AI summary by TwinMind

Can you solve the basketball riddle? - Dan Katz

TED-Ed

6m 13s

Problem Setup

  • Engineer created basketball-playing robot "Dunk-O-Matic"
  • CEO falsely advertised robot can "automatically adjust skill to create fair game"
  • Original design: robot shoots baskets with variable success, takes turns with human
  • No adaptive performance capability initially built
  • Robot has adjustable probability feature: can set probability q for success on each attempt
  • Team has dossier with probability p for each human participant's basket-making success
  • Goal: calibrate robot so human has exactly 50% chance of winning each match

Game Rules & Structure

  • Human shoots first, then robot alternates
  • First successful basket wins the match
  • Robot probability q can be adjusted between opponents
  • Need to account for first-player advantage

Mathematical Analysis Using Geometric Series

  • Simple approach (q = p) fails due to first-player advantage
  • Example: if both players have 100% success rate, first player always wins
  • Solution requires geometric series analysis

Geometric Series Fundamentals

  • Infinite sum where each term = previous term × common ratio r
  • If |r| < 1, series has finite sum
  • Sum formula: a/(1-r) where a = first term

Human Win Probability Calculation

  • Round 1: probability p of immediate win
  • Round 2: requires both players miss first attempt
    • Miss probabilities: (1-p) for human, (1-q) for robot
    • Probability: p × (1-p) × (1-q)
  • Round 3: requires two rounds of double misses
    • Probability: p × [(1-p) × (1-q)]²
  • Total human win probability forms geometric series:
    • First term: p
    • Common ratio: (1-p) × (1-q)
    • Sum: p / [1 - (1-p)(1-q)]

Solution Derivation

  • Set human win probability = 1/2
  • Solve: p / [1 - (1-p)(1-q)] = 1/2
  • Algebraic solution: q = p / (1-p)
  • Constraint: if p > 50%, q would exceed 1 (impossible)
  • Fair game impossible when human's immediate win chance > 50%

Alternative Reasoning Approach

  • Robot's first-round win probability: (1-p) × q
  • For fairness: robot's first-round chance should equal human's
  • Set (1-p) × q = p
  • Solve: q = p / (1-p)
  • Same result, simpler derivation
  • Logic: if first-round odds equal, subsequent rounds effectively restart with same odds

Demonstration Outcome

  • Solution implemented successfully
  • Engineer chose honesty over deception
  • Publicly explained company's false promises and improvised solution
  • Bad press directed at employers, not engineer
  • Presentation volunteers owned competing robotics company
  • After intellectual property litigation, engineer joined healthier workplace
  • Gained regular spot on pickup basketball team

TwinMind

Get summaries like this for your own meetings

TwinMind takes notes in the background during meetings, lectures, and calls, then writes the summary for you.

Free plan to get started

No bot joins your calls

Phone, desktop & Chrome apps

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome