Click two places on the graph and then click 'Plot line' to draw your line. The score is calculated as the absolute difference between the mean squared error of your line and the mean squared error of the least squares best fit. A lower score indicates a better guess. Only the first attempt at each dataset will count towards the top score.

Warning: Points chosen around the boundaries of the graph may not appear on the graph.

Data description

Linear regression parameters

The following plot shows generated data with a given linear correlation value.

Note that the data generated here are already standardised, that is x and y both have mean 0 and standard deviation 1.

Created by Aidan Boland.