Episode 3.43 - A Book On Tape

3.43 jumps into some programming notes (0:41), reads through the Sloan 2020 paper submission: Predictive error estimates for baseball statistics using clustering algorithms and covariance analysis (1:28) including the following sections: 1. Introduction (2:04), 2. Methods (3:48), 2.1 Data Discussion (4:39), 2.2 Cluster Computation (8:35), 2.2.1 Choosing the k-means clustering algorithm (9:20), 2.2.2 The Mike Method: A Conceptual Model for determining k (12:01), 2.2.3 The Rentz Routine: A Geometric Model for determining k (13:12), 2.2.4 Clustering Assignment (14:35), 2.3 Player Prediction & Estimation (18:42), 2.4 Prediction Validation (23:33), 3. Case Studies for Christian Yelich (27:43) & Mike Clevinger (31:02), 4. Conclusions (32:04), and concludes with some initial thoughts on the paper (34:16).