x-axis is time, y is approval/disapproval, width of line represents percent neutral. But then where's the excuse to play with the neat video library?
Maybe it would make more sense to show approval (or disapproval) and undecided?
The numbers support this: on average, 32% of the population is neutral towards a newly elected president. After 200 days, this has gone down to 19%. The same numbers for approval ratings remain steady: from 64% to 63%, while those for disapproval necessarily go up: from 4% to 18%.
This means that even the people who voted against the winning candidate (always between 40% and 60% of the turnout[0]) are initially mostly neutral towards them. But while the people who voted for him stay loyal for a bit (possible because they don't want to feel they've been deceived), the people who voted against him have no qualms about voicing their disapproval.
I have to admit that this "let's wait and see" attitude was pretty surprising to me. It seems like people refrain from judging a president until after he has actually made policy, as opposed to immediately upon taking office, as I would have expected them to do.
But to answer your question, I'd imagine that a plot of unemployment rate vs. approval rating would be quite enlightening.