Time bomb back testing
-
I know back testing the time bomb block doesn't produce accurate results as it's not exposed to the real time tick flow but I'm wondering how it actually works during a back test.
Is the time bomb block ignored and skipped during testing? Automatically passed?
Does the time bomb block pass continuously because ticks are being thrown at it almost instantly?
Does the block produce random signals...? if so, based off what? -
It is not passed, but the seconds you have in the Tester are not accurate. The lowest timeframe is 1 Minute and all ticks inside that minute are generated. Their time is simply fake
-
I'm slightly confused, you say the block is not passed but I've done some testing and it does produce results although I know they are fake. So I'm wondering what actually causes it to pass on some occasions but not constantly given data is being thrown at it really fast. Does the backtest data attempt to create time values out of thin air causing the entry on some occasions but not others.
-
Oh, I mean passed, that it works... but with generated data, so don't believe the results. And I think the block is very simple, it collects data for the certain seconds, then it dumps that data and starts collecting new data for the new period. This is something I should fix I guess, but in the same time from a long time ago I was thinking to hide those blocks in this category
-
Hello.. you say "The lowest timeframe is 1 Minute and all ticks inside that minute are generated", but in MT5 we can in fact use real ticks from the broker in strategy tester, so I assume we can actually trust time bomb blocks as long as real ticks are used. What do you think?
As far as I noticed, the real ticks have millisecond resolution, matching the exact time they occurred in the market, and also account for spread changes. See image below:
