First thing is day of the week and which profiles and templates may exist for that particular day

Only focus on setups within the true day

Again asian range standard deviations just like CBDR

You need to decide where youre going to go after the 4h, make it a personalised approach, which timeframe are you going to after the 4h, 1h is nice but it still has to be refined a bit more. ICT likes to go from a 4h into a 30m or 15m and then further refine it to the 5m if possible, usually ICT looks for his setups and executes on the 15m minimum, the best is to get down to the 5m

The key is that we look for a timeframe below the 4h that produces FVGs, the 1h may be muddy and the 15m may have clear FVGs. So we can select our PD arrays on the 15m as well.