The worked example handed you the steps. Here you don't get them. Each case drops you into a real sports question and asks you to decide: is this even a regression problem, what's X, what's Y, and what do you do first? The arithmetic stays light — the thinking is the point.
Start with the Dodgers arrow_forward
Case Files drop you into a fresh scenario and make you drive all four phases — Frame, Plan, Crunch, Decide. Coach the Rookie hands you a flawed analysis to spot and fix — including the cases where SLR is the wrong tool entirely.
Frame it, plan it, crunch it, decide.
Do they score more in hotter weather? Runs vs temperature.
Start → LiveBasketball · Case FileDo teams that shoot a better 3-point % win more? Wins vs 3PT%.
Start → LiveNFL · Case FileDo offenses with more passing yards score more? Points vs yards.
Start →Spot the flaw in someone else's analysis, then fix it — including when SLR is the wrong tool.
They divided by n. Find the mistake and fix the variance.
Start → LiveCoach the RookieA strong fit became a bold claim. Spot the leap from correlation to cause.
Start → LiveCoach the RookieA great fit, a wild prediction. Spot the extrapolation.
Start → LiveBoundary caseA trap: when is SLR the wrong tool entirely?
Start → LiveBoundary caseA near-zero R² on curved data. Why won't a straight line fit?
Start →Spaced repetition — a "due today" queue that resurfaces these on a schedule — is coming next.