Answer:
Think of MAD as the mean of a mean, or the average distance of a point to the mean. You primarily use it to see the spread of the digits, so if there was a plot made of digits: {1,1,2,2,2}, and a plot made of digits {-10,0,1,17}, the mean would be the same, but the MAD would be very different. This way, when you look at the two data plots, without knowing the digits, you can use MAD to know if the digits are closely or widely spaced, as opposed to not being able to know just knowing the mean.
Let me know if you have any other questions!