We Only Saw the Planes That Came Back
by Anonymous · ⭐ 0 · 🍴 0
✈️ We counted the bullet holes on 100 planes that came back, part by part: 101 on the wings, 119 on the fuselage — but only 7 on the engine and cockpit. Where should the steel plate go? Not where the holes are «many», but where they are «fewest» — planes hit there never came back to be counted. 💥 The second tab brings the lost ones back. Bullets land «evenly» across the five parts (20% each). Set deadliness to zero and the two bars are identical. Raise it and engine hits drop from 19% of all hits to 12% among returners, while the wings climb from 20% to 28%. What vanished was not the holes but the planes. 🔍 Third, spotting it. «The morning routines of 10 successful founders», «old buildings are sturdier», «so many reviews say it cured them» — all three ask the same question: who is missing from this number? This app was made with code too · Remix it and change the deadliness.
About this app
An interactive lesson on survivorship bias using the classic example of counting bullet holes only on aircraft that returned. Explore which aircraft part should be reinforced, adjust the lethality of hits to compare all aircraft with only the returners, and answer three questions about who was excluded from the data.
Use cases
- Students learning statistics and critical thinking
- Teachers introducing survivorship bias
- Adults practicing data literacy
- Anyone who wants to question conclusions drawn from selective samples
Features
- Clickable aircraft-part exercise that reveals why the area with the fewest holes matters
- Lethality slider comparing all dispatched aircraft with returning aircraft
- Dynamic counts and percentages showing how the observed sample changes
- Three-question quiz about missing failed founders, collapsed buildings, and ineffective remedies
- Tabs for returned aircraft, lost aircraft, and identifying missing data
- Reset and retry controls with optional sound toggle