How do narrow resonance widths quantitatively suppress random combinatorial pairing versus true cascade events?

Your question touches on a subtle but important distinction in cascade physics: how narrow resonance widths act as a *filter* that distinguishes genuine multi-step cascade events from spurious random pairings that look superficially similar.

The most direct answer comes from resonance compositeness. When a resonance has a narrow decay width, it signals that the intermediate state is genuinely bound or coherent rather than a random collection of decay products. [AG-2024.03-1181] introduces a probabilistic compositeness scheme that explicitly excludes resonances with "unphysically large decay widths," showing that near-threshold resonances with small widths have correspondingly *small* composite fraction—meaning they are less likely to arise from accidental combinatorial overlap of unrelated particles.

The physical intuition: a narrow width means the resonance persists long enough to be a real intermediate object. A broad width—or no resonance at all—suggests random kinematic coincidences. By quantifying compositeness as a function of decay width, you can suppress false positives where two uncorrelated decay products happen to land in the same invariant-mass window.

In cascade dynamics more broadly, [AG-2026.02-995] shows that true cascade events—where "fragile bound or coherent structures are formed through intermediate states"—obey non-Markovian memory effects over a characteristic *memory time*. Random pairings, by contrast, lose memory of their history instantly. So a narrow resonance width (long lifetime) naturally preserves the memory signature that distinguishes genuine cascades from noise.

The provided preprints do not quantitatively compare suppression factors (e.g., ratio of cascade-event yield to combinatorial background as a function of width) in the context you're asking.