File talk:Clickthrough rate vs. degree and position.png

Regarding the Relative Link position graph, the initial elevated values are clearly hatnotes and the Lede of the article, possibly infoboxes and other high importance templates. The last two values reflect junk that sometimes end up below the refs section... dismal value collapsed navigation templates, external links, and other random junk. The very interesting thing about the graph is how amazingly flat the stats are for the body of the article. My initial guess would have been at least some downward slope. Alsee (talk) 20:08, 2 March 2016 (UTC)[reply]

I was surprised by that too. The only possible explanation I can think of is that people very often jump to the section of their interest by using the TOC; alternative explanations (the text is skimmed regularly or is read from word by word, cover to cover) seem unlikely. Nemo 10:44, 24 March 2016 (UTC)[reply]