Philip Thompson (1922–94) pioneered innovative approaches to weather analysis and prediction that blended determinism and probability. He generally posed problems in terms of simplified dynamics that were amenable to analytic solution. His preciseness in problem formulation and presentation in a forceful didactical manner are linked to his early home-schooling and experiences with a coterie of young intellectuals. Four of Thompson's contributions are examined with the intention of highlighting their impact on the current state of operational analysis and prediction.