Free · graduate level · no sign-in
Six fictional organisations, seven modules, one way of learning: read a short case, see the theory with its equations, run a lab where the mechanism moves in front of you, then earn XP by answering challenges on real numbers from a fixed data set. Every answer can be checked; every module ends with a board meeting.
Written for M.Sc., MS and MBA students of statistics, data science, economics and business, and for anyone who learns better by doing. Modules build on each other, so start at 0 unless you already know the earlier material. Progress is saved in your browser only.
A grocery-delivery start-up, a payments company, an EV-charging network, a village council, a ride-hailing firm, a microfinance lender. Each case file is a problem someone there actually has, and the statistics is what solves it.
Three to five paragraphs with the equations, textbook terms first, then a line of further reading to standard texts (Tukey, Tufte, Blitzstein & Hwang, Ross, Wasserman, Casella & Berger, Rice, James et al., Hastie et al., Wooldridge) and the original papers.
Orders onto a live histogram, transactions through a rule engine, cars through a charging hub, surveys through a sampling machine, trips through a regression line, loans down a decision tree, customers pulled toward cluster centres. Most labs simulate unit by unit; all have Play, speed and Reset.
First-try answers earn 100 XP, hints cost 25, each cleared file earns a badge, and five cleared files unlock the board meeting. Ranks rise with XP. Nothing is graded by anyone but you.
Each module uses one fictional data set that never changes, so the numbers in the challenges are reproducible by hand, in R or in Python.
Use it as pre-reading, as a lab hour, or as revision. Modules 0–4 follow a first graduate statistics course; 5 and 6 add the core of a first machine-learning course. Each goes a little beyond the usual syllabus.