Why do the linear regression and the exponential regression give different results

Ted and meg have each drawn a line on the scatter plot shown below: the graph shows numbers from 0 to 10 on x and y axes at increments of 1. dots are made at the ordered pairs 1, 8 and 2, 7.5 and 3, 6 and 4, 5.5 and 4.5, 4.8 and 5, 4 and 6, 3.5 and 7, 3 and 8, 1.9 and 9, 1.2 and 10, 0. a straight line labeled line p joins the ordered pairs 0, 8.5 and 10.1, 0. a straight line labeled line r joins the ordered pairs 0, 8.5 and 7.8, 0. which line best represents the line of best fit? line p, because it is closest to most data points line p, because it shows a positive association line r, because it is closest to most data points line r, because it shows a negative association