Discussion of article "How to choose an Expert Advisor: Twenty strong criteria to reject a trading bot" - page 2

 
Hely Rojas #:

A good expert advisor software should be like the windows calculator. It does not require support from the programmer, its results are the same in Win95, Win98, XP, etc. I agree with the opinion of the author of the article.

Vendor: I am selling you a Nut. It is the best on the market. But it only works on brass screws. Oh... and for best performance, you should use a pipe wrench of my own making. Sometimes when the weather is humid, the nut rusts, but don't worry, if you contact me, I can advise you to clean it up and make it functional. ATT: Seller001

I just canceled my Netflix . This is better 

Agreed , and when you buy bread it should taste good no matter what you put in the sandwich otherwise its not good bread !

Put cheese mayo nutella fishsticks , it does not taste good ? Bad bread!

Put a bread on the bottom a bread on top and an iphone in the middle . What you broke your tooth ? Bad bread!

Buy pizza put in in the bread fry it dip it in butter , add mayo , eat it . What you gained weight ? Bad bread!

 
Cristian Mihail Pauna #:

Yes, indeed! This article was wrote for the software buyers. Thank you for mentioning, this is an important observation!

If we draw a circle , professor , angle 0 and angle 360 are on the exact same point . What do i mean (or try to imply) with this ?

Are there scammers here ? ow yeah 

Does that mean we should strikeout (strikeout -> method)  an approach or a method because it was used by a scammer or because our anecdotal (means few examples , not the Greek meaning which is ridiculous) evidence reports failure.

Even worse , should we completely write off a method or approach just because a scammer claimed that's what they were using ?

So , let's pickup a random point , the changing a parameter should not lead to failures . Okay , if a vendor internalizes all the parameters and slaps a neural net on the ea (which i will agree it is not easy to tell whether something is a scam or has internal "AI" , that is sad and true but also difficult to address by Mq because if you cut the scams off you cut the "AI" off as well) , does that fall under the same classification ? Is it a scam ? What if you have an internal variable in your products , are those not features ? Why do these variables have a specific value would you change the variables ? 

Now you will probably complain or already have about my counter arguments to the administration but let me assure you , the SEO of this site is stellar that means : 

You will not lose any traffic to your shop with these comments . In fact having your opinion posted on an official article will pass some of that search engine juice down to your profile and shop.

So all is well . Have a good weekend 

 
Lorentzos Roussos #:

i bet the author is describing 20 things his products don't have . 

  • the most interesting point i found was 11 , clearly , they don't teach risk reward wherever he got his PhD 
  • 3.we can clasify news in two categories , expected , and , unexpected , the author claims filtering for none is better than filtering for the expected at least .
  • 5.If it has array out of range it will not "No, it is still running,"
  • 6.If the broker decides to change their server time you will change the time parameter in the bot
  • 3." These kinds of bots are sensitive to big event news and can generate losses in unprecedented cases" , no , all bots are sensitive to big event news and can generate losses 
  • 4.if the author updates the bot isn't the act of downloading the update " If a trading bot asks for your action from time to time" ? 
  • 10.there are no such bots on this market 
  • 17.It is impossible for all parameters to be profitable (18 same)

I personally do not believe in high win rate. I think a real forex business relies on low win rate and high RR ratio.

But: In case you want to subscribe to a channel you don't have a lot of time to wait and see when the results of the bot converge.

 

Please publish an article about comparing existing profitable EAs in market that you tested.

thank you

 
b.mohammadi #:

Please publish an article about comparing existing profitable EAs in market that you tested.

thank you

This cannot be done as it would effectively be promoting the EAs that do the best.

 
Yashar Seyyedin #:

I personally do not believe in high win rate. I think a real forex business relies on low win rate and high RR ratio.

But: In case you want to subscribe to a channel you don't have a lot of time to wait and see when the results of the bot converge.

Yeah , in theory any trading system is a subset . If we ask "is this trading strategy improving within the sample we have or is it getting worse" do we stop there ?

I mean even in building an ea , the one he published one day before the article was approved for example , the op did some selection .

He can't have taken a strategy as is and expected it to work. That would be random .

That is my disagreement with this article . The author criticizes eas that don't work if you adjust the settings as you please , and i'm saying , what if you hide the settings ?

Is this a good ea because the user cannot mess it up ? Clearly his eas have variables and parameters , if he exposes them to the user , and then the user adjusts them and the ea performs worse is it a bad ea ?

My 2 cents

 

This article is only a METHOD to select between thousands of available EAs on the market.

Not anyone must be agree with a method. For me, this method is working perfectly anytime.

--
And one more thing: I didn't paid to publish this article.
The site paid me to publish this here, more than 2 cents!

 
Cristian Mihail Pauna #:

This article is only a METHOD to select between thousands of available EAs on the market.

Not anyone must be agree with a method. For me, this method is working perfectly anytime.

--
And one more thing: I didn't paid to publish this article.
The site paid me to publish this here, more than 2 cents!

Exactly , thank you for clarifying . 

PS : You should also edit the description of your latest ea , here are some tips 

 

I'd like to ask about this " 11. Reject if there are more negative than positive trades! " . Usually, I agree, this is a rule for a trader doing scalper. Do you have a swing rule? What percentage of positive trades would you expect? 35% to 50%?

 
Ricardo Rodrigues Lucca #:

I'd like to ask about this " 11. Reject if there are more negative than positive trades! " . Usually, I agree, this is a rule for a trader doing scalper. Do you have a swing rule? What percentage of positive trades would you expect? 35% to 50%?

Hello, I usually accept an EA in my portfolio only if it generates at least 80% positive trades in the last 3-5 years.