Who won the Niblets? If you revisited the voting page late in the voting and looked at the poll results, you may already know. But I have a few different takes on the data that weren’t in the poll results. So I hope you’ll keep reading.
Preliminaries
There were a total of 1347 ballots cast in the 2008 Niblets. The WP-Polls plugin records each voter’s IP address, so I’m counting a ballot as an IP address from which at least one vote was submitted.
In just looking at the data, I found a couple of cases that appeared to be clear ballot stuffing. Now I know that many of us vote more than once in the Niblets if we have access to more than one computer. But these were much more egregious. Several repeated ballots from the same IP address range (i.e., differing only in the last 3 digits) were submitted, separated by only 2-3 minutes. I assume this means the voters were able to force their ISPs to reassign a new IP address at will, and they were using this ability to cast many extra ballots. But whatever the reason, I excluded extra ballots submitted from the same IP address range within a one hour window. There were 25 such ballots, leaving a final ballot count of 1322.
The first question I looked at what how many ballots answered each of the 19 questions (note that for Question 19, which allowed as many as three answers to be chosen, I counted a ballot as answering it if at least one answer was chosen):
Voters clearly found some questions more interesting than others. As you can see, the questions about blogs (Q1-4) were answered most, and at the other end, we may have heard enough about Mitt Romney (Q15).
Also interesting is the fact that even the most frequently answered question was answered on only about 60% of ballots (Q1: 796/1322 = .60). This was really unexpected. I wondered if perhaps the same person’s vote was inadvertently being spread across multiple IP addresses—if their ISPs had automatically reassigned them in the middle of a ballot. I looked at IP addresses for a bunch of incomplete ballots cast close together in time to see if there was a pattern of similar IP addresses answering complementary sets of questions (e.g., one ballot answers Q1-5, the other answers Q6-19). I found no evidence of this pattern at all.
The only other explanation I could come up with for so many incomplete ballots was the one already discussed on the Niblets voting thread: single-issue voters. That is, lots of ballots were cast by people who visited only to ensure that their chosen candidate won on a single question. To check this possibility, I looked at how many of the 19 questions each ballot answered:
Wow! Over 30% of ballots cast (411 of 1322) answered only a single question. And the second most popular number of questions answered was two. Clearly there is something to this explanation.
Let me look at one more figure before I get to the results. So which question did the 411 single-question voters answer?
So it looks like the one-question voters cast lots of votes on the blog questions (Q1-4), some on Best doctrinal post (Q12), and a few on the last four questions (Nicest evil villain, Navel-gazing, Biggest contribution, Write-in) but almost none on any of the questions in between. The first four and the last four groups of questions kind of hang together, but I don’t know why all the love for Best doctrinal post.
Results
In order to account for single-issue voters, I’ll count votes in three ways:
- Using all 1322 ballots
- Using only the 911 ballots that answered at least two questions (thus excluding the single-question voters)
- Using only the 373 ballots that answered at least half of the questions (9+ of the 19), to get at the people most interested in the majority of the Niblets (and therefore ideally those interested in the Bloggernacle community more broadly)
For some questions, these different vote-counting approaches didn’t lead to different rankings, so I’ll only show one set of results. But for the questions where the rankings are different depending on how votes are counted, I’ll show them all, as I’m sure they will provide a basis for many fun arguments. 
Q1: Best big blog
Past winners:
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ or 9+ questions, so I won’t show the figures for them.)
BCC wins for the third year in a row! Who can displace them?
Q2: Best group blog
Past winners:
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
(Answers that moved up or down are bolded. Arrows show which direction they moved relative to the previous figure.)
Depending on which set of results you like, either Segullah or Juvenile Instructor gets the Niblet. Past winner ZD is booted all the way down to fourth (or third). Was it something we said?
Q3: Best solo blog
Past winners:
- 2005: Dave’s Mormon Inquiry
- 2006: Dave’s Mormon Inquiry
- 2007: Category not used; Dave’s Mormon Inquiry received more votes than any other solo blog in the “Best small blog” category
2008 results:
Counting only ballots that answered 2+ questions:
Counting only ballots that answered 9+ questions:
Ardis Parshall’s Keepapitchinin breaks DMI Dave’s strangehold on the category! But then, I guess this was kind of anticlimactic since Dave wasn’t even nominated. Perhaps we lost track of him in his move to Beliefnet?
Q4: Best new blog
Past winners:
2008 results:
Counting only ballots that answered 2+ questions:
Counting only ballots that answered 9+ questions:
Another win for Ardis’s Keepapitchinin!
Q5: Best blog layout/graphics
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
Depending on which results you prefer, either My Regis Blog or Nine Moons gets the Niblet.
Q6: Best commenter
Past winners:
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
I previously identified Ray as the most prolific commenter (in 2007). Now he’s also the Niblet winner as the best commenter.
Q7: Most memorable comment
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ or 9+ questions.)
DKL takes both first and second place! I guess that’s similar to taking a one-two punch from Jesus. 
Q8: Best overall blogger
Past winners:
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
Ardis, not satisfied with taking best solo blog and best new blog, also takes best overall blogger, and by a sizable margin.
Q9: Best humorous post
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
Here’s a question where the results differ quite a bit depending on how the ballots are counting. When all ballots are counted, TAMN of Seriously So Blessed is the narrow winner. When only ballots answering 9+ questions are counted, one of the BCC Police Beat Roundtables is the winner by a pretty large margin.
Q10: Best historical post
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
Ardis also takes best historical post, adding to her virtual trophy case.
Q11: Best spiritual post
This category hasn’t been used in previous years, but the “best post” winners in both 2005—Wilfried Decoo’s “Coffee“—and 2006—Ardis Parshall’s “Dressing the Dead“—fit well into it.
2008 results:
Counting only ballots that answered 2+ questions:
Counting only ballots that answered 9+ questions:
Either Kristine or Russell Arben Fox prevents Ardis from winning a Niblet in every single category in which she was nominated.
Q12: Best doctrinal post
This category hasn’t been used in previous years, but the “best post” winner in 2007—Kristine’s “Why I Liked President Beck’s Talk (Mostly)“—might fit into it.
2008 results:
Counting only ballots that answered 2+ questions:
Counting only ballots that answered 9+ questions:
The Niblet goes to AdamF or Julie M. Smith. It looks like lots of the single-question ballots that answered only this question were voting for AdamF’s post, as it fell dramatically in the rankings when only ballots that answered at least two questions were counted.
Q13: Best current events post
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ or 9+ questions.)
TAMN of Seriously So Blessed takes the Niblet!
Q14: Best SSM/Prop 8 post
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ or 9+ questions.)
Caroline of the Exponent takes the Niblet!
Q15: Best Mitt Romney post
This category hasn’t been used in previous years. It should be noted, however, that “Most overdone theme: Mitt Romney” took third in the write-in category in the 2006 Niblets, and in the 2007 Niblets, “Mitt Romney ad nauseam” was voted the worst thing in the Bloggernacle and “Theme to be least missed in 2008: Mitt Romney” took second in the write-in category.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
The Niblet goes to Scott B. Or to Scott B. in a tie with Bruce Webster.
Q16: Nicest “evil villain”/wolf in sheep’s clothing — Now corrected to include Andrew S!
This category hasn’t been used in previous years.
2008 results:
Counting only ballots that answered 2+ questions:
(There were no differences in the rankings when looking only at ballots that answered 9+ questions.)
The Niblet goes to Bridget Jack Meyers. She clearly must have very realistic-looking sheep’s clothing if she’s able to fool the hard-nosed T&S crowd into having her guest blog!
Q17: Most blatant example of navel-gazing
This category hasn’t been used in previous years.
2008 results:
Counting only ballots that answered 2+ questions:
(There were no differences in the rankings when looking only at ballots that answered 9+ questions.)
That’s two and a half Niblets (the half is for Best humorous post) for Seriously So Blessed if you’re keeping score. Ardis still has the lead with four.
Q18: Best contribution to the Bloggernacle
This category hasn’t been used in previous years.
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ or 9+ questions.)
Tracy M takes the Niblet!
Q19: Write-in
Past winners:
- 2005: (The category was not used.)
- 2006: Best podcast: Mormon Stories
- 2007: Best controversies: Mormon grad students on welfare (Devyn S. at Mormon Mentality); Waldorf style nursery (Artemis at FMH)
2008 results:
(There were no differences in the rankings when looking only at ballots that answered 2+ questions.)
Counting only ballots that answered 9+ questions:
BCC’s sideblog takes the Niblet!
Concluding Bits
With the voting dataset assembled, I was able to ask two more potentially interesting questions. I’ll share what I found in case you’re interested too.
First, on what days were most ballots cast? I expected that there would be many cast at the beginning, and then fewer with each passing day. This is what we got:
So I was approximately correct, but I didn’t anticipate the small recovery near the end. Perhaps this is when everyone put in their single-question ballots!
Which questions did voters spend the most time on? The WP-Polls plugin records the time each question is answered, but only to the nearest minute. As you might guess, most votes were cast less than a minute apart. I tried looking at the average time between consecutive questions, but these values are highly skewed, and influenced by a few people who appear to have left the page entirely in between questions only to return another day to finish. I ended up settling on counting how many people spent more than one minute between each pair of questions. Since varying numbers of people answered each question, I looked at this as a percentage:
Note that the figure starts with Question 2 because I don’t have a time before Question 1 to count from.
It’s difficult to draw inferences from this figure because the percentages are based on only those ballots where people voted for two consecutive questions, which as we saw at the beginning, is a fairly small number of ballots. Those who are missing from this calculation may have in fact pondered one of the questions at such great length that they never got around to answering it. But to the degree that we can make an inference, it looks like voters ran quickly through the beginning questions about blogs, but slowed down with the later questions. This would make sense given simply the amount of material to read even if you only read the post titles and comment summaries, for example. The question voters were most likely to spend some time on was Question 7, Most memorable comment. Perhaps this suggests that they were following the links to go read the comments, which would have been good. I can definitely see doing this while not doing the same with the nominated posts, as posts tend to be so much longer than comments, and I tend to have a short attention span.
Okay, the floor is yours now. Did your favorite blogs, bloggers, or commenters win? Or were they unfairly dismissed? Is there more (or less!) you’d like to see in next year’s Niblets?
