memory-based model evaluation

hey all :)

I have a set of predicted and a set of real values for variable X in range {-2,2}.
the model used for the prediction was NOT regression but memory based (that is, based on a distance metric I chose the closest neighbor and assigned its value to the query sample) So I ended up with the two sets....

how do I say if the results are good or bad? are the values of RMSE and MAPE errors enough?

thank you :D