Predicting with a model: commute time

Subject: the time from stepping out your front door to sitting down at your office desk. This time, don't stop at building the model — use it to make a 'prediction that could be wrong' and design the verification.

(1) Model commute time simply with 2–3 inputs → relationship → output. (2) Derive one verifiable prediction from that model. It must include conditions and numbers (e.g., 'leaving after 8:10 takes 12+ minutes longer than leaving before 8'). (3) Write how you'll verify it this week — what, when, and how you'll record, and what you'll compare it against. (4) Write which part of the model you'll suspect if the prediction is wrong — distinguishing what kind of miss means a variable is missing versus what kind means the relationship is wrong.