The Decoded 7-Day Think-in-Models Challenge

A seven-day habit of noticing the models, assumptions, and uncertainty hidden inside everyday claims.

Type: Challenge
Difficulty: Beginner
Length: 7-day challenge

Professor Juneja spends his days turning messy, uncertain situations into models, and then being honest about what those models can and cannot say. This challenge borrows that habit and shrinks it to something you can practise in a few minutes a day.

Each day is one small act of noticing. There is nothing to install or sign up for, and your progress is saved only in your own browser. The goal is not to become a mathematician in a week. It is to start seeing the assumptions and uncertainty that sit quietly behind the numbers you meet every day.

Loading your progress…

DAY 1Spot a model in the wild
Find one prediction you saw today, a weather forecast, a delivery time, a sports odds line, and name what it is really a model of. What does it keep, and what does it ignore?
DAY 2Find the hidden assumption
Take a confident claim about the future and write down one assumption it depends on. If that assumption were wrong, how far off could the claim be?
DAY 3Turn a certainty into a probability
Catch yourself, or someone else, saying something 'will' happen. Rewrite it honestly as a rough probability. Notice how it changes the way you'd act on it.
DAY 4Draw the error bars
Find a single headline number today and ask what range it probably hides. Would the story change if the true value were 20 percent higher or lower?
DAY 5Hunt for exponential thinking
Find something that grows by multiplying rather than adding, like interest, followers, or an outbreak. Estimate where it lands after ten steps. Did your gut undershoot?
DAY 6Separate the tool from the claim
Find a headline about AI or technology. Rewrite it to say precisely what was done, stripping out any word that claims more than the evidence shows.
DAY 7Make one decision like a modeler
Take a real choice you face. List the options, the key uncertainty, and what you would need to know to decide well. Then decide, knowing the model is a guide, not a verdict.

Your progress is saved only in this browser. Clearing your browser data will reset it.

All Episode 4 resources Listen to the episode →