Probability weighting

The Prelec function

Move α to see how probabilities translate into decision weights.

One parameter. A changing curve.

w(p) = e^{−(−ln p)^α}

Shape parameter α 0.60

0.10 · More curvature1.00 · Linear

Inverse-S weighting

Look for the fixed point.
Every curve passes through p = w(p) = 1/e ≈ 0.368. The endpoints stay at (0, 0) and (1, 1).

A decision weight measures a probability’s influence on choice. It does not change the actual chance of an outcome.

━━ Prelec w(p)┄┄ Linear benchmark w(p) = p

Actual probability p	Weight w(p)	Difference w(p) − p

One-parameter Prelec model · 0 < α ≤ 1 · w(0) = 0; w(1) = 1.
Based on Prelec (1998), The Probability Weighting Function.