Author Topic: Time machine inconsistent across many tests with different data  (Read 1070 times)

Tags:
  • All members
  • Posts: 2
Hi,

I've been trying Time Machine for some weeks now running the same triggers and setup on the same historical markets many times. Some free data and some market data I have paid for.

I am not impressed so far. Every time I run it, it produces different results, whether in high speed mode or standard speed. It is not possible to test any sort of plan with it when it produces totally inconsistent results.

Am I am missing something here?

Regards,

Martin.


  • Tim Vetrov
  • Administrator
  • Posts: 4870
  • Gender: Male
*
Hello Martin,

It's not a bug but a feature ;)
TimeMachine simulates real life situation - unstable connection, missed refreshes, out of sync trigger execution.
Depending on trigger and trigger variables values the execution could bring to different results.
If you need more details please share your trigger to [email protected] 
Thank you.
Proud to be 🇺🇦
I'm happy to help Monday - Friday, 08:00-18:00 GMT
Буду рад помочь с понедельника по пятницу, 08:00-18:00 GMT

  • All members
  • Posts: 2
Hello Martin,

It's not a bug but a feature ;)
TimeMachine simulates real life situation - unstable connection, missed refreshes, out of sync trigger execution.
Depending on trigger and trigger variables values the execution could bring to different results.
If you need more details please share your trigger to [email protected]
Thank you.
Thank you for your reply Tim. Reproducibility of results is a fundamental requirement of Science, Maths, and Statistics. What you have just said is that it is not possible to reliably test any edge, any system in Time Machine

  • Tim Vetrov
  • Administrator
  • Posts: 4870
  • Gender: Male
*
Not exactly.
I mean the results could vary in a reasonable limits of deviations.
Proud to be 🇺🇦
I'm happy to help Monday - Friday, 08:00-18:00 GMT
Буду рад помочь с понедельника по пятницу, 08:00-18:00 GMT

  • All members
  • Posts: 905
Unfortunately, betting can never be an exact science. TM is a very good tool to test triggers and evaluate strategies.  If betting in-play the results will probably show more variation than when the markets are idle.

 

Please note, BetFair is seems to be currently OFFLINE