Each chapter has a short browser game that drills its working ideas. A game runs on a phone or a laptop, needs no login, and works offline once the page has loaded. Every round asks you to do something (fit a line, tap the outlier, order the steps, match the terms, find the wrong sentence) and then explains the answer with a pointer to the exact book section it came from.
Recursive review. Every game after the first plays ten rounds: seven on the new chapter, plus three review rounds sampled from the pools of earlier chapters, with one always drawn from the chapter just before. Review rounds carry a “Review: Chapter N” badge and favor the older ideas the new chapter leans on most, so the deck grows with you and keeps what you already learned within reach while it adds what is new.
The 16 games¶
Chapter 1: Regression as a way of thinking. Description, prediction, and causation, with Galton, Anscombe, and the Challenger warning.
Chapter 2: Simple linear regression. The least-squares line, residuals, and the five-stage modeling workflow.
Chapter 3: Inference for simple linear regression. Tests and intervals for one slope, the ANOVA table, and R-squared.
Chapter 4: Correlation. What r measures, Fisher’s transform, and the traps a single number can hide.
Chapter 5: Randomization and bootstrap inference. Permutation tests and resampling for when the formulas run out.
Chapter 6: Matrix algebra for regression. The matrix objects (X, X’X, projections) every later chapter is built on.
Chapter 7: The general linear model. Least squares, the hat matrix, and the geometry of the fit in matrix form.
Chapter 8: Multiple regression in practice. Partial slopes, extra sums of squares, and the general linear test.
Chapter 9: Model diagnostics. Hat values, studentized residuals, and Cook’s distance for spotting influence.
Chapter 10: Remedial measures and transformations. Logs, Box-Cox, weighted least squares, and where each one stops working.
Chapter 11: Categorical predictors and interactions. Dummy coding, interaction terms, and polynomials.
Chapter 12: Multicollinearity, selection, and validation. VIF, selection criteria, cross-validation, and shrinkage.
Chapter 13: Logistic regression. The logistic model, odds ratios, and fitting by maximum likelihood.
Chapter 14: Poisson and the GLM. Count models and the one family that ties linear, logistic, and Poisson together.
Chapter 15: Regression with time. Autocorrelation, the Durbin-Watson test, and honest forecasting.
Chapter 16: Path analysis. Direct and indirect effects, and a look back across the whole workflow.