Statistics, optimisation and "lucky coin" .... - page 6
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
A little correction to.... Question of the thread: meaning of statistics and value of TS optimization in trading. I question the absolutism of obtained statistical data in determining the robustness of the TS, because I tried to show that "good" statistics can be obtained in any way for any "Trading idea" when increasing the parameters to be changed.
...
The point is that "good" statistics cannot be obtained on the RNG. As a simple example, take any strategy. Put it into the optimizer and start sequentially going through the parameters. Start measuring the total return of each outcome. If the strategy is fit, it will give a plus on half of its parameter space and a minus on the other half. But the total result will be zero - which means it cannot be traded. But if the strategy is working, even on a large parameter space, we can find stable clusters that will shift the overall result into the positive zone. We can go further and build for example tests for stability of shadows of multidimensional clouds (a step forward in the concept proposed by Robert Pardo). And in this case you will observe interesting effects untypical of any RNG.
The point is that "good" statistics cannot be obtained using LFG. As a simple example, take any strategy. Put it into optimizer and start subsequently searching through the parameters. Start measuring the total return of each outcome. If the strategy is fit, it will give a plus on half of its parameter space and a minus on the other half. But the total result will be zero - which means it cannot be traded. But if the strategy is working, even on a large parameter space, we can find stable clusters that will shift the overall result into the positive zone. We can go further and build for example tests for stability of shadows of multidimensional clouds (a step forward in the concept proposed by Robert Pardo). And in this case you will observe interesting effects untypical of any RNG.
to continue what I've said:
and the first Expert Advisor I found on the market:
experiment is not clean, optimizing balance max, just looked that by sequential search for three million combinations)
But it is clear from the upper chart that the Expert Advisor does not really care about combinations of parameters, the positive outcomes are stably much higher than the negative ones even at the beginning of optimization, which can't be said about the second strategy.
to continue what has been said:
and the first Expert Advisor I found on the market:
experiment is not clean, we optimize balance max, I just saw that there are three million combinations by successive search)
But it is obvious from the upper picture that the Expert Advisor does not really care about combinations of parameters, the positive outcomes are consistently much higher than negative ones even at the beginning of optimization, which cannot be said about the second strategy.
For my part, I can add that if the optimizer is skillfully handled (it is the optimizer), we can easily find any Expert Advisor whose source code is not available and whose forwards are not (for example, all EAs from Market).
Off topic again =(
The topic starter asked if it's possible to make money with a coin... And everything was reduced to statistics...
FOREX statistics (!) is the past. It does not exist and will never exist. Why do you need statistics? Test it in the real world.
Off topic again =(
The topic starter asked if it's possible to make money with a coin... And everything is reduced to statistics...
FOREX statistics (!) is the past. It does not exist and will never exist. Why do you need statistics? Test it in the real world.
The point is that "good" statistics cannot be obtained on the MSG. As a simple example, take any strategy. Put it into optimizer and start consecutively working through the parameters. Start measuring the total return of each outcome. If the strategy is fit, it will give a plus on half of its parameter space and a minus on the other half. But the total result will be zero - which means it cannot be traded. But if the strategy is working, even on a large parameter space there will be stable clusters that will shift the overall result into the positive zone. .....
Take any strategy and start to go through the parameters one by one. If we are not satisfied with the total profit of each outcome, we start adding new parameters. Isn't it kasher? Why not? If we initially take a strategy and start optimizing it, it's because we are not entirely confident in its initial "goodness". If additional parameters are needed to improve it, then methodologically nothing prohibits adding these parameters. So we can get a strategy that uses a lot of different parameters, which will give us any positive result in the first half, in the second half, and even in the third half....:)...... As a result, we get a simple fit (or otherwise, we need to claim to have discovered a "Universal Market Formula", colloquially referred toas the "grail" ....).
I.e., you have finally come to realise the futility of finding a marketable pattern.... :)
Without any banter: there is a certain generally accepted idea, which is discussed on various forums, it is talked about at lectures for young traders, books are written, it is applied in practical activities, and everything like that....... I.e., it is believed that every competent trader must use this idea, and the most interesting thing is that this idea is really quite popular... I have tried to build some sort of logical construct that questions this idea. I'm not trying to sell anything, nor am I looking for support for solutions to any problems... :) I am interested in the same logical arguments of the adherents of this idea in its defence. Unfortunately, apart from almost religious indignation, there is very little constructive evidence (and if there is, there are so many logical inconsistencies...). It turns out that the application of an idea is not based on its comprehension but on its "general acceptance"?
It may sound categorical, but everything they say in lectures and books is utter nonsense. I do not deny that once it could bring profit, but markets change, sometimes invisibly, but quite noticeable for this or that strategy. Without going into too much detail about what happens to a trading idea when it becomes publicly available, one could say that the market, as the number of participants tied to a particular pattern increases, simply grinds down those areas of inefficiency, literally turning them into information noise. Anyone who wants to make money in the market for any length of time has to keep his findings secret. So anyone who uses "generally accepted" ideas is either cheating himself, not understanding what's really going on, or is leaking. Either deceives AND leaks. Or in order to make the right impression on others pretends that all is well (of course, the generally accepted system, the classics!) - You have to avoid such people a mile away.
It may sound categorical, but everything they say in lectures and books is bullshit. I don't deny that at one time it might have been profitable, but the markets change, sometimes invisibly, but quite noticeably for one strategy or another. Without going into too much detail about what happens to a trading idea when it becomes publicly available, one could say that the market, as the number of participants tied to a particular pattern increases, simply grinds down those areas of inefficiency, literally turning them into information noise. Anyone who wants to make money in the market for any length of time has to keep his findings secret. So anyone who uses "generally accepted" ideas is either cheating himself, not understanding what's really going on, or is leaking. Either deceives AND leaks. Or in order to make the right impression on others pretends that all is well (of course, the generally accepted system, the classics!) - You have to avoid them a mile away.
It is good that MT has a tester, and any bullshit can quickly check whether it works or not.
Good thing MT has a tester, and any bullshit can be quickly checked to see if it works or not.
That said, the topic of testing is strikingly persistently avoided in books a la "this is how to make money in the market").