Dilution effect

Biological Parallel

Predicting rabbit population growth from current abundance, weather data, and predator counts—high accuracy. Adding irrelevant variables (lunar phase, wind direction, atmospheric pressure) dilutes the model: noise drowns signal. The dilution effect: more data isn't always better. Immune systems face this: T-cells exposed to too many similar antigens simultaneously show reduced response to the primary threat—diluted activation across targets. Predator recognition dilutes with varied alarm calls: vervet monkeys with precise leopard/eagle/snake alarms respond faster than species using generic danger calls. Biological forecasting suffers when adding marginally relevant variables (distant climate indices, tangentially related species) that worsen predictions. The mechanism: finite resources (attention, activation, data) spread across more targets increase noise-to-signal ratio. Parsimony isn't elegance; it's predictive necessity. Companies dilute brand identity when expanding product lines too broadly—the same signal degradation.