Fresh analyses of AI, self, and sport
We examine claims by making their measurements, models, evidence, and uncertainty visible — whether the claim is about AI, about you, or about tonight’s game.
What you can explore
Three ongoing programs, one method: define the claim, measure it, compare the evidence, calibrate the conclusion.
The Evidence Desk
A living review of prominent claims about how AI affects thinking, learning, and work — each rated for the strength of its evidence and the attention it draws.
Measures of Your Mix
One psychological measure at a time: take it, understand what it does and doesn’t capture, and build your own survey sheet. Everything scores in your browser.
Forecasts & scorecards
Daily probability screens for the NWSL, Premier League, and NFL — models graded in public against the market, plus the complete World Cup 2026 archive.
On the desk right now
A current sample from each program.
“You can’t tell if AI is making you worse.”
Holds up Strong evidence, modest attention — the self-assessment literature backs it.
№4 · Life satisfaction
The Satisfaction with Life Scale: five items, two minutes, scored entirely on your device — with the norms and caveats that make the score meaningful.
NWSL championship race
Match forecasts, the table, and a public scorecard of how our model is doing against the market — updated daily through the season.
Who writes this
Data for Thought is written by a cognitive scientist (psychology PhD) working in applied statistics and forecasting. The through-line of every piece is measurement: what a number actually captures, what it misses, and how confident the evidence lets us be. Forecasts are graded in public; measures ship with their limitations; claims are rated against the primary research.