Hi everybody,

don’t want to speak for GG, but I have it confirmed by him that real-time bars are sliced by ATR(5) and then are compared against UNsliced history bars.  Attached is a simple indicator that shows how many times each bar shoud be counted for.

Hope it helps.

Vlan

I don’t think we should use ATR to split the candles because the ATR I think will cause mismatches with the history. The candles should be I think be split based on the average bar size of the entire dataset instead. But we can try each one. It shouldn’t be too difficult to test each case 

How should we go about ” “slice” big-bars and make sure not to include fractals among the slices”.

If the bar is lets say 10pips in height (up bar), how should we split it up?

Thanks!

Hi Pedma,

Regarding your point (a) ,  I think GG does not change anything to DB. The gaps handling and splitting the bars is done in his algo …

Saver,

1. No, DB is as it was before. I was lazy to do it during testing differences.. and I already had the Close price in DB. So modifications to the algo seems simpler. in retrospect: a mistake…

2. Same as #1 –

3. Just “slice” big-bars and make sure not to include fractals among the slices.

G.

Ohhh, here we go again…

Stinky, there is no EA per your “requirements”, and there will be no “proof” for claims that wasn’t made.

There is an indicator that does what it shows.

Most of the charts posted are at end of session, so you can match last prediction with whatever came next. Very simple task for you, I’m sure.

If you don’t have…[Read more]

Looks great GG!

Thanks for posting these screenshots.

I have a few questions.

Are your fractals DB now based on closed price fractals?
Did you clean out the data to remove any gaps (weekend gaps, general price gaps)?
Is there anything else that we need to do to prepare the data?

Thank you! 

Nice work! All trades were in profit.

Recent charts after this week close.
2 EURUSD – M5 & M15
2 GBPUSD – M5 & M15
1 USDJPY – H4 – Longer terms version

G.