Forecasting, scored
Find out how good your judgement actually is.
Forecast real questions alongside people you know. Every prediction is scored against what actually happened, so the feedback is the point — not the winning.
Points are free, and they only ever stay points. Nothing here converts to anything outside the app.
Every resolved forecast is scored the same way, and the score is public on your profile.
Nothing to pay, nothing to top up, and no way to turn points into anything outside the app.
A private circle keeps its own questions and its own standings. Nothing crosses over.
Change your mind as often as you like. The last forecast before close is the one scored.
Questions your circle actually argues about
Public questions sit alongside ones only your circle can see. Scores never cross between them — a circle’s standings are its own. The three below are examples.
Closes 18 March
Will the Fed cut rates at the March meeting?
The crowd started this question at 31% and drifted to 62% over six weeks. Every step of that is on the record, including yours.
Did the club finish Middlemarch before 1 March?
Will our Q3 launch date hold?
Three steps, then the part that matters
Say what you think
A probability, not a side. “62%” is a different claim from “yes”.
Write down why
Private to you. It is what makes a later review honest instead of reconstructed.
Change your mind
Revisions are kept. The last one before close is the one that counts.
Then it resolves
You get a Brier score, your calibration curve, and where you sat against the crowd. That is the product.
Start with one question.
Forecast something you already have an opinion about, and find out in a few weeks whether the opinion was worth having.