Does it matter which side of the golden band price is on? Yes, a little.
Sitting inside the Golden MA band makes support/resistance and trendline levels perform slightly worse, though volume profile shows no effect at all.
Four independent families are compared (support/resistance, trendlines, grid, volume profile); volume profile shows no effect, and the number for the golden MA family itself carries a construction caveat, since the band is defined by those same MA lines, so it is not blended in as a fifth independent example.
Worked rate
Out of every 100. The line marks 50.
The two Golden Moving Averages – a near one and a far one – form a band between them. We asked: does it matter, for every other kind of level in the system (the grid, support/resistance, trendlines, volume profile), whether price is sitting inside that band or outside it when the level gets tested?
Yes, in the same direction, across the independent families. Being inside the band makes a level's outcome a little worse – the move that's supposed to happen is a bit less likely to actually get there. The clearest case is support/resistance: 51.8% worked inside the band versus 55.6% outside, a gap that holds up even accounting for the size of the samples (4,528 and 30,253 cases). Trendlines show the same direction and also hold up (49.6% vs. 53.0%). The grid shows the same direction too, but its gap is small enough that it doesn't clear the same bar on its own.
Volume profile is the exception – no effect at all. Inside the band, 55.7%; outside, 55.9%. That's noise. Whatever is happening with the band, volume-profile levels do not seem to care about it.
The Golden MA's own number needs a caveat, even though it's the biggest gap in the table (51.2% vs. 58.1%). The band is defined BY the two Golden MA lines, so "inside the band" for this one family mostly just means "price is between its own line and the other line" -- it isn't an independent regime the way it is for the other four families. We checked whether that entanglement was hiding a fake result (splitting by approach side didn't make the gap disappear), so the number is real, but it can't be presented as the strongest example of a cross-family regime effect, and it shouldn't be blended in with the four independent families' numbers.
Reading this plainly: sitting inside the band looks like a choppier, more contested kind of regime for most level types. It's a small effect, a few points, not a strong edge on its own – but it repeats in the same direction across independent families, which is what makes it worth keeping on record.