Creative counting on github?

I like to look at the statistics of my projects – not that they are overly popular, but it is still interesting (gratifying as well!) to see spikes in activity. However, I’m just not sure what underlying arithmetics github uses for counting. Have a look at the following graph:

github_clones

Github reports 7 clones and 6 unique cloners in the two week window. Yay! But wait, where does the 7 come from? Counting the peaks, you have either 6 or 5 of them. With neither axis you can arrive at the magic seven… To me it looks more like that there were 6 clones and 5 unique cloners.

Thanks github, for making me feel better by adding one ghost clone! 🙂