💥 Outlier Hunter – Find the Point That Breaks the Model

Three rounds of increasing difficulty

In each round you see a scatter plot and the current best-fit line. Identify the outlier by clicking the correct point in the list (or reasoning from the graph). Then decide whether removing it improves the model. You must think — the outlier is not always the most extreme-looking number.

Round 1

–

Current MSE

–

Current R²

Click the point you believe is the outlier:

Outlier Hunter Complete

You learned that:

Outliers can dramatically change the slope and intercept.
Removing a true outlier usually lowers MSE and raises R² for the remaining points.
Always investigate why the point is extreme — it might be a data error or a genuinely special case.
Never remove points just to make the graph look prettier; have a reason.