We ran a Datathon. Here's what 200 students taught us

Two hundred students, forty-eight hours, one very tired organising team. Here's what running the club's flagship Datathon taught us that no retrospective slide ever could.

Logistics is the product

Everyone remembers the winning project. Nobody remembers the wifi that didn't drop, the coffee that never ran out, or the schedule that actually held. But those invisible things are exactly what let the good ideas happen. Spend your energy there first — the magic is downstream of the boring work.

An event is a project with a deadline you can't move — so plan like the deadline is real, because it is.

What beginners build under pressure

The surprising pattern wasn't in the technology. Beginner teams consistently over-scoped in the first hour, and then — somewhere around hour thirty — made the single best decision of the weekend: they cut.

The teams that shipped weren't the most skilled. They were the ones that chose one small, honest thing and actually finished it. Watching that happen two hundred times in a weekend is the best data-science lesson we know how to teach.

We'll run it again next year. Bigger, calmer, and with twice as much coffee.