Data Quality - page 2

Rogerio Figurelli
Moderator
61343
Rogerio Figurelli  
ssn:
Thanks for the replies so if am not using M1 price data I should not read much into the history quality
Note that using M15, H1, etc., you also can reach 100% test quality, if you enable tick by tick simulation option.
Alain Verleyen
42249
Alain Verleyen  
figurelli:
Note that using M15, H1, etc., you also can reach 100% test quality, if you enable tick by tick simulation option.
History quality (%) has nothing to do with the execution mode (Every tick, OHLC, ...). You can obtain 100% quality if you have complete M1 dataset for the testing period, independantly of the timeframe and the execution mode you used.
Documentation on MQL5: Standard Constants, Enumerations and Structures / Environment State / Symbol Properties
Documentation on MQL5: Standard Constants, Enumerations and Structures / Environment State / Symbol Properties
  • www.mql5.com
Standard Constants, Enumerations and Structures / Environment State / Symbol Properties - Documentation on MQL5
Rogerio Figurelli
Moderator
61343
Rogerio Figurelli  
angevoyageur:
History quality (%) has nothing to do with the execution mode (Every tick, OHLC, ...). You can obtain 100% quality if you have complete M1 dataset for the testing period, independantly of the timeframe and the execution mode you used.

Do you mean that we can obtain 100% of Test Quality results without tick by tick execution mode option enabled?

I think we are talking about different metrics: History Quality (as show by MT5) and Test Quality (real backtesting quality).

Note that MT5 History Quality can obtain 100%, but if you don't use tick by tick, the number of ticks will be so low in M15, H1, etc., that the Test Quality, for me, is not 100%.

At least, I never would trust in an EA without a final tick by tick test.

Alain Verleyen
42249
Alain Verleyen  
figurelli:

Do you mean that we can obtain 100% of Test Quality results without tick by tick execution mode option enabled?

I think we are talking about different metrics: History Quality (as show by MT5) and Test Quality (real backtesting quality).

Note that MT5 History Quality can obtain 100%, but if you don't use tick by tick, the number of ticks will be so low in M15, H1, etc., that the Test Quality, for me, is not 100%.

At least, I never would trust in an EA without a final tick by tick test.

What are you talking about ? There is no such thing as "Test Quality" in MT5, only History Quality sometimes called Data Quality. The question of the topic starter was about Data Quality, nothing else.

Obviously a backtest with execution mode "Every tick" will be more accurate than other execution mode.

Documentation on MQL5: Standard Constants, Enumerations and Structures / Environment State / Symbol Properties
Documentation on MQL5: Standard Constants, Enumerations and Structures / Environment State / Symbol Properties
  • www.mql5.com
Standard Constants, Enumerations and Structures / Environment State / Symbol Properties - Documentation on MQL5
Rogerio Figurelli
Moderator
61343
Rogerio Figurelli  
angevoyageur:

What are you talking about ? There is no such thing as "Test Quality" in MT5, only History Quality sometimes called Data Quality. The question of the topic starter was about Data Quality, nothing else.

Obviously a backtest with execution mode "Every tick" will be more accurate than other execution mode.


I will repeat my original phrase before you comment: "Note that using M15, H1, etc., you also can reach 100% test quality, if you enable tick by tick simulation option."

This is just a real market advice to ssn, that I don't know if is so obvious to him.

If so, don't forget that obvious things, sometimes, must also be said here, to help and advice other users.