherwise it will be a lot of lyricism. The selection of the signal in a real trading EA will be almost random, and with a large number of factors and minimum trades will be comparable to roulette. False data will generate stochastic, AO, AC and probably Ichimoku values.
I understand the reason for the oscillator not showing the real picture. It turns out that some of the signals are floating. Indeed, a number of parameters change throughout the period, the signal fades, then disappears with every tick. So it turns out that what the tester displays differs significantly from the real one. Consequently, profitable Expert Advisors can become unprofitable. The solution is not to analyze the current bar, but the previous one. Otherwise it will be a lot of lyricism. The selection of the signal in the real trading EA will be almost random, and with a large number of factors and minimum trades will be comparable to roulette. False values generate stochastic, AO, AC and maybe Ichimoku.
To avoid floating readings, most of the indices and oscillators in SSB are set to open prices. For them the price fluctuations within the bar do not play any importance. Moreover, Expert Advisors in the output are also tuned to trade at the open prices.
However some other oscillating indices and oscillators may change their values inside the bar, because their algorithms calculate closure prices and/or maxima and minima:
1. Ichimoku
2. Stochastic
3. AC
4. AO .
To avoid a large number of trades, we introduced a filter in SSB4, according to which TS with less than 100 trades on the history are not even considered and ranked by the program. Since the final choice of TS in SSB is made by a user, he/she has to decide whether a strategy is potentially fit or not.
After all, no one forbids running additional forward tests for SSB-derived Expert Advisors and testing them on demo accounts for more confidence.
The result is a complex checking of all TSs using R. Pardo's method
1. Using the SSB repository, i.e. checking by time, on different financial instruments and timeframes using distributed computing.
2. After SSB, forward and other tests.
Some of the Expert Advisors will pass these very tests and be considered suitable for auto trading or semi-manual trading. A considerable part of the TS will be filtered out and found inadequate at the stage of automatic (SSB) or manual (user) testing.
There are no miracles. No one gave a guarantee that all TCs obtained via SSB are super profitable. SSB does only the most routine part of the work of rejecting the inadequate TS, which saves a lot of time and nerves for users. The automatic creation of Expert Advisor codes according to the obtained and tested TS allows to avoid mistakes inherent in manual programming and debugging.
The whole process, from the formulation of a potentially adequate TS to getting an Expert Advisor tested by this or that TS, is considerably reduced in time.
And if you are not satisfied or not satisfied with the SSB, you can run this whole process manually from beginning to end. No one is forced to do this.
But a small part of the indices and oscilloscopes can change their readings within a bar, because their algorithms contain calculations on closing prices and (or) on highs and lows:
The result is a comprehensive check of all the TS according to R. Pardo's methodology:
If you are not satisfied or not satisfied with SSB, you can run this whole process manually from beginning to end. Nobody will be forced to do it.
Maybe you, Mr. Reshetov, are involved in luring money from traders and have a conspiracy with DCs?
For the system to work it must rely on static data, to avoid this the system for these indicators must refer to the previous bar!!!
"Incorrect optimization can lead to tweaking and other serious errors. If you overlook these errors, the resulting trading model will show very good results in the optimization process and very poor performance in real trading"(c) Pardo
This is what causes the good results of your systems. Let's correct an algorithmic error. Referring to the zero bar is just adjusting instruments to historical data and has nothing to do with trading systems.
Maybe you, Mr. Reshetov, are involved in luring money from traders and have a conspiracy with DCs?
For the system to work it must rely on static data, to avoid this the system for these indicators must refer to the previous bar!!!
"Incorrect optimization can lead to tweaking and other serious errors. If you overlook these errors, the resulting trading model will show very good results during optimization and very poor performance in real trading"(c) Pardo
This is what causes the good results of your systems. Let's fix the algorithmic error. The reference to the zero bar is just a fitting of tools to the historical data and has nothing to do with the trading systems.
Of course, all I do is systematically swindle money from already poor traders in close collusion with the most vile kitchens. You can rest assured of that.
For these nefarious and very selfish purposes, I purposely introduced into SSB the algorithmic error you identified, which uses for TA readings on a zero bar, and not a thousand and even less a hundred-thousandth bar. Additionally, for more mercantilism I chose MetaTrader4 as a trading platform, because only this terminal shows "An improper optimization, which may lead to tweaks and other serious errors" as stated by (c) R. Pardo.
For that very reason I will not fix the error and will not change the trading platform, even if you don't ask, beg or persuade me to. Otherwise what good will it do me?
For this very reason I will not correct the error. What i will profit from it otherwise?
Only the strong of the world can fall, then get up and climb to a higher peak. But the strong of all things know how to get rid of negative emotions.
In FSB, for example, there is a reference to the previous bar. And this is not done because this bar is more informative.
The idea is still very strong.
New tools must be added and then no doubt profitable strategies will be found.
Though I have corrected it - strategies remain profitable, in spite of these errors. But not all of them.
Only the strong of the world can fall, then get up and climb to a higher peak. But the strong of all things know how to get rid of negative emotions.
In FSB, for example, there is a reference to the previous bar. And this is not done because this bar is more informative.
The idea is still very strong.
We must add new tools and then we will surely find profitable strategies.
Although I have corrected - strategies remain profitable, despite these errors. But not all of them.
I'm not going to change anything to suit the whims and lusts of individual fate dissatisfied users. The SSB has already taken sufficient measures to deal with fitting. There is still no way to avoid fitting 100%, as the financial markets are inherently unsteady and even more so at the moment due to the global crisis. Someone offering 500 pieces of useless advice based on the fable "Quartet", i.e. to swap the bastard for the bastard, is not going to change the results. You cannot change horses ten times a day, because SSB uses distributed computing through a single repository and all strategies in that repository have exactly the same ranking algorithm.
The moral of this fable is this: you guys are not good musicians.
Once again, for the especially gifted, who do not like it, can not use SSB, but switch to, for example, FxSB, or indulge in their own developments and techniques. Using one programme or another is purely voluntary.
Once again, I repeat for especially gifted people who do not like it, can not use the SSB, and go to, for example, FxSB or indulge their own development and techniques. Using this or that program is purely voluntary.
Yes, I use both products, I would use the third one too.
The results really don't change much, because close practically=open.
Problems arise in real trading, it's not hard for me to change it. This is more of a problem for beginners.
Still, this forum is dedicated, hopefully, to solving these issues, and I think the lifetime of version 4 is not eternal. The 5th version should take into account all nuances and disadvantages of the previous ones and why not discuss innovations in the new version. After all, although SSB users are capricious and demanding, and the author of SSB does not constructively accept any criticism, I think everyone is interested in making the product more effective.
So what is this forum for, Mr. Reshetov?
Probably for gratitude.
Thank you!!!
and I don't think version 4 will last forever. And the 5th version should take into account all the nuances and shortcomings of the previous ones, and why not discuss innovations in the new version. Though SSB users are really picky and demanding and the author of SSB does not accept any criticism constructively, I think everybody is interested in making the product more effective.
When there is constructive criticism rather than 500 useless tips, then it will be possible to cooperate in a constructive way.
The 4th version of SSB will not last forever, since I am a user of this software and in my spare time I try to improve it. I really have much more free time now, because I managed to switch a significant part of trading to autotrading and the trading stability has noticeably improved.
But at the moment there is nothing concrete in terms of rationalisation with any tangible result, because everything is hindered by the non-stationarity of financial instruments. It is impossible to defeat unsteadiness because it is independent of traders. Also it is impossible to bypass it, because market participants create it themselves to cheat and lure money from other participants. The introduction of additional indicators and oscillators is also meaningless, because they only add to the already considerable information redundancy to the readings of existing indicators and oscillators. The characteristics of strategies improve only on the fit, and on the forward tests they deteriorate noticeably.
Maybe the non-stationarity limit of this system has already been reached. Maybe something else that really improves the efficiency can be found? It is difficult to say anything definite yet. I.e. the search continues, but no appreciable progress has been observed so far.
At least the results of the current algorithm of strategy ranking in the repository are satisfactory. But I would like something better.
Use of this or that program is a purely voluntary matter.
Totally agree. :)
I also want to note, that as far as I understand, if indicator value is always calculated in the tester only at bar opening (and does not change for this bar anymore), the generated strategy will work correctly, although indicators on the chart may show different values from the tester.
But it is good, if for example I know this for a long time and take it into account in my code. However, very beginners may not understand these features and get "unplanned" losses. That is why it makes sense to warn about them. Or discuss it somewhere (why, by the way, not in this thread?)...
Because you yourself, Yuri, from time to time make comments on other programs and experts. For example, with MT4 (remember, at least incorrect processing of orders in the tester), or FxSB (in particular, the Fibo indicator), etc.. And your comments are needed! Nobody is saying to you - "Well, if you do not like MT4 - then do not use it"... :)
Maybe already reached the limit of this system in non-stationarity, and maybe something else that really improves the efficiency can be found? It is hard to say anything concrete yet. I.e. search proceeds, but visible progress is not observed yet.
Probably because you continue your search purely by yourself and your wishes to SSB are inadmissible? :)
When we get constructive criticism instead of 500 useless tips, then we can cooperate in a constructive way.
And could you clarify what you consider to be constructive criticism? (I'm quite serious.)
But if this is the way you usually criticise everyone..... :) (Not serious anymore...)
But at the moment there is nothing concrete in terms of rationalisation with any tangible result, because everything is hindered by the non-stationarity of financial instruments. It is impossible to defeat unsteadiness because it is independent of traders. Also it is impossible to bypass it, because market participants create it themselves to cheat and lure money from other participants. The introduction of additional indicators and oscillators is also meaningless, because they only add to the already considerable information redundancy to the readings of existing indicators and oscillators. The characteristics of strategies improve only on the fit, and on the forward tests they deteriorate noticeably.
Maybe the non-stationarity limit of this system has already been reached. Maybe something else that really improves the efficiency can be found? It is difficult to say anything definite yet. That is, the search continues, but there has not been much progress so far.
At least the results of the current strategy ranking algorithm in the repository are satisfactory. But I would like something better.
In my opinion, the market is changing and the need to find new strategies will never disappear. It is not only speculators in the market and there are more methods to influence the market in the modern world, so it will always be possible to swindle money, but it is just a matter of being more sophisticated. I'm trading on stock market in parallel and I really see how much more professional and intelligent forex is. And here you need powerful tools.
It is not really possible to change anything, because otherwise we would lose the ability to test at the opening prices. The introduction of other instruments will not be redundant, but it is necessary to limit the maximum in one system to e.g. 12. There are many more interesting tools than the standard ones, not to mention the paid ones, which are designed for more efficient trading. There won't be an overabundance here, you can limit the tester to a random set of samples to reduce the performance. Or you could assign instruments a rating.