Tracking Break-Ups via Facebook Status Updates

Here is a CNN.com article about a journalist and graphic designer who minded Facebook status updates to try to discern when people broke up the most. (And, of course, this is a sensible way to do that. On a similar note, one of my collaborators is involved in a project that tracks the nation's moods---and, e.g., election updates---by following tweets.) However, I don't understand the plot that goes with the article. There is supposed to be a spike around Christmas, but the graphic that goes with the article clearly does not show such a spike. What gives?