Diego Sarceño·Jun 23Probability Is Not a DecisionWhy a World Cup predictor ships four different answers from one forecast — and what pari-mutuel game theory has to teach data scientists
Diego Sarceño·Jun 20200×–1500×: Caching and Vectorizing a Monte-Carlo Tournament SimulatorHow a memoization table and a NumPy rewrite turned an intractable simulation into a coffee-break job
Diego Sarceño·Jun 19One Source of Truth, Eight Layers, Zero SurprisesEngineering a reproducible ML system — and the config that silently did nothing for months
Diego Sarceño·Jun 18The Champion List Is a LiarBuilding a leakage-free cross-tournament backtest as the sole arbiter of every model change
Diego Sarceño·Jun 16Why a Skellam Distribution Beats a Score GridThe statistical backbone of a World Cup forecaster — and the calibration bug that hid inside a 5×5 matrix
Diego Sarceño·Aug 6, 2025🧠 ¿Por qué amamos los números… incluso si decimos que los odiamos?Ciencia, datos y deportes: la mezcla inesperada que conecta más de lo que crees