Professor William Irvine created a python simulation that produces mock data and then plots and fits that data to show how uncertainties contribute to the quality of a fit. The program allows you both to alter the statistical distributions from which each data point is drawn and to add systematic shifts to the data. In this way, you can explore how changes in the types (and magnitudes) of uncertainties lead to changes in the overall chi-squared distribution.
Play around and explore! mock_data.py 2 |video
 |  |
 |  |
 |  |