Lens Harbor

Found something interesting when looking at lens stats for one of my bottom dwellers recently. The lens was created in January, 2009. I had an idea, built the lens and made a couple edits through January. Then did nothing with it. No promotion, no updates, nothing, and it got zero traffic for months.

You can see the effect of all this inactivity on lensrank by looking at the smooth curve between section A (when I updated the lens) and section B in the screen shot below. That’s the lensrank algorithm at work, steadily lowering my lensrank over time. It’s very graceful, actually.

Then in April the lens got its first visitor and lensrank soared about 74,000 spots. It got another hit or two into May which helped it tread water for a while. Then the rank dropped, the effect of all that traffic had expired.

But look at the right hand part of section B on the chart. Instead of going back to the same point it was at before the visits, the lensrank dropped further. It went all the way down to the point on the curve where it would’ve been if no one had visited the lens. Connect the lines under the traffic blips and you’ll still see that same graceful curve in action.

This pattern repeats in section C on the chart when the lens had brief periods of traffic in June and July. There were no clickouts, ratings or interaction with polls or guestbooks for the lens during the chart’s time period so we can eliminate those as variables.

What does this mean?
1. The effect of traffic visits are short lived. Once that effect is over the lensrank returns to where it would’ve been if the traffic never existed.

2. Lens updates matter. No surprise here, but notice in section A that lensrank got a bump from my updates and continued the curve from that new higher point. It didn’t revert back to the previous curve like it did for traffic. Or maybe it does revert back but at a much slower rate.

For the record, I’m not trying to solve the lensrank code. The behavior I’ve casually observed over 3 years is mostly consistent with what I’d expect from a ranking algorithm. Even so, I was intrigued at seeing such a clean curve.

Now it’s time to get that lens out of the cellar…