As most of you now know, Climate Audit has been declared co-winner of the 2007 Weblogs Award for Best Science blog (together with Bad Astronomy.) This decision was made with the agreement of all parties.
Some of you have been understandably a little puzzled and seeking to interpret the matter. Here’s a bit of the background.
There was an incredible volume of voting for Best Science blog on Nov 8. Eventually both blogs surged to vote levels much higher than even high-trafffic political blogs such as Arianna Huffington and Michelle Malkin. You can see some interesting real time reports here.
On the afternoon of Nov 7, the situation was close, but quiet, with BA having a very small lead. At 3:44 pm blog time (4:44 pm EST), the vote was:
CA 6,349 BA 6,631
Voting on Nov 7 was plagued with difficulties as there seemed to be problems with the 2007 Weblogs Awards site and all parties had trouble getting through. These problems were resolved some time between the afternoon of Nov 7 and the morning of Nov 8 (closing day – 5pm closing). One reader reported at 7:23 am EST (6:23 am blog time):
CA leads 8602/8586!
An hour later at 8:16 am EST (7:16 am blog time) after another reader reported the following:
CA: 9232 BA: 8751 Pharyngula: 5353 JS: 3532
By 9:30 am EST (8:30 am blog time), CA had increased the lead to nearly 1000 votes:
CA 10208 ba 9264
Between then and 5 pm EST (closing), both blogs continued to amass votes at about 1300 votes per hour – at almost an identical rate. At 5:03 pm, a reader reported the standings at:
CA 20131; BA 18907.
Votes changed somewhat over the next few minutes, which people at the time thought might have been due to caching. So by 5:17 pm EST, the votes had changed a little to:
CA 20242 BA 18983
In a post published just after closing, I refer to votes at closing of
CA 20242 BA 18993
However matters didn’t end there. After the closing of the vote at 5 pm EST, Bad Astronomy continued to narrow the gap – organizers later advising us that someone had hacked into their system. The progress of post-closing votes was documented by a poster at the Weblogs forum here:
3:59 PM …18828….17294…1534
4:16 PM …19259….17798…1461
4:44 PM …19735….18575…1160
4:53 PM …19783….18755…1028
5:03 PM …20131….18907…1224
5:17 PM …20242….18983…1259
5:31 PM …20247….19060…1187
5:54 PM …20555….19144…1411
6:22 PM …20613….19175…1438
6:34 PM …20615….19181…1434
8:20 PM …20634….20681….-47
A poster at CA here showed some further details on the Bad Astronomy surge after 8 pm EST with a post at 8:19 pm EST (7:19 blog time):
Numbers just a bit ago: CA: 20,631 BA: 19,219
Just refreshed: CA: 20,632 BA: 20,673
The voting was finally shut off just after Bad Astronomy got into the lead.
I went out to play squash league (with dinner and beer) after the vote had closed and returned about 5 hours later to find Bad Astronomy in a slight lead. I thought that this was rather amusing: what would an American election be without hanging chads? The organizers were aware of the problem and had undertaken that the post-closing votes would not count; however, they thought that they would not be able to sort out the problems until this week.
My size-up was more or less as follows. The post-closing votes would be relatively easy to back out and there was little doubt that CA had a healthy lead of about 1200 votes at closing. The post-closing votes for both blogs strongly indicated the presence of bot votes for both blogs – something that had obviously crossed many people’s minds as the huge voting volumes accumulated through the day. The vote had obviously attracted interest, but how much interest? The pace of votes may also have seemed implausibly close (but the Best Technology blog also had a very close and hotly contested race.)
Given the bot activity in the post-closing vote, this was certainly an indication of bot activity pre-voting that would be sufficient to trigger an audit in any election where there was something on the line e.g. a congressional race. It seemed to me that it would be a formidable job for the organizers to try to determine bot activity pre-closing, that whatever determination that they made would simply lead to bad blood. The contest was supposed to be fun and to bring attention to blogs – something that the contest had succeeded in doing.
Accordingly, after midnight EST, I sent an email to Phil Plait suggesting that we agree to a draw, something that he was willing to do. We then agreed to jointly send a letter to the organizers (that I drafted at Phil’s suggestion) thanking them for organizing the competition, noting the spirit of the competition and suggesting that they declare a draw. The organizers agreed with this suggestion with alacrity – actually that doesn’t describe accurately enough how quickly they agreed. It wasn’t obvious what number should be picked for the vote total – I suggested that a round number of 20,000 each be chosen and that a note on the record state that the parties had agreed on the matter.
There’s an old saying in business: bulls sometimes win, bears sometimes win, but pigs never win (or hogs get slaughtered). Phil Plait said that he wanted the award to help promote an upcoming book: it’s no skin off my nose if a young guy like him gets a boost with his book and so the tie was a simple alternative. Everyone gets on with their business and the arduous audit procedure for something where there is no money involved is avoided.
My negotiations with Phil Plait were very cordial. The one thing that frustrated him about the result appeared to be that no one would ever know who “really” won. This is a state of affairs that’s probably more familiar to me than him. There are lots of things where it’s impossible to know what “really” happened. Because BA got more [bot] post-closing votes than CA, one could guess that they also got more bot pre-closing votes than CA, but it would really just be a guess.
If CA had been declared a winner based on 5 pm closing vote and the results of the vote were being used in something like an IPCC assessment report, then I’m sure that many people would have called for an audit of the results. But not of a temperature reconstruction. Just one of life’s many ironies.