Posted in:
ABA Blawg Contest Rigged?
Turns out Slaw didn’t do so bad in the ABA blog awards Tech category. After a recount, we’ve placed second!
Now for the fun part… Google cache (while it lasts) shows a significant number of votes were removed from the system from the time the contest closed and the final numbers being released. (Hat tip to Bob Ambrogi)
Before the recount:
- FutureLawyer, 1930 votes
- Technolawyer blog, 1545 votes (Corrected as noticed by Neil Squillante below)
- Mac Lawyer, 509 votes
- Ross Ipsa Loquitur, 348 votes
- Slaw, 317 votes
- Jim Calloway, 169 votes
- Real Lawyers Have Blogs, 132 votes
- Ernie the Attorney, 93 votes
- Dennis Kennedy, 81 votes
- Inter Alia, 64 votes
AND after the recount:
- Technolawyer blog, 1499 votes
- Slaw, 295 votes
- Mac Lawyer, 256 votes
- FutureLawyer, 241 votes
- Jim Calloway, 150 votes
- Real Lawyers Have Blogs, 105 votes
- Ross Ipsa Loquitur, 102 votes
- Ernie the Attorney, 93 votes
- Dennis Kennedy, 82 votes
- Inter Alia, 64 votes
Not to throw stones without blame here — we did have 22 votes removed (for which I’ll take full credit – I voted on my mother’s computer at Christmas, and once on IE, FF, Safari, etc.) — but… I KNEW there must be some shenanigans going on.
Top marks to ABA Journal staff for checking duplicate IP addresses. Well done, and crisis averted! … At least for the contest itself. :)
Nice analysis. One correction. Our raw vote total was 1,545. We lost 46 votes after the recount. Congratulations on receiving mostly honest votes as we did.
Thanks for the correction Neil. All in all, a nice testament to the TL community. Congratulations. :)
Thanks, Steve. This is good news: I’m happy to be runner up to TechnoLawyer any day. (Good thing I didn’t give in to my impulse to ask people to do the Mayor Daley thing: “Vote early, vote often.”) And thanks to all of those who voted (once) for Slaw!
Ah! I was able to vote multiple times from my same computer, same browser. I thought it was set up that way. Still, I probably only voted about 3-4 times total. Others may have done the same, however.
FWIW, “rigged” isn’t really the word I would use for this. I was looking forward to something genuinely seedy – on the part of the ABA – but per Connie’s comment, this seems to have been more a question of (relatively) innocent incompetence than “rigging.”
Btw, how did Dennis Kennedy’s tally *increase*?…
Gee, I don’t know: 1700 duplicate votes is getting close to seedy.
I really liked the ABA Journal’s headline on this issue: “Some People Love the Blawg 100 a Little Too Much”
The problem is that hundreds of law students voting from the same law school would still have the same IP.
IP based voting controls are not perfect either.
Same would go for networks that only have one external IP address, which would include many firms.
Omar’s point is well taken. Law schools and firms coming off the same proxy server will have the same IP address, and an email to a local group to drum up support (which should be fair game) could have a similar effect.
However, duplicate IPs weren’t the only factor considered. From the ABA post Shauna mentions, they noted multiple votes in short succession (read: seconds & automated) coming from the same IP, along with C-class IP ranges from overseas. Plus there was a human review by the ABA.
So while IP voting isn’t perfect, I’d say they did their diligence, and injected human eyes when it helped. FWIW.
Dennis Kennedy’s vote total didn’t really increase. The Google cache Steven found is from a few hours before the voting ended. That’s why he had the wrong raw vote total for us as well.
Regarding the IP address issue, that’s above my pay grade but I believe the ABA allowed a certain number of votes per IP address and that they also looked at patterns. Honest voting tends to look random whereas someone deleting cookies, voting, deleting cookies, voting, etc. occurs in a recognizable pattern.
One thing I do know a lot about is email. We knew that some of these votes were not coming from email (as ours were).
Connie, I think the ABA’s primary concern was someone voting hundreds or thousands of times..
How can they tell they’re “automated”? If, in fact, a law school was responsible, it would be entirely possible for a large class, upon hearing from a classmate about the voting, to all vote in a short period of time.
When I was in my last year of law school, wireless access was added. During class, many students ran chat programs like Instant Messenger and passed information around.
And, yes, how did someone’s tally actually go UP?