tick mode

• 07/06/2017 at 2:04 PM #39955
  Seemore Profit

   

  When i backtest never really get a high tick mode. This one test not really doing anything major tricky or different but i get tick mode 68+ any ideas why this could be?

   

  Thanks

  Jamie

  

  Seemore Profit

  Participant

  Topics: 5

  Replies: 21

  Been thanked: 7 times
  07/07/2017 at 8:26 AM #39985
  Nicolas

  I don’t understand what you mean by “get a high tick mode”? Please explain.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4775 times
  07/08/2017 at 6:38 PM #40104
  michel

  I also have a “tick mode” field in the vars optimization report (see last column of the attachment) but have no idea what the number shown means. Help will be appreciated 🙂

  PS: my PRT config is in spanish.

  

  michel

  Participant

  Topics: 0

  Replies: 7

  Been thanked: 2 times
  07/08/2017 at 7:16 PM #40113
  Nicolas

  @michel

  @GraHal and I have answered this question in another thread this week, you’ll find it here: https://www.prorealcode.com/topic/tick-mode-help/

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4775 times
  07/08/2017 at 7:19 PM #40114
  michel

  Thanks Nicolas I’m aware of this thread but I’m afraid I don’t really understand it. In a few words the less the better?

  

  michel

  Participant

  Topics: 0

  Replies: 7

  Been thanked: 2 times
  07/08/2017 at 7:31 PM #40117
  Nicolas

  The optimisation engine don’t avoid the old “0 bar” problem (to be faster), while the simple backtest in tick mode is the most reliable test you can have (without the “0 bar” issue).  You’ll find many discussions about the “0 bar” or “zero bar” phenomena in the forums (use the search form).

  In a few words, the internal bar price movement is not checked in optimisation tool.

  1 user thanked author for this post.

  michel

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4775 times
  07/08/2017 at 7:48 PM #40119
  michel

  Merci beaucoup Nicolas, j’essayerai de comprendre à fond ce que tu viens de me dire.

  michel

  1 user thanked author for this post.

  Nicolas

  

  michel

  Participant

  Topics: 0

  Replies: 7

  Been thanked: 2 times
  07/09/2017 at 11:22 AM #40163
  GraHal

  @michel the result of changes in price within / inside the bars / candles are ignored by the optimisation calculator.

  @Seemore Profit price must be touching your stop loss 68+ times with the variable values in that ’68+ backtest’ row?

  2 users thanked author for this post.

  Seemore Profit, michel

  

  GraHal

  Participant

  Topics: 214

  Replies: 8885

  Been thanked: 2225 times
  07/09/2017 at 4:38 PM #40178
  michel

  @Nicolas, @GraHal: PRT 10.3 was supposed to solve some issues like the zero bar problem (http://bit.ly/2u0nueX), so “ignored by the optimization calculator” means that I still can’t trust results consistently? Is it worth doing this optimization or is it a false friend to avoid? I’m doing a large job based on this premise in order to test deeply my last script and the tick by tick backtest results are too good pro tempora. I distrust and that’s why I want to clearly understand the tick mode in practice.

  

  michel

  Participant

  Topics: 0

  Replies: 7

  Been thanked: 2 times
  07/09/2017 at 9:38 PM #40182
  GraHal

  “ignored by the optimization calculator” because to calculate an optimised variable would take forever with the current optimisation engine.

  However, when you select and run a set of optimised variables then tick mode kicks in (is NOT then ignored) and the results you see in the Equity Curve and in the ‘Detail Report’ are what you might expect to get running in real time / Live.

  You say …  tick by tick backtest results are too good … please confirm you have enabled (ticked the box) for running in tick tick mode?

  

  GraHal

  Participant

  Topics: 214

  Replies: 8885

  Been thanked: 2225 times
• Author
  Posts