I don't think you need all, I mean, there were space constraints. This is why a bar graph can be clearer than a line graph for discrete data though. They may be boring, but they are often effective at communication. :-) Oh and it's helpful to label both ends of the graph, from Q1 2005 to Q4 2014, then fill in what labels you can in the middle. If you'd picked random Qs or even just every year, perhaps to shade the backgrounds of each year, that might have made for a more understandable graph. Our fault for misunderstanding it in the first place, of course... This is also why some places link to tables of data used in the graph. You could click to view the table if you had any questions about what you're seeing in the graph. Stephen Few's an accessible read on this subject.