tedster - 2:54 am on May 24, 2011 (gmt 0)
I absolutely agree that we need to keep things as simple as possible. However, even before Panda people were calling every ranking loss a "penalty" and looking high and low for what they did wrong. And that use of the word "penalty" was not fruitful either for people who needed help. Simple and over-simplified are not the same thing.
With algo changes, it's often the absence of something right, rather than being "punished for your sins." A Panda loss cannot be removed by any process parallel to escaping a penalty. A penalty is removed by 1) time itself, 2) undoing a guidelines violation, 3) a reconsideration request.
But escaping from Panda does not appear to be about removing some violation, it's about creating some new factor that registers positively in this algorithm. So I don't think the penalty model is a fruitful way to think about the challenge and how to meet it. Not all ranking drops are penalties - and they never were.
The thread's title asked if it's a time penalty. No, it is not. It also asks if it's a trust thing. That's a bit closer, although Google already had trust calculations in the algo. I have heard about good new backlinks helping to lift pandalyzed sites. If that's true, then conventional trust may play a part.