Let’s start off by acknowledging that this data would have gotten me kicked out of statistics class back in grad school days. Our sampling is random, but it might also include far fewer than 114 people, despite the 114 responses: it’s the interwebs, and we have no idea who really voted!
Plus, of our 115 voters, only 6 people definitely read all 10 books, and one of those 6 forgot to actually vote (despite answering how many he or she had read). Then, because of our own lack of thinking this all through, we also don’t even know how many books our first 22 voters read, since that question wasn’t there at first.
And–yes, there’s more!–we had a handful of votes where only first or first and second places were filled out, which skews the data, since a book with dozens of third place votes can outpoint a book with a few first place votes, all else being equal.
So, you know, bad bad data.
But hey, why let that slow us down?