🎲

Is Success Luck or Hard Work?

Replicating Veritasium's NASA astronaut selection experiment: when you select the very best from a huge pool, winners end up surprisingly lucky — even when luck barely counts.

What is this?

Imagine 18,000 people applying to become NASA astronauts. Each person has a skill score (their ability, 0–100) and a luck score (random circumstances, 0–100). Their final application score is mostly skill, with just a small slice of luck. The top 11 are selected as winners.

The surprising result: even when luck counts for only 5% of the score, the average winner is luckier than 90% of the population. This simulation runs that selection 500 times to prove it's not a fluke — it's the unavoidable math of extreme competition. At the very top, tiny differences in luck break ties between people with near-identical skill.

How to read the charts
  • Scatter (Skill vs Luck): Each dot is one applicant. Yellow dots are winners. Notice they cluster in the top-right corner — winners need both exceptional skill and above-average luck.
  • Avg Winner Luck per Iteration: Each data point is one run of the full 18,000-person selection. Watch the line stabilise as more iterations are averaged together.
  • Luck Distribution histogram: Blue bars show where the whole population's luck sits (centred near 50). Amber bars show where winners' luck sits. The rightward shift is the key result.
  • Skill Distribution histogram: Same layout. Winners cluster at the extreme high end — near-perfect skill is the baseline to even be in contention.
  • Compare Luck Weights chart: Runs six simulations at different luck weightings (0% → 50%). Even a tiny 1% luck weight dramatically raises winner luck scores.

Scatter — Skill vs Luck (showcase iteration)

Avg Winner Luck per Iteration

Ready

Status Log