How is The WOT Project doing after six weeks?
July 28, 2011 4 Comments
Back in March I blogged about Web of Trust and how it could be a powerful tool for skepticism. Web of Trust is a crowdsourced web site safety rating system, that can warn unsuspecting internet users (and now, everyone on Facebook) when they are about to visit a site that contains scams, malware or other potential danger. I suggested that by rating sites selling products based on superstition or pseudoscience, skeptics could turn WOT into a tool for skeptical outreach. Indeed, many skeptic targets such as PowerBalance already sport negative WOT ratings.
In June Canadian skeptic Erik Davis launched a site called The WOT Project. His focus is the opposite side of the equation: protecting the WOT ratings of skeptic sites. Each week the WOT Project posts a list of skeptic sites and encourages participating skeptics to give them a good rating in WOT. The sixth such set was published on Monday.
Since WOT has an API, the ratings can be measured over time. Since my two most recent blog posts were about measuring skeptic outreach on Wikipedia, I thought it would be appropriate to do this for WOT as well. So let’s see how well has WOT Project done in their efforts to protect skeptic web sites on WOT.
My first blog post on WOT contained some stats on WOT’s current ratings for skeptic and believer websites. I generated those using a script to pull the WOT ratings en masse using their published API. So I have data on the WOT ratings of several hundred skeptic websites as they stood in March that can be used as a baseline.
Of the 50 websites posted in the first five WOT Project posts, I had data from March on 43. I then used the same script to pull current WOT ratings for all 50 sites. It is straightforward to compare the two sets of data. (I didn’t use the sites posted in WOT Project Volume 6 because they were just posted on Monday – there hasn’t been much chance for them to improve yet).
WOT compiles four scores (trustworthiness, reliability, privacy and child safety) with a confidence value reported on each score. As before, I’m going to focus on the trustworthiness score here, as it is clearly the most relevant to skeptic sites.
Here’s how the sites stood back in March, click for a larger version:
As you can see, we weren’t in bad shape to begin with (as my original blog post had indicated) – a total of 88% of the sites had Good or Excellent trustworthiness, and thus would have a green icon in WOT. But at the same time about 12% had either bad scores or no score at all.
Confidence values are not terrific, with only about 28% of the sites having high confidence scores (top two results). A larger number of sites have zero confidence (33%). That’s not good because it means our site scores could potentially be changed easily by those with an axe to grind about skeptics.
Here’s how the 50 sites in the first five WOT Project sets stand as of July 25 (about 6 weeks of crowdsourced work). Again, click for a larger version:
Clearly, a big win. All fifty sites now have green icons, and a much larger percentage have “Excellent” as opposed to “Good” trustworthiness.
Confidence is much increased as well, with no sites left with zero confidence, and 42% of them up in the highest confidence tier.
Although many skeptic sites already had good ratings in Web of Trust, The WOT Project has clearly demonstrated its value in its first six weeks. Overall ratings have been improved, and none of the selected sites show yellow or red icons any longer. Confidence values have improved as well, presumably helping to insulate the ratings of skeptic sites against malicious attacks by believers.
The WOT Project is an excellent example of an online skeptical tool with a simple goal that makes great use of crowdsourcing. I congratulate Erik on his efforts and encourage all web-surfing skeptics to get involved.