How Not to Be Wrong: The Power of Mathematical Thinking - with Jordan Ellenberg

Mathematical Thinking and Problem Framing (Abraham Wald Story)
📌 Abraham Wald, during WWII, was asked by generals to reinforce bombers based on where bullet holes were concentrated on returning planes.
🧠 Wald recognized the flawed assumption: planes hit in critical areas (like engines) did not return; thus, armor needed to be applied where no bullet holes were observed.
💡 Mathematical thinking involves not just computing formulas, but interrogating the underlying assumptions of a problem, sometimes overturning the question entirely.

Expected Value and Lottery Strategy (Cash Windfall)
💲 The general mathematical case against lotteries is that the expected value (average value) of a ticket is less than its cost (e.g., $1.50 expected value for a $2 ticket).
📉 The Massachusetts Cash Windfall lottery’s expected value was only 80 cents per $2 ticket, but a "roll down" rule made expected value jump to $5.53 on specific days.
💰 MIT students (Random Strategies) exploited this positive expected value, buying up to 80-90% of tickets on roll-down days, effectively acting as the house rather than beating it.
🏛️ The state continued allowing the practice for six years because the large volume of ticket sales generated an estimated $10–$15 million in extra revenue, as the state profited regardless of who won.

Risk Mitigation via Combinatorial Design (The Fano Plane)
❓ A key mathematical puzzle was why the MIT group hand-filled tickets, unlike others using quick-pick machines—suggesting a strategy to minimize risk despite the same expected return.
🎲 By simplifying to a small "Transylvanian lottery" (3 numbers out of 7), the speaker demonstrated a set of 7 tickets based on projective geometry (the Fano Plane).
🔒 This geometric configuration guaranteed a fixed return of $6.00 (the expected value) across all outcomes, eliminating the risk of winning less than the average (hedging).
🎯 For the actual 46-ball lottery, a similar combinatorial design using roughly 230,000 tickets could theoretically guarantee specific high-tier payouts, essentially hedging away all risk of loss.

Key Points & Insights
➡️ Question Assumptions: True mathematical thinking requires challenging the premise of a problem, exemplified by Wald realizing where the planes *weren't* being hit.
➡️ Positive Expected Value is Key: Exploit situations where the average return outweighs the cost, but understand that doing so in high volume makes you the *system*, not the gambler.
➡️ Risk vs. Return: Even with the same expected value, investors (or high-volume lottery players) prefer lower-risk, hedged bets, especially when using other people's money (leverage).
➡️ Geometry and Design: Complex, risk-free strategies in combinatorial problems (like lotteries) can be rooted in advanced mathematical structures like projective geometry and combinatorial designs.

📸 Video summarized with SummaryTube.com on Dec 01, 2025, 15:52 UTC