PREDICTING
“romance: questions. products: want the bad version. people: don't change.”
The Lesson
Prediction tips by domain: (1) Romantic interest. Do they ask about your life? Do they find excuses to touch you? (2) Product success. If people want the bad version (first iPhone, first car), the good version will be huge. (3) Politics. Follow the money (direct or indirect). (4) News. Assume danger is exaggerated; news profits from fear. (5) Complex models. Don't trust long-term predictions from complicated models; too many subjective assumptions. (6) People. Assume they don't change. (7) Everyone is selfish (even altruists get meaning from helping). (8) Slow disasters get solved (ozone, Y2K). (9) 'Rogue experts' are wrong 95% of the time.
Real-World Example
A founder evaluates a new market. They apply filters: Do early users desperately want the buggy prototype? (Good sign.) Is competition following money or ideology? Are industry predictions based on complex models? (Discount them.) Has the key competitor's CEO 'changed'? (Probably not.) Layer multiple prediction filters for better odds.
Watch Scott explain this lesson
Watch Lesson on Locals →