I gather this posting is a tad tongue in cheek. There was a Kenneth Deffeyes predicting Thanksgivings Day 2005 as the date for peak oil. A prediction which itself was a bit tongue in cheek according to Mr. Deffeyes himself.

In person he would say he made the prediction tounge-in-cheek.  He then says it could be a few weeks to either side.
He made the prediction (and the error bars) based on his application of the linearization method, but both the prediction and the error bars are only as good as the data he used. If his data aren't quite complete or accurate, then neither are his results. But it's a fun idea anyway!