Is this the Grail? - page 7

New comment
 
paukas:

That's the right one, it's more costly under a rolling stone!

f.t.:, By the way, the gravity meter gave me 66%. Is that good or bad?

"The script calculates the number of trades similar to grailing and normal trades and displays the analysis results. The higher the value in the "Looks like grailing" section and, consequently, the longer the bar of vertical dashes - the more the trade looks like a grailing trade."
 

hmmm... 16 trades!!! or even 6!!! what are we talking about? it's not a system - it's just a few trades :)))))

I don't see any order lines on your screenshots - maybe you "wrongly" launched the script (not in the window of a running EA with results), or you ran it before the end of testing? Or you deleted all order open/close marks? The lines of all orders must remain on the chart after testing is over - the analysis is performed by them, since we do not have access to the history of deals of the test that has been finished.

 
paukas:

f.t.:, By the way, the gravity meter gave me 66%. Is that good or bad?

It means that 66% of your trades are "potentially gray". They might have been triggered not by real ticks but by simulated ticks, which have nothing in common with real ones, except OHLC values. In reality, they would be triggered at completely different prices, so it's not certain that the result is the same as in the tester. It could have been either better or (more likely) worse. The most "accurate" results can only be obtained on minutes when modelling "at opening prices".
 
"Potentially grave" is apparently a bad thing......
 
f.t.:
This means that 66% of your trades are "potentially gray". They may have been triggered by simulated ticks, which have nothing to do with the real ticks, except the OHLC values. In reality, they would be triggered at completely different prices, so it's not the fact that the result is exactly the same as in the tester. It could have been either better or (more likely) worse. The most "accurate" results can only be obtained on minutes when modelling "at opening prices".

I ran it on real trades, not tester trades, on the clock. On the minute chart it shows 89% on the same trades.

I have a suspicion that the script isn't calculating the right thing.

 
f.t.:

hmmm... 16 trades!!! or even 6!!! what are we talking about? it's not a system - it's just a few trades :)))))

On your screenshots, I don't see any order line - maybe you started the script incorrectly (not in a working Expert Advisor window with results), or ran it before the end of testing, or deleted order open/close tags? The lines of all orders must remain on the chart after testing is over - the analysis is performed by them, since we do not have access to the history of deals of the test that has been finished.


The deals will not be shown on the TF M1. Are 30 trades not enough? In general, I do not understand: why does the script show drastically different results on TF D1 and M1?

And about the system I have one criterion - that it should earn well and not lose. Even if it's 1 trade a day.

 
paukas:

I ran it on actual trades, not tester trades, on the clock. On the minute chart it shows 89% on the same trades.

I have a suspicion that the script does not calculate it correctly.

The script is not intended for estimating real trading! It is an evaluation of the results obtained in the "tester". Well I gave you the link there all described what the script counts - the number of trades made inside the BAR. It's these trades (in the absence of ticks or at least minute data inside them) - it's not testing but emulation. In real life - everything is honest there is nothing to measure. But in the tester - not a fact, rather unlikely, and how unlikely - the script shows.

The percentage of deals performed completely inside the bar can be considered as an estimate of the Expert Advisor's graality: the more deals there are inside the bar, the more similar the Expert Advisor is to the tester's grail.

 
Fullness:


Aren't 30 trades enough?

Sorry, but for me it's not "not enough", it's nothing at all:(

In general, I do not understand: why does the script produce drastically different results on TF D1 and M1?

See the answer to the previous post ;)
 
f.t.:

Well, I gave you a link where everything is described that the script counts - it is the number of trades made inside the BAR. It is these trades (in the absence of ticks or at least minute data inside them) that are not testing but emulation.


This is what I am talking about. On H1 it shows a lower figure than on M1.

If it's counting intra-bar trades it should be the other way round. That's weird.

 
paukas:

That's what I'm talking about. On H1 it shows a lower figure than on M1.

If it's counting intra-bar trades it should be the other way round. It is strange.

and the length of history on H1 and M1 is the same? Try to set the same testing period for both charts so that the number of deals is the same and they occur in the same period on both charts.

i.e. before buying. you can count your ownintra-bar trades. But why count them? :))) in the tester you just need to avoid them. ;)

123456789
New comment