I would say only if one of your data points is the origin. But your experiment could have started with a non-zero velocity, for instance, which would rule out the origin as one of your data points. Even so, a "best fit" is not meant to be perfect, it is only meant to be the best that you can do with your particular data set.
