← read the signalhow to play

Guess the R²

Read the scatterplot. Make your guess. Reveal the fit.

round 01/5score 0
changing settings starts a new game
sample 00180 observations

Generating your first scatterplot…

Find the fit.

How much of the sample’s y-variation does a straight-line fit explain?

0.50Your R² estimate
0 · No linear fit1 · Perfect fit

Upward and downward patterns can both have a high R².

Use arrow keys on the slider · Enter to submit

Closer guesses. Higher scores. Up to 1,000 points per round.

How to play & what you’re estimating

Make the call

Choose a value with the slider, then lock it in. The amber regression line appears after your guess. Complete five rounds and compare your estimates with the exact sample values.

R² or correlation?

R² is the fraction of sample y-variance explained by an ordinary least-squares line with an intercept. Here, R² = r², where r is Pearson’s correlation. r preserves the direction; R² does not. Neither establishes causation.

Scoring

Points = round(1,000 × max(0, 1 − 3 × normalized error)). Normalized error is the absolute error divided by the metric’s range: 1 for R², 2 for r. Scores use full precision; displayed answers are rounded.

Scores last for this game only. Refreshing starts over. All points are generated in your browser.