Discussion of article "Extract profit down to the last pip" - 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
Coming back months (almost years now) later to the article, 2 questions remain:
1. Why did NOKSEK come to mind and what is it remarkable for?
2. Up to the section "Etching of jewellery" everything is written quite detailed and clear. Just take it and do it step by step. Here you have run the genetics for 3 months, you have chosen good results in terms of profit, you have run with your hands on the whole history, you have made sure that it looks decent. But then the clarity fades. Were you running a complete oversample or several times genetics over the entire interval from the largest drawdown to today? Did you just select these best passages from there, or was something else done?
Basically it doesn't matter if it's according to the article or how it's already done now with all the changes. Was it possible to formalise or what is the algorithm of actions? I.e. to choose from a pile of exactly those that will go into battle, when there is a symbol and TC in hand, on which hopes are pinned.
traveller00:
Coming back after several months (almost years) to the article, there are still 2 questions:
I don't remember the content of the article anymore, a lot of water has passed.
In principle, it is not important, according to the article or as it is already happening now with all the changes. Was it possible to formalise or what is the algorithm of actions? I.e. to choose from the pile the very ones that will go into battle, when there is a symbol and TC in hand, on which hopes are pinned.
Of course, like everyone else, new tools, approaches, etc. could not have appeared during such a period of time. To discuss them would require a new article to be written.
Some studies for quite a long time I can't bring to a fighting state. A lot of time is spent on questions about the trading platform.
It will be possible to say that I was able to make progress without empty shaking of the air after some results of a combat robot are revealed. For now, everything is at the level of the MO branch.
Is it worth waiting for a new article? Personally, I would appreciate reading it.
Maybe you could at least outline the basic methods, principles and actions. There is a symbol suitable, there is a TC and belief that it is good on this symbol. What do we do next in order to choose exactly those N Expert Advisors that will be used in real trading? I.e. describe the very struggle for MO for future generations to be able to repeat. And issues of robustness are of interest here, of course. I.e. what has been done (and has anything else been done additionally at all?) to make these N last longer?
The grit that ruins everything. White swans in history and in reality.
When optimising the TS, it is possible to run into such situations.
When optimising, I find trends in history and run against the trend, i.e. when rising only sell and vice versa, all plums, almost ).
Is it worth waiting for a new article? Personally, I'd appreciate reading it.
Maybe you could at least outline the basic methods, principles and actions. There is a symbol suitable, there is a TC and belief that it is good on this symbol. What do we do next in order to choose exactly those N Expert Advisors that will be used in real trading? I.e. describe the very struggle for MO for future generations to be able to repeat. And issues of robustness are of interest here, of course. I.e. what has been done (and has anything else been done additionally at all?) to make these N last longer?
I'm poking around like a blind kitten, so no secret knowledge. There's just nothing to pass on. Here's a steep chip off my ideas about faith today.
I'm thinking, what the hell? There are a thousand rollover trades on Training, no white swans, PF > 3. And most importantly, the implicit optimisable parameters are less than a dozen. I mean, it's not even an MO model. Therefore, wild overfitting is excluded - it is a regularity! I'm looking at OOS - bummer.
And so I threw a lot of similar cases into the basket. And then I foolishly decided to look at OOS more, and there it is.
What is this short PrevOOS interval (this is OOS from the first screen), which forced to fill the trash from the TC? And most importantly, I still didn't understand whether it was rubbish or not.
Here's a silly question: isn't it good to have an OOS equal to Training, or better yet bigger by a factor of 4?
Wasn't a big fan of the OOS. It's not like they're scrapping something that's trading well right now.
What was that short PrevOOS interval (it's OOS from the first screen) that made the trash from TC fill up? And most importantly, I still didn't understand whether it was rubbish or not.
And what are the dates there?
What are the dates on there?
They are the stormy two months of March and April 2020. These two months have almost nothing (market patterns) in common with what came before and after. For that reason, it is reasonable to exclude them completely from analyses, including MO studies. There was a klondike there for the brazen.