Locard's Exchange Principle (1910): "every contact leaves a trace" — when two surfaces touch, material is transferred both ways. Real trace evidence (fibers, dust, glass, gunshot residue) transfers probabilistically, not universally, and the transferred trace then decays with time and handling.
Transfer probability (per particle):
P = 1 − exp(−k · pressure · roughness), k = 2.2
Persistence over time:
S(t) = exp(−λ · t) (population survival curve)
each particle draws retain ~ Uniform(0,1) at the moment
it transfers, and stays detectable while S(t) ≥ retain.
Because retain is uniform, P(a given particle still visible
at time t) = S(t) exactly — the population fraction remaining
always matches the exponential curve (verified numerically).
- Contact pressure — how hard the two surfaces are pressed together; higher pressure drives more particles across the interface.
- Recipient surface roughness — a textured/fibrous recipient (e.g. wool) catches and holds more trace than a smooth one (e.g. glass).
- Donor trace density — how many trace particles start on the donor plate; a real crime-scene surface may carry far more or far fewer than a lab swatch.
- Make Contact — presses the recipient plate onto the donor plate; at the moment of contact each donor particle transfers independently with probability P.
- Days since contact / activity level / Play decay — every transferred particle is assigned a random retention threshold at the moment it transfers; as simulated time and activity (washing, abrasion, brushing) increase, the survival curve S(t) drops below more thresholds and those particles are marked lost — this is why trace evidence has to be collected quickly.
The contact-area panel is draggable and scrollable (mouse wheel / pinch) so you can pan and zoom into dense clusters of particles; the two side panels track the same population in aggregate — a live survival curve and a histogram of retention thresholds — so you can see the individual particles and the statistics agree.
Real-world relevance: this probabilistic transfer-and-persistence model is the same reasoning crime labs use to judge whether an absence of trace evidence means "no contact occurred" or simply "the trace didn't survive" — a distinction at the heart of countless real forensic cases.