Does support or resistance hold the same way on every timeframe?
Hold rate stays flat around 70-72% across all eight rungs, but the same flat pattern shows up with non-phi constants too, so it is not proof of anything special – the size of the reaction once a level breaks does grow more than threefold from the fastest rung to the slowest.
Hold rate is pooled across support and resistance, all eight rungs; the flat hold-rate reading also reproduces with non-phi constants (sqrt(2), pi, e), so it is read as not contradicted rather than proven – the reaction-size scaling does not carry that caveat.
Hold rate
Reaction size
Out of every 100. The line marks 50.
The ladder has eight rungs, from a 21-minute clock up to a 610-minute clock. We checked whether a level "holds" (price bounces off it instead of breaking through) at the same rate no matter which rung you're looking at.
Yes, on the surface. Every single rung sits in a tight band -- 70.1% to 72.1% – with sample sizes from about 1,750 up to about 3,600 per rung. That is a two-point spread across eight very different timeframes, which is about as flat as a number gets.
But there's a caveat, and it matters. A separate check (see the linked alt-constants entry) built the exact same kind of grid out of made-up numbers that have nothing to do with this system's ratio – and those fake grids print the same flat 70-ish-percent hold rate too. That means the flat hold rate is mostly telling us about how the touch rule itself is built, not about anything special to this system's structure. So this half of the finding should be read as "not contradicted" rather than "proven."
The size of the reaction, once a level does break, is a different story and does not have that caveat attached. It climbs steadily as you move up the ladder – from about 2.6 times the average bar range at the fastest rung to about 9.1 times at the seventh rung, more than a threefold increase – before dipping slightly at the slowest rung. That climb makes sense once you remember the reaction is measured in each rung's own units: a bigger structure produces a bigger, slower move, measured on its own ruler.
Stated plainly: the shape of the reaction repeats across rungs (how often it holds), but the size of the reaction does not stay the same number – it scales up with the rung, by design. That is the correct, expected picture of self-similarity, not a contradiction of it.