Bug info:
Hi, while backtesting my existing strategy i noted a drastic change in backtest results. I cannot understand why this can happen with same parameters since last one week. Please clarify if its because of any update/ change in the historical data / software?
We have updated the logic of “Maximum deals per higher timeframe”.
If you have any settings related to it, please try to turn on/off the option.
Yes, that helped.
once again i see a change in backtest results. Any changes updated in the historical data?
There were some corrections regarding the calculation of some values. Maybe that affected all results and shows more accurate values now?
There were changes due to Backtest issues - #2 by aressanch
In some circumstances backtest might give wrong entry points on tf lower than lower indicator
thats not helping then. My strategy was based on backtest data and now all of a sudden the values have somehow changed. i have been frequently checking back since i applied my strategies now suddenly there is a major change. Not sure this will give confidence to all of us as community. Needs to be seen at priority and the exact change should be communicated.
Can you share backtest so I can check.
Links of affected bots are appended below:-
It depends on Backtest issues
Backtest timeframe is less than indicator. In some circumstances it gave wrong entry points.
I think to make it like before you need to add 15m indicator that is always true. Example, RSI < 100
ok let me try please
Yes, you require an indicator that is always true and triggers every 15 minutes. That’s one way to do it.
what do u mean by Always True?
though i still believe the issue is not exactly that. As my initial strategy included 15 min TF but results are no more the same
RSI is always bigger than 0 and smaller than 100.
If you backtested with 1 minute timeframe maybe you have to increase the timeframe of the test to match the one of your lowest indicators? Nevertheless, I think that backtests should be able to work with every timeframe lower than that of the lowest indicator, if it can adapt to the extra data.
the first backtest results have been restored (due to adding 15 min TF) but the 2nd one (as mentioned below) is still showing wrong info.
Gainium app
finally restored both backtests. Thanks for prompt responses.
