What does a residual ei=yi−y^i=0e_i = y_i - \hat{y}_i = 0ei=yi−y^i=0 imply for that specific data point?
The model perfectly predicts the observed value.
The point is an outlier.
The model is biased.
The slope is zero.